Durova

Telling knitting from crochet

2010-04-14T11:32:00.000-07:00

It came as a surprise three years ago when I started crocheting to hear people call it "knitting". Nobody who crochets or knits confuses the two, and most people who do a few textile arts can tell the difference at a glance. But a large number of people who don't do either craft haven't any idea how to distinguish them.

The sofa blanket and pillow above are both crochet. If you want to get fancy, the blanket is filet crochet. Filet crochet is a technique that uses two stitches of different sizes to create patterns, using the negative space generated by the smaller stitch for contrast.

If you see someone making fabric and are trying to tell whether it's crochet or knitting, the absolute giveaway for crochet is a hooked tool. If you aren't standing quite close enough to see whether there's a hook on the end, count the implements. One short tool is always crochet.

Crochet also holds very few stitches. If you see only one loop of yarn on the tool (or just a couple), you're looking at crochet.

Knitting uses at least two sharp ended needles with dozens of stitches all looped around the needles at the same time.

Knitting can be reproduced by machine, but crochet has to be made by hand. So everyone owns a lot of knitted items. T-shirts, socks, sweatpants, and mass produced sweaters are all knitted items. Crochet has its own distinct stitch patterns that won't look like anything in your sock drawer.

Knitting is good at creating soft stretchy fabric such as socks. Crochet tends to be stiffer so it's good for sturdy items such as placemats or tote bags. It's easier to make fancy patterns with crochet and crochet produces fabric somewhat faster than hand knitting.

Crochet is inherently denser than knitting and uses one-third more yarn to produce the same sized item. That sofa blanket in the photograph weighs about ten pounds.
----
Image credits:
http://en.wikipedia.org/wiki/File:Pink_knitting_in_front_of_pink_sweatshirt.JPG
http://en.wikipedia.org/wiki/File:Crochet-round.jpg

Neglected subjects

2010-04-13T18:17:00.000-07:00

This morning Peter Weis and I were chatting, and I mentioned that I was learning to use a knitting machine. He replied that Singer was famous for their knitting machines, and it was a little while before we realized that he was thinking of sewing machines rather than knitting machines. So if you've never seen a knitting machine before, this is what one looks like. Sewing machines join fabric together; knitting machines generate knitted fabric.

That red strip hanging down from the device is a piece of knitting. It curls at the side because it is a stockinette stitch (stockinette stitches like to roll and curl; that's how they behave). This knitting machine has 50 hooks threaded, half of which I'm using to produce this item. Each hook holds one stitch of fabric. The purple item at right is a shuttle, which pushes the hooks and creates knitting stitches as it moves across the hook bed. I operate this shuttle manually; higher end knitting machines can be punch card driven or computerized.

The advantage of using a knitting machine is that it can generate an entire row of stitches in the time it would take to produce two or three stitches with regular knitting needles. This device can create rows up to 100 stitches wide, and with accessory extensions I could extend that as wide as the work table.

Knitting machines aren't like regular knitting needles, though. This doesn't pack into a tote bag and carry to Starbucks. It needs a fairly large work table with specific dimensions and it has to be clamped down. It also requires engineering skill: a lot of things can go wrong with the assembly and operation.

Each knitting machine can only do certain things. The hooks are sized to a given size of yarn and spaced at a fixed spacing, which means that one machine could produce baby blankets or heavy sweaters, but not both. This flat bed knitting machine will never make socks (unless you like a seam in your socks, which nobody does--seams in socks are uncomfortable).

Before getting the flat bed machine I tried the smaller circular machine at right. It operates on a crank and generates thin rounded pieces of knitting that are suitable for cords and edging.

The smaller machine is worth getting if this intrigues you. This device doesn't work on as many types of yarn as the manufacturer claims. I've had success only with acrylics and wools in baby yarn and sock yarn. It's almost a necessity to have a lace making crochet hook in order to correct for machine errors: these devices are prone to skipping and jamming.

Although once a piece gets started properly, this is far faster than manual knitting.

In recent years basic terms and concepts in textile arts have been fading from general knowledge. These days, "needlepoint" and "knitting" and "braiding" often get misapplied to things that are not needlepoint or knitting or braiding, because a lot of people don't know enough about these subjects to tell the difference. I've created a lot of the Wikimedia Commons textile arts categories; one has to know what one is looking at to sort the content.

I uploaded these photographs to Wikimedia Commons today; they might find their way into articles soon.

Wikipedia's Arbitration Committee implements genuine reform.

2010-03-31T17:25:00.000-07:00

Good news at last: Wikipedia's Arbitration Committee has implemented a new system that saves weeks and months of hassle without any loss in quality over their previous system. Check it out here.

Referencing comes full circle

2010-03-20T09:43:00.002-07:00

I wonder whether amigurumi was invented to keep guys away from things. Drop an amigurumi on the remote control, send a shiver down his spine, and he leaves the room to hunt for leftover pizza while a woman settles in happily to watch the shopping channel. The cuteness factor on these things is absolutely repellent.

One of the areas where Commons and Wikipedia need more attention is textile arts. It's not as if men never touch the subject (here's a photograph of WMF volunteer coordinator Cary Bass's crochet project), but let's face it: women are underrepresented among editors and it's usually women who take an interest in the subject.

If a guy's going to work on textile arts it probably won't be amigurumi. It's a crochet form that originated in Japan to make miniature animals, such as the owl at right. I don't make them, but the quantity of amigurumi photographs at Wikimedia Commons had reached the point where a category was justified. So I made a new category today. It's the sort of housekeeping work that makes Wikimedia Commons useful and anyone can do it on subjects they happen to know.

Back at Wikipedia I added a {{Commonscat}} template at the amigurumi article to link to the new category. Then I looked at the article, which is a dreadful little stub whose only reference is an AOL blog that probably isn't reliable. So I looked for something better. Mostly the references on this type of subject aren't very good: they tend to be how-to books with a lot of patterns and minimal context. But that's better than nothing, right?

Um...

Here's an excerpt from the top return on Google Books: Amigurumi!: Super Happy Crochet Cute by Elisabeth A. Doherty

In Japan a crocheted or knitted doll is called an amigurumi (ah-mee-guh-ROO-mee). Disclaimer: Please, people. I do not speak Japanese. This pronunciation guide should be taken with a grain of salt. There is very little information available in English on Amigurumi, so I have decided to go with my own definition based on information I found in an online encyclopedia. Ami is a shortening of the word amimie, which means “stitch.” Gurumi is a shortening of the word nuigurumi, which means “stuffed doll or toy.” Smoosh the two together and you get amigurumi.

The book was published in 2007 and that's reasonably close to Wikipedia's definition as it appeared at the end of 2006. I don't know if we'll ever get a reliable source on this stuff, but I'll keep looking.

The good, the bad and the ugly

2010-02-25T20:03:00.000-08:00

The following is a guest post by Peter Weis:

What is chromatic aberration?
“In optics, chromatic aberration (also called achromatism or chromatic distortion) is the failure of a lens to focus all colors to the same point”, is the first sentence in the Wikipedia entry on chromatic aberration. Basically light enters a lense in the same, aligned direction but due to the mechanical advice and its imperfection – the light leaves the lense in several unaligned directions.
We should leave the technical question at this point as well. For further information I recommend the full article.

How does that affect an image?
Chromatic aberration looks like “purple fringing”. On the image below you can see longitudial chromatic aberration. In this case all edges of your subject have purple fringes. If there are additional green fringes you speak of lateral chromatic aberration. Again for further technical information I recommend reading the article on purple fringing.

The good

The most convenient solution for chromatic aberration is using Adobe's Camera Raw. For those amongst you who use Photoshop but never headed for Camera Raw here you go.

Camera Raw is freeware that enables you to work with the raw material from your DLSR. To explain the full workflow and the advantages and disadvantages of Camera Raw would be enough information for a second article – so let us focus on dealing with chromatic aberrations and purple fringing. I recently edited a map of the South Pole which had some problems.

This image was cropped at 400% - the occurring greenish/reddish fringing is between 1-3 pixels. First step is to open your Raw file with Photoshop. If Camera Raw is already installed, this should be possible by double-clicking the file.

Note: If your file is not a Raw file you could nonetheless open it using Camera Raw. Here's a short how to:

Photoshop >>> File >>> Open as…

>>> [yourfile].[yourimageformat] >>>

Open as: Camera Raw

In the next step the Raw dialog should open up. The third tab from the right is “lens correction” – that’s where we will start working. The section “Chromatic Aberration” is almost self-explanatory. Depending on the colour fringing of your image you either adjust red/cyan or blue/yellow. The larger the colour fringing the easier to remove it – but don’t panic if otherwise.

In this case you could use “Defringe: All Edges” or “Defringe: Hightlight Edges“. “Defringe: All Edges” will remove the colour fringing, but also results in subtle grayish lines. For my image this solution appeared to be the optimum. The defringed image tends to be a little bit desaturated, so I recommend you to slightly increase the saturation and if needed contrast. It might be helpful to have the original layer and the adjusted layer in your layers palette to directly compare what values of saturation/contrast fit the most for your project. “Defringe: Highlight Edges” is to be used if most of the colour fringing occurs in edges of highlighting, which is quite often the case. Just feel free to play around with the sliders.

After a few hours of work including temperature, levels, curves, saturation and restoring the map the finished version looks like this:

Writing articles for a change...and oh what a change

2010-02-24T16:55:00.000-08:00

Strange as it may seem, I've also been known to write for Wikipedia. This last day has been more interesting than usual in that regard. You're going to love how this post ends.

Found the topic from a short mention at an administrative noticeboard.

It appears that this article on a California company may be part of that company's pre-IPO publicity blitz. The company's website, linked from the article, shows a countdown clock for tomorrow at 9:00 am local time. Given that just a few contributors wrote most of the article over the past few days, is this cause for concern? User:LeadSongDog come howl 21:37, 23 February 2010 (UTC)

The good thing about blogging is it gives the chance to be more candid. So not to mince words, this looked like astroturfing. Here's how it appeared shortly before my first edit. Red flags for promotional writing include awards in the lead and a single sentence of criticism backloaded at the tail end of the article. The section entitled "Competition" used one source that basically stated the major players had recently pulled out of the field. If you check the references the quotes turn out to be cherry picked.

Actual press coverage is mixed: plenty of positives, also skepticism over feasibility and hype. The company was profiled by 60 Minutes a few days ago and has been doing a media blitz since then. It's a tech firm that wants to sell green energy and is very cagey about its company secrets. They held a press conference this morning with Arnold Schwarzenegger, Colin Powell, and Michael Bloomberg. Is this pre-IPO hype?

Astroturfing jobs at Wikipedia these days are getting contracted at one or two levels of separation from the client company. So the geolocation of this major contributor is less significant than its singlemindedness. Read the edits for yourself.

http://en.wikipedia.org/w/index.php?title=Special:Contributions&limit=500&target=58.179.137.71

Following are a few highlights from the IP's attempts to reinsert promotional content into company history such as a tangential NASA connection that the CEO had prior to founding the startup:
*(the idea came from somewhere....look at the Microsoft Page....)
*(one fine day in 2001, Mr Sridhar became an expert in fuel cells!)
*(the fucking Google page says where the name Google came from!!!!! ok, bitch?)

The last of those comments looks to have been intended toward me; the IP user appears not to have noticed that by this time the IP was actually edit warring with a male administrator. This IP editor doesn't seem to understand much about the underlying engineering or science. At one point I stopped to explain that hydrolysis consumes energy rather than produces it; shortly afterward the IP shuffled sentences so it looked like a venture capitalist was using a cost per kilowatt hour argument to rebut a criticism about fuel cell manufacturing costs.

The firm has $400 million in venture capital and a burn rate of $85 million a year. Everyone loves green energy--or would if it were cost effective--but this looks like a very high stakes gamble. Bloom Energy's business plan is based upon a bet that a notoriously unstable technology can be mass produced by a startup firm headed by a Ph.D. who never got any closer to NASA than managing a university department that built prototypes. Instead of Schwarzenegger and Powell, I wish that company had introduced its VP of production at the press conference today.

It turns out that Sprint Nextel owns 15 patents on similar fuel cell technology and has been using their own models in the field for five years. The Department of Energy gave Sprint a $7 million grant last year to expand their program. Quite a few other firms are active in the area. It isn't clear what Bloom Energy has that they don't--other than a star studded board and a hype generator. It could be the start of a new era or it could be the next Pets.com.

Among the interesting tidbits I found was a piece from Wired that had located a 2009 patent award to Bloom Energy for a device that seems very similar to its Bloom Energy Server. Perhaps the vaunted "secret ingredient". Did you know...

... that Bloom Energy was awarded a patent in 2009 for a power device that utilizes yttria-stabilized zirconia?

That lovely bit of information is on Wikipedia's main page right now.

Time to go back onsite and politely congratulate that IP coauthor. Entries at Wikipedia's "Did you know?" feature remain live about six hours. Wikipedia's main page typically receives 1 million page views during that time.

Lead room

2010-02-06T13:24:00.000-08:00

One factor to consider when cropping a portrait is lead room. When a subject looks to one side viewers become curious where the person is looking. Lead room adds a sense of meaning to facial expressions and gives compositions a satisfactory emotional balance. This 1935 portrait has a beautiful use of lead room: it was taken by photographer Ben Shahn in Jackson Square, New Orleans. A wide space to the left enhances the wistful expression in the subject's eyes.

To see what a difference the lead room makes, let's try cropping out part of the photograph so he looks centered. The effect isn't nearly so pleasing.

Human beings take cues from the eyes of other people. We want to know what catches their attention; there's an urge to glance in the same direction. Even when we know that we're looking at a very old portrait we feel that urge; it's instinctive. And although this subject's expression is exactly the same the cropped version seems caged; we glean less from it. The slouch and the knitted brow don't convey as much.

Last night the concept of lead room intruded on a search for historic Irish portraits that Alison and I were doing. She wanted Countess Markiewicz and we found a 12 MB portrait that was feasible but not ideal.

This doesn't have Ben Shahn's artistry and the lead room is on the wrong side. Plus the image has more damage at right than at left. There isn't much room to crop in any direction and the crop I want to make out of sheer laziness would get rid of that bright vertical band. But then we would lose precious lead room and this photograph doesn't have enough of it anyway.

It only takes a moment to perform a crop, but the decisions that go into it mean a great deal. Historic material sometimes carries unavoidable flaws. Alison was willing to work with this anyway; Countess Markiewicz was the first woman in Europe to hold a cabinet minister position in a national government; she was Minister of Labor for Ireland from 1919-1922 (Alison might want to throttle me for not spelling that L-a-b-o-u-r but I'm an ignorant Yank).

She's doing most of the work herself. I offered to crop. There's no really good choice here and a lot of bad ones.

I wanted to crop away at left to balance the lead room but couldn't really go very far: her long skirt, the andirons, her fingers, and the books on the mantel all got in the way. So I left a lot of the bright vertical band at far right. It's a correctable problem. Alison will need extra work to fix it--and somehow I suspect I may be helping with those touches. But the result will be worth it.

For comparison here's a "lazy editor's crop" alternate. Less work to restore but it would never be as good.

So cheers to Alison! (And now, since I'm such an evil wiki witch, she'll just have to do this restoration so you can all read the followup. Excuse me while I drop another newt into the cauldron...)

Copyright limbo

2010-02-05T19:43:00.000-08:00

Today's example of Sigmund Freud isn't so much an example as a contemplation of copyright law. The portrait was taken in 1914, which makes it public domain under United States law. If the portrait had been taken ten years later then Wikimedia Commons wouldn't be able to host it. Although all of Freud's writings entered public domain in the European Union at the start of 2010, United States law doesn't recognize all of his work as public domain. It's a really weird quirk that causes big obstacles.

Copyright terms in the European Union run for the life of the author plus seventy years. Freud died in 1939. So on January 1 2010 all of his copyrights expired in Austria and the United Kingdom. Most countries reflect that lapse in their own law by a provision known as the rule of the shorter term, but the United States doesn't follow the rule of the shorter term.

So I could republish all of Freud's writings from Vienna where they're public domain, but not from the US where they remain under copyright. The Wikimedia Foundation servers are located in the United States. So because the servers fall under United States law, Austrian Wikimedians can't bring Freud's later writings onto the German language Wikisource (which hosts free licensed texts).

United States copyright law is complex, but one rule that holds true nearly all of the time in the States is that material which was published before 1923 is public domain (no matter when the author died). So this 1914 portrait is fine to reuse but a 1924 portrait wouldn't necessarily be free. That copyright gap is going to widen: Anne Frank's autobiography will lapse into public domain in The Netherlands in 2015 but will remain under copyright in the United States. Broadly speaking, United States copyrights are in a holding pattern until 2020. So 1923 works won't lapse public domain for at least another ten years.

For the last thirty years U.S. copyright law has been getting a series of extensions. Those extensions have something to do with lobbying by the Walt Disney Corporation: Mickey Mouse was created in 1925. It doesn't actually protect the value of Mickey Mouse to keep Sigmund Freud under copyright in the United States, though. The United States could adopt the rule of the shorter term without harming Disney's profits.

Other culturally valuable material is getting tied up because of this legal hitch: William Butler Yeats's works entered public domain in his native country of Ireland this year. I can republish this 1911 portrait of him freely, but I can't republish his late poems.

And one thing worth wondering is this: with a normal copyrighted work it could be possible to contact the copyright owner and request a release under copyleft license. How does one seek access to material that remains under a copyright which no one seems to own?

More than one way to skin a cat

2010-02-03T13:49:00.000-08:00

Last month I posted about restoration work on a portrait of opera singer Jenny Lind. Today I came across another graphic artist's approach to restoring her portrait. It's an interesting peek at how different two approaches can be. Robert C. McLaughlin's restoration is a composite of two different photographs including the one from the Library of Congress that I've been using.

He specifically avoids using the healing brush and works toward a much more artistic effect. The result looks more like a pencil sketch than a photograph.

The fast food hamburger of digital editing

2010-01-31T23:14:00.000-08:00

Today I'm going to do the unspeakable: praise the Photoshop auto levels function. This is something like a fine chef admitting to liking fast food. Auto levels is notorious as the one click tool that gives mediocre results. It's reliable, quick, easy, and not very good for you. Sort of like a hamburger. And yes, I have a use for it. Not the usual use, though.

NativeForeigner is working on another Crimean War lithograph. He asked me to check his progress, which was very good except for something he hadn't noticed yet. A problem loomed in the upper left corner.

Don't see it yet? Neither did Native Foreigner. And neither would I unless I had worked with this type of image before. Now let's peek at what the auto levels function reveals: big stains!

A lot of nineteenth century lithographs have subtle problems with vertical banding that show up in the sky. I suspect this is because they had been rolled for storage at some point. This type of problem tends to seem very faint until the editor adjusts the histogram. And then there's just no ignoring it.

If I suspect an image might have this type of problem I preview it in auto levels to gauge the severity.

And to all the other digital image editors who are gasping at this statement, please remember:

That's preview.

pre-view

p-r-e-v-i-e-w.

Please don't pounce on me like you're all chefs who've just caught a fellow chef munching French fries.

Using the History option we step back from that auto levels version. It's too early to really change the histogram permanently. That preview gave a very clear look at the side, shape, and darkness of the staining problem. Now we're going to resolve the problem. There are different ways to do this. A lot of people like the Dodge and Burn tools. I prefer masks and brightness adjustments.

The way I solve this is to draw a free form selection and add feathering. Feathering a selection blends its borders. This edit used 10 pixels of feathering and increased the brightness by 2 in the selected area. If you like to be cautious you can do these edits within new layers. Then redraw another selection, make a very slight adjustment, and keep going until the stains vanish.

Here's the effect of the first edit in another levels preview. Once NativeForeigner understood this he did the remaining work himself.

Library of Congress starts open source initiative

2010-01-30T10:14:00.000-08:00

Considering how much of the historic media at Wikipedia comes from the Library of Congress, it may surprise you that the Wikimedia Foundation has no formal partnership with the Library of Congress.

Yesterday Slashdot picked up on a Library of Congress initiative to do more of their work with open source software. It would make a lot of sense if the Library of Congress interfaced directly with the world's most successful open source nonprofit: Wikimedia. The Library of Congress has been absolutely wonderful about making its material available to the public at high resolution. Today's post expresses appreciation for that openness in the hope that this valuable synergy will be appreciated and built upon.

Eight of the images that ran on the main page of the English Wikipedia this month came from the Library of Congress collection. The image above is a landscape of Havana, Cuba painted in 1639 by Johannes Vingboons, which I restored. Wikipedia's main page received 4.0 million page views while it ran at the Picture of the Day feature on January 1, 2010. The image itself received 11,900 direct page views that day and a total of 13,382 direct views this month. All of the page view statistics for Wikipedia's main page are confirmed here.

The Picture of the Day for January 6, 2010 was a seventeenth century chiaroscuro woodcut by Bartolommeo Coriolano, who was knighted by Pope Urban VII for his artistic work in engravings and woodcuts. I did the restoration. Wikipedia's main page received 5.1 million page views on January 6 and the image itself received 7937 direct page views this month.

This 1868 portrait of an Argentine gaucho was Picture of the Day on January 8. Wikipedia's main page received 5.2 million page views that day and the image received 16,427 direct views this month. This was another of my restorations. We'll be seeing other volunteers' work too.

This 1890s photochrom print of the quays at Waterford, Ireland was restored by Jake Wartenberg. It was Picture of the Day on January 14 when Wikipedia's main page received 5.3 million page views. Jake's restoration has received 33,629 direct page views so far this month. Good work, Jake!

The artist for this woodblock print of a women's bathhouse was Torii Kiyonaga, 1752-1815. Wikipedian editor Torsodog performed the restoration. It was Picture of the Day on January 18 when Wikipedia's main page received 5.6 million page views. The image has received 39,158 direct page views this month.

On January 20 Wikipedia's main page ran this albumen print of Moroccan snake charmers, which was created during the latter half of the nineteenth century. This was another of my restorations. Wikipedia's main page received 5.3 million page views that day. The image file has received 17,884 direct views this month.

On January 21 another one of Jake Wartenberg's restorations ran on Wikipedia's main page. This is an 1856 lithograph of the hospital at Selimiye Barracks where Florence Nightingale worked during the Crimean War. The main page had 5.3 million page views on January 21 and the image file received 21,840 direct views this month.

This theatrical advertisement from 1900 was restored by Adam Cuerden. While it was Wikipedia's Picture of the Day for January 22, the site's main page received 5.3 million page views. The image file hosting page has received 11,211 direct views this month.

Altogether that totals 41.1 million page views this month for Wikipedia's main page while media from the Library of Congress was running on it, and 161,468 direct page views for the eight Library of Congress images that were highlighted as Picture of the Day. These numbers are typical for the attention the Library of Congress collection is receiving through Wikimedian volunteer efforts.

I'd like to coordinate directly with the Library of Congress management to utilize this synergy better. And if the Library of Congress isn't interested I'll be happy to work more closely with institutions such as the Tropenmuseum that see the potential.

SMS Moltke

2010-01-29T22:23:00.000-08:00

Staxringold asked me to collaborate with him on this restoration of the SMS Moltke at Hampton Roads, Virginia in 1912. There was a difficult repair in the foreground on this image, which was 147 MB at full resolution in uncompressed TIFF format. At preview size the area looks like two white marks in the waves about two-thirds of the way to the left.

This image was digitized from a glass plate negative. Glass plate photography was developed in the mid-nineteenth century and was widely used until the early decades of the twentieth century when photographic film was introduced.

One of the problems with this format, though, is that the photographic emulsion is prone to damage. Once damage occurs the emulsion can peel away from the glass. That's starting to happen in this section. The challenge I faced was to reconstruct the appearance of choppy water. The sequence you'll see below was the progressive work on this area that I showed to Staxringold so he can do this type of area himself next time. The following sections are screen shots at 200% resolution.

If it seems a little nutty to work at 200% resolution on a 147 MB digitization of a negative that was no larger than 5" x 7", maybe it is. But glass plates are a high resolution format. Film gained dominance in the consumer market because it was less fragile and easier to work with. Glass plates remained in use for technical purposes such as astronomy and medical imaging until digital technologies took over at the very end of the twentieth century.

All in all, that makes a pretty good useful lifespan for a technology that came into wide use during the American Civil War.

D-Day

2010-01-28T22:00:00.000-08:00

Occasionally digital restoration raises new questions. This photograph depicts a synagogue in New York City on D-Day, 1944. The Library of Congress didn't identify which congregation. A Wikipedian who edits under the username Pharos has identified it as Congregation Emunath Israel on West Twenty-third Street.

I've requested this to be Wikipedia's picture of the day for June 6, 2010.

It would be really wonderful to make contact with this congregation before that day so they know about it. A few of their older members might even be able to name who these women are. Unfortunately they don't have a website or publish an email address. Their telephone goes directly into a voice mail system.

So if you live in New York City or know a friend who does, would you be willing to help contact this congregation please? I'd really appreciate the help.

This stinks in Cologne

2010-01-27T16:41:00.000-08:00

The fourth largest city in Germany actually is quite beautiful. The photograph here was taken by a German Wikimedian named Thomas Wolf and it is featured in three languages. The thing that stinks is that the city's archive building collapsed last year; Cologne's archive had been one of the very few that had survived World War II completely intact. Last March the city was constructing a subway on the same street. Then suddenly the archive tumbled.

Two people died. Thousands of documents going back nearly 1100 years were in rubble. Half a million photographs were housed there. Disasters such as the Cologne archive collapse demonstrate why it is important to digitize cultural records. Digital versions help protect and preserve a heritage.

So with thoughts of Cologne, here's a photochrom print circa 1910 of the Eisenbahn Bridge at Cologne from the Library of Congress collection. I haven't restored it yet; maybe someday.

Looking ahead

2010-01-26T07:21:00.000-08:00

Congratulations to Coffee on his first digitally restored featured picture: a nineteenth century whiskey advertisement so politically incorrect it's almost quaint. It really is wonderful and rewarding to help new people such as NativeForeigner and Coffee learn digital restoration. Yet tutoring is very time consuming.

So today's post is brief: it's time to scale up this digital restoration work. One way to do that is to collaborate with art schools. What I'd like to do is get this incorporated into the curricula of advanced digital editing classes with the best student work to be selected for exhibition in museums. We have the contacts on the museum side to make that happen.

If you have contacts with an art school, please put them in touch with me: nadezhda(dot)durova(at)gmail(dot)com

Beginner's luck

2010-01-25T16:12:00.000-08:00

Would you believe that this image is not only an editor's first restoration, but his first digital editing project of any sort? NativeForeigner asked me to help choose his first project a few days ago. After discussing several options I pointed him to a series of lithographs about the Crimean War. The Library of Congress owns a wonderful set created by William Simpson in the mid-1850s that's been well preserved. He installed GIMP and Skype, we traded screen shots and shared ideas, and he performed most of the edits himself. Here's a view of the unedited original.

NativeForeigner inspired me to think of this when he showed an interest in Roger Fenton's Crimean War photography. The Simpson lithographs make wonderful beginner projects.

So today we'll share a few highlights from the collaboration. NativeForeigner's first question was how to address the area in this screen shot.

How many of these marks are intentional, he wondered? I sent back a draft edit of a suggested first pass to the section. The key concept here is that it's easier to take away than to readd data. So when something is uncertain, leave it for later and work on the easier material while you grow more familiar with the image and gain an understanding of its context.

NativeForeigner trusted my advice and did the first three passes himself. He only really needed to pause for advice again to address a stain at the lower left corner: a dark spot surrounded by a fainter orange mark. He corrected that problem with clone stamping.

After his third pass I found a damaged portion in the sky at upper right, which he corrected. Then he transferred the file and I gave the image a pass to correct for scratches, creases, and subtle stains. Then a Curves adjustment and a color balance provided the final touches.

We're creating the featured picture nomination together at Wikimedia Commons as I write this post. Once it goes live we'll add the image to articles on Wikipedia. This illustration, which was published in April 1855, shows the conditions that Florence Nightingale confronted when she arrived at Balaklava in late 1854. Her work to improve care for the wounded and sick during that war established her reputation. In many ways that turned nursing into a modern and respected profession.

Headhunters

2010-01-24T10:52:00.000-08:00

"You have to appreciate that the deforestation of Nias started when they were no longer headhunting! So it was functional." Gerard Meijssen often expresses unusual perspectives that make me laugh, but he's seldom wrong. This photograph was taken in Indonesia. Here's Gerard's translation of the Tropenmuseum's caption:

Repronegative. The best known island of the range to the west of Sumatra is probably Nias. It was at any rate the biggest and most populous. In the time of the VOC, the island was best known as an exporter of slaves to Atjeh, Padang and Benkoelen. In this way the gentry of the hierarchical Nias earned the gold needed for dowries and the ritual feasts. Nias was a society of warriors who not only enslaved people, they also went head hunting, for instance for the burial ritual of a nobleman. The colonial government tried to end this (P. Boomgaard, 2001). A group of subdued headhunters, Nias “a group of head hunters came to submit themselves” Nias, Northern Sumatra, ca 1920

Go ahead and double check the museum's official metadata if you can read Dutch. Gerard seems perfectly serious, which is a little hard to stomach if, like me, you envision reruns of Gilligan's Island when the word "headhunters" enters a conversation.

Headhunters really existed. And after too many formative years of squandering brain cells on the best bad sitcom of all time I'm restoring an actual photograph of headhunters. Here's a thumbnail version of the image the Tropenmuseum provided at high resolution.

This was probably photographed in or near the village of Bawemataloeo. If Gerard is right, these people took heads as trophies in battle. An effective way to put fear into one's enemies! Although (per Gilligan's Island) I'm slightly embarrassed to discuss this subject, it is fascinating to see the people who lived that way. And a relief to know it's historic. Then I remember the deforestation and my political correctness meter breaks.

This seems to have been digitized from a print on textured paper with directional lighting. The scratches aren't clear lines: they look like little horizontal nicks stacked over each other. It's quite difficult to work on. I've been spending most of my time at 200% and 300% resolution with a healing brush selection three to six pixels wide. The hardest part of it is repairing this man's face.

When an image has problems that are solvable yet difficult, the mellow approach is best: address the parts that are easy to fix and go work on a different area when it starts to seem frustrating. Come back after the subconscious has pondered the situation, work on the tough area again, and slowly turn big problems into little problems.

One of the things I do is keep an unaltered version of the image handy and pause occasionally to view sections side by side. That helps to distinguish photographic damage from real data. It also sustains my morale through a hard restoration. Here's how that section looks so far.

Delurking in the shadows

2010-01-23T12:57:00.000-08:00

Everyone likes the person who can make a portrait look better. A tool in Photoshop called Shadow/Highlight that is equally useful whether the portrait is a nineteenth century engraving or a twenty-first century photograph. Today we'll put that tool to use in both situations.

Our first example depicts Senator Henry Clay of Kentucky who was one of the influential legislators of the nineteenth century. The edit at the top of this post was done by a classical guitarist who volunteers for Wikipedia under the username Jujutacular. Jujutacular has also uploaded several musical performances under free license and fellow editors have selected his work as featured sounds. An example of Jujutacular at guitar is here; it plays best on the Firefox browser. Jujutacular also restores historic images and yesterday I helped him with the image edit, which he had done from this Library of Congress original.

Jujutacular's dirt and scratch removal was very careful, yet the image remains badly faded. Here's what that fade looks like on a histogram. A histogram assigns all the pixels on an image a brightness rating from 0 (for black) to 255 (pure white). Faded images lose data at the left and right ends of the graph.

Image editing programs offer several ways to correct for fade. The Shadow/Highlight tool in Photoshop is a powerful way to coax detail out of the shadows.

This tool has three sets of sliders. The top section controls shadows and the middle controls highlights. Those are the only two sections we'll need for today. The top bar in each section, "Amount" controls the amount of enhancement. The "Tonal Width" and "Radius" bars control the data selected for enhancement.

The shadow/highlight tool shows previews in real time. So the best way to get familiar with it is to open it up and play around with a few images. A good fade correction brings out the information you want to see without overpowering changes.

The final version of the Henry Clay engraving had minimal editing after the Shadow/Highlight adjustment. Notice how much clearer it becomes.

Although I admire the detail of this engraving, overall it's a period piece not quite to my taste. The postures are stiff and theatrical. He looks more like a third rate actor playing a senator than an actual senator. Yet it's a good example of an artistic style that was popular during his career. Have a look and we'll move on to a twenty-first century portrait.

One type of problem that's probably happened to everybody is the indoor portrait taken next to a bright window. The human eye adjusts to brightness differences better than any camera. So this type of pose usually seems like it ought to look fine but gives disappointing results. Here's an example with Philippe Beaudette who works for the Wikimedia Foundation.

Poor Phillippe looks like he hasn't been outdoors in a month. He doesn't really live in a cave; it's just the lighting. The first edit was a mild perspective crop to take out distracting architectural elements at left and right.

Then I wanted to put texture and color back into his face, but not too much. The aim is to give his forehead definition without exaggerating contours so much that the skin looks blemished. On the highlight slider an unusually large radius setting (92 pixels) produced a pleasing effect. The edits to the shadow mostly brought out details in his sweater and hair. I don't want to bring out very much of the shadows, though, because too much could produce the dreaded five o'clock shadow.

Unfortunately the current version of GIMP doesn't have an equivalent editing tool to Photoshop's Shadow/Highlight feature. GIMP users can coax somewhat similar effects from the Curves tool.

Afterward I increased yellow by a small amount in the midtones and highlights, then boosted the saturation slightly. Very small increases to saturation often make a subject look healthier. The edits almost gave him a tan.

Shades of gray

2010-01-22T11:00:00.000-08:00

How do you color balance a black and white image? Today's restoration is an Egyptian camel transport photographed at Mount Olivet near Jerusalem in 1918. It was scanned from a print in an old album. Paper tends to get deficient in blue as it ages. Old photographic prints turn color in similar ways to old newspapers. So this photograph appears to be shades of brown; what do you do about that?

Some people think that black and white photography ought to look black and white. Others prefer sepia tones. We probably can't resolve differences of taste, but it's worth a look at what happens with both approaches. They each have their shortcomings. You'll also see the solution that I think works best.

Desaturation is a simple edit. Here's a screen shot of how to perform it in Photoshop.

If you're editing in GIMP the Desaturate option is on the Colors menu. Here's the result you'll get in either program.

This may be totally justifiable on technical grounds, but desaturation tends to convey an emotional coldness. The original is a gorgeous panorama. This doesn't have the same appeal; I don't feel the urge to lose myself in it the way I did with the original.

So why not keep those warm sepias just as they are? If you're really restoring the image that won't be feasible. The trouble develops when you work with the histogram. Without getting too technical about it, old images usually fade. And when you correct for that fade (in terms of brightness), any discoloration will seem to be intensified. As a demonstration we'll do a simple levels adjustment on a partially restored version of this image.

Notice how artificial this looks. This photograph was well preserved and had a good histogram for something ninety years old. In an average print of similar age those browns could turn bright orange.

One possible solution is to leave both the histogram and the colors alone, and to only perform dirt and scratch removal. That would keep the sepias untouched but lose whatever detail might be recovered from the shadows. My solution is a compromise: I worked on the histogram with the Photoshop Curves tool (which is a more sophisicated cousin to Levels), then I adjusted the color balance by hand and did a partial desaturation.

Since old paper is deficient in blue, a color balance adds blue. This means moving the bottom slider toward blue and the top slider toward cyan. Photoshop allows the user to adjust highlights, midtones, and shadows differently. I like to start with highlights because usually blue is more deficient in highlights. That makes sense if you think of paper prints turning color like old newspapers: the dark areas have more ink, which conceals some of the change that occurs in the paper.

For this adjustment I moved the sliders four units toward cyan and blue in the highlights, then three units toward cyan and two toward blue in midtones, and one unit each in shadows. Don't worry too much about remembering those exact numbers. When you're editing just use the preview function and adjust the settings until they feel right.

Next I reduced saturation in the Hue/Saturation dialog. This lets an editor take out some of the color while still leaving enough to to keep that emotional allure. The sepias won't be overwhelming anymore, but they're still there.

The equivalent editing tools can be accessed in GIMP by choosing Layers > Colors > Color Balance for color edits, and Layers > Colors > Hue-Saturation.

Here's the end result of my edits. You might prefer a slightly different final version. Photography is an art; so is digital image restoration. The unedited Library of Congress file is available here. Soon I'll be uploading my work on the image to Wikimedia Commons including an interim edit that doesn't include my final changes to the histogram, color balance, or saturation. Feel free to work with that if you'd prefer something a little different.

Talk like a Pittsburgh Pirate

2010-01-21T12:03:00.000-08:00

There's a wonderful Wikipedia editor named James Ringold who puts his volunteer time into baseball articles. He edits as Staxringold and recently nominated a featured topic about Major League Baseball awards where seventeen of the eighteen articles are featured articles. It really shows how much the baseball editors put into their work. When James restores images he's equally diligent; he recently completed work on a poster about the Pittsburgh Pirates' 1909 National League championship. That's Honus Wagner at the upper right corner next to the pennant. The full resolution restoration is available here; it makes an interesting comparison against the original.

Each day Wikipedia's main page runs a section called Picture of the Day. Scroll about halfway down the page to see it. You can see from next month's upcoming schedule that Babe Ruth will be on the main page on his birthday, February 6. Sometimes special occasions get scheduled per request.

James and I both think the Pittsburgh Pirates would be perfect for International Talk Like a Pirate Day. If you'd like a peek under the hood at how Wikipedia operates, here's our conversation with the Picture of the Day director:

POTD request for September 19

Arrrrr!

Seems perfect for Talk Like a Pirate Day. ;) Durova⁴⁰¹ 22:58, 16 January 2010 (UTC)

"Talk like a Pittsburgh Pirate"? Heh. I think there's another pirate-themed FP lying about, however. howcheng {chat} 23:55, 18 January 2010 (UTC)

NOOOOOOOOOOOOOOOOOOOOOOOOO! :) Staxringold _talk^contribs 01:46, 21 January 2010 (UTC)

----
Many thanks to the Library of Congress for making these fine images available to the public at high resolution.

Camera tilt

2010-01-20T11:05:00.000-08:00

Today's edit techniques are equally useful whether the photograph depicts a World War I battleship or yesterday's birthday party. You're going to like this if you're new to digital editing and want good skills that are quick, easy, and make your friends think you're an expert.

This ship is the USS New Jersey (BB-16), painted with camouflage in 1918. It's one of the subjects of a Wikipedia content drive called Operation Majestic Titan, which is working to provide good articles about every battleship ever built. An administrator who works with that project, The ed17, thought this photograph might have the quality to become a featured picture.

Ed asked me for help after the first reviewer noticed how tilted the horizon is. He didn't think the defenseimagery.mil image was high enough resolution to merit a full restoration--he just couldn't figure a good way to correct this tilt. Ed's dilemma was that he had no leeway. The photograph already cuts off part of the ship's superstructure. He didn't want to lose any more information.

If the solution looks confusing at first, smile and think how much it will impress your friends.

The first thing to do is straighten that horizon: this image needed two degrees counterclockwise rotation. That's why the framing looks tilted within a black box. The next step is the crop. The dotted lines on this screen shot show where the crop was about to happen.

Ed understood the basic mechanics of editing but he wasn't sure how to apply them to this situation. This image had enough room to crop left and right and a little too much foreground, but that superstructure left no good choice at the top of the frame.

Obviously we aren't going to propose a featured picture candidate with black triangles in the sky, but that's just sky. We can patch empty sky.

These edits used the clone stamp and then the healing brush to blend a natural effect. Altogether it took less than five minutes to perform the rotation, crop, and the repair on the sky.

If you don't already own software that has these features, an open source program called GIMP can do it. The Wikipedia article about GIMP includes external links for download and tutorial.

And if you're more likely to edit birthday party photos than World War I history, a good habit to get into is to zoom out a little while shooting. The editing is much easier when all four sides of an image have enough room to rotate and crop.

Pattern recognition

2010-01-19T08:57:00.000-08:00

The human nervous system is wonderful at pattern recognition. That's bad news when photographic damage forms a distinctive pattern because even when restoration removes nearly all of the damage our brilliant brains still see a pattern. There's a trick to overcome that effect.

There are always more good images to restore than time to restore them. A friend is collaborating on the Haitian restoration and the Indonesian restoration is nearly finished. Today's post looks at a different part of the Indonesian portrait where the challenge looks difficult and is actually even harder than it seems. Here's what we're dealing with:

Vertical streaks at close parallel are a tough kind of damage. We want to erase those streaks without losing the folds of the garment or the shape of the earring. I worked on this with a healing brush at six pixels diameter, which was just the right width to erase those lines.

Healing brushes work by sampling one area of an image and performing a complex mathematical operation to blend data from that region with a different area. If you think of this intuitively, the way to heal a scratch that goes from a light section of fabric to a dark section of fabric is to choose a source region that has similar light and dark areas. One by one those scratches start to disappear.

The progress feels rewarding for a while until you reach a point where the scratches are nearly gone. But darnit, you still see 'em.

This is one situation where reviewing an area at 200% resolution is not sufficient: subtle patterns emerge when you pull back to 50% resolution. Our brains recognize lines--especially lots of parallel lines. We distinguish lines even when they're faint and discontinuous. This gets very frustrating when you're trying to erase them.

Fortunately you don't have to sit at the computer for hours until you start climbing the wall.

The secret to finishing work on this area is to recognize that your brain is just reacting to a pattern. All you really need to do at this point is break up that pattern and randomize it. So instead of editing in an endless series of vertical strokes, increase the pixel selection and follow the curves of the garment. Clone stamp the delicate areas such as the necklace.

A neurological phenomenon called lateral inhibition might have something to do with why we see this. If the phenomenon interests you, the Mach bands illusion and the Cornsweet illusion make good reading.

The cabinet secretary's chin

2010-01-18T09:40:00.000-08:00

A Wikipedia administrator who goes by the username Wehwalt has contributed a lot of featured articles about historic politics. He's one of the reasons Wikipedia's biographies of Neville Chamberlain and Nikita Krushchev are as good as they are. The portrait at right is lead illustration at another of Wehwalt's featured articles. Franklin Knight Lane was United States Secretary of the Interior during Woodrow Wilson's presidency.

Wehwalt and I were discussing a possible restoration on an image from one of his articles. Something unexpected turned up last night when I saw the full resolution version of Lane's portrait. This doesn't happen every day: it looks like a crude job of retouching tried to hide a double chin.

Notice how the dark marks begin just below his actual chin line although it's obvious that the actual skin in that area and below the side of his jaw is not in shadow.

So here's an invitation: you help me decide. Should I restore Mr. Secretary Lane's original double chin or should I complete the retouching properly? Normally I don't try to make encyclopedic subjects more handsome than they actually are, but this alteration looks historic.

My inclination is to reconstruct the original chin (both of them). Yet I'm really curious what you think.

Either way, here's hoping you chuckle as much as my friends did. There's nothing like a little bit of vanity even if it's 97 years old.

Preserving culture

2010-01-17T16:05:00.000-08:00

The other day Jose Antonio Vargas wrote a piece for the Huffington Post about online coverage of the Haitian earthquake. He had very kind things to say about Wikipedia's article on the earthquake--hundreds of people from around the world were prompt about compiling relevant information from newspapers and other sources.

Internet communication brings the world a whole lot closer. It can bring out the best in us, yet the comfort of distance also produces a few howlers. When Jose asked for research assistance with that story I stumbled across these highlights from a two hour sequence in the article edit history. These edit summaries all seem well intended even if not all of them are equally clueful:

18:40, 13 January 2010 Scribeofargos (talk | contribs) (41,751 bytes) (→Casualties: UN has confirmed Annabi is missing and updated the latest confirmed casualty numbers released from UN) (undo)
19:07, 13 January 2010 Schuster gph (talk | contribs) m (47,425 bytes) (Put Costa Rica on the list of solidarity. Waiting for reference as it is happening now.) (undo)
19:24, 13 January 2010 Krenakarore (talk | contribs) m (47,896 bytes) (International aid - Brazil) (undo)
19:39, 13 January 2010 Br10ta10 (talk | contribs) m (48,908 bytes) (Added the fact that the main Port-au-Prince jail collapsed (and prisoners escaped) ~~~~) (undo)
19:51, 13 January 2010 RobNS (talk | contribs) (49,108 bytes) (Added image of National Palace before it was destoyed) (undo)
20:47, 13 January 2010 Moncrief (talk | contribs) (52,306 bytes) (→International aid: why oh why would someone do this. This is an encyclopedia. SPELL OUT "AND") (undo)

So, during this time while fresh water and mass burial are urgent concerns is it inappropriate to discuss photography? Here's hoping, with respect, that it's sufficiently clueful. One of the things that gets lost in a disaster is culture. What happens to the museums and libraries? The photograph above is a Port-au-Prince street scene from 1901. Since I can't get to Haiti to clear rubble from the streets I'll repair a piece of Haitian heritage.

Here's a section near the lower left corner. Someone wrote a description on the image. It also has a large fingerprint and several small dirt specks.

A good habit to develop when doing restoration work is to repair small damage first. The edits so far were all done with the Photoshop healing brush at default settings and 4 pixels diameter. Now that this has a good region at right to sample from, I increased the healing brush to 21 pixels to address the thumbprint.

With the preparation work done properly, the fingerprint disappeared in under a minute. The trap to avoid is the urge to rush things. Nothing looks more artificial than repetitive fiber marks from faulty cloning. The practice I follow to review a section at 200% resolution. If the result looks natural enough that I can't see where the edits were made, then it's ready.

This image is not going to be finished in an hour, though. Similar to the Guadalcanal photograph, it has scratches through important sections. Whenever a scratch crosses a boundary that boundary area needs special attention to repair. So a large scratch across a busy street scene requires much more effort to fix than a scratch across a clear sky. Here's a view of the work on another part of the photograph. Not exactly before and after, but before and interim. The portion at right shows the work in progress. Click in closer to see the details.

Selecting a project

2010-01-16T14:12:00.000-08:00

This week a Marine who's stationed in Afghanistan posted to my Wikipedia user page to ask for assistance learning digital image restoration. He said he found me through the site's Military History Project, which is a place where editors organize work on the subject, and asked me to select his first project.

It's important to work on a subject one cares about so I headed to the Library of Congress website and ran a search for the keyword "Guadalcanal". It was the first major Allied offensive operation in the Pacific during World War II and the Marines saw heavy action there. Found a number of high quality images, not all of which are in public domain. After researching the rights status the best I could find was this image of a team of people aiding a man on a stretcher. It's an excellent photograph in terms of composition and expression. But it wouldn't be a good beginning restoration project. A view at full resolution reveals a large number of surface scratches at important areas. The injured man's face at right is a sample. The problem gets worse at lower left where they're dressing his wound.

So I ran a second search for Iwo Jima and located the Second Battalion, Twenty-Seventh Marines landing at the beach on February 19, 1945.

The Iwo Jima photograph also has scratches and even fingerprints, but none of the most serious damage occurs in a crucial area. At right is a crop of the fingerprints at full resolution. It's fairly easy to repair a problem of that size when it occurs on a sandy beach or in a sky: the historic value of this image is not tied to the placement of a few grains of sand.

The important information is mostly in the human figures. As you'll see below, the work to be done there is at a level a beginner could handle. The ideal thing to go for with a starter project is a project whose damage in critical areas is limited to small dirt and fiber issues.

Now I'll be getting to work on a new selection for my own projects: a photograph of Port-au-Prince from 1901. More about that tomorrow.

The eyes have it

2010-01-15T11:36:00.000-08:00

Received a beautiful comment from an Indonesian blogger about last month's Balinese dancers. So here's another item from the Tropenmuseum's Indonesian collection. The portrait at right was taken in 1918. Curators have selected it for restoration as having special significance and sent over a 25 MB file; I'm not certain exactly why they chose it. The metalwork on this woman's headgear is fascinating.

Today's post is about the importance of eyes. Other aspects of an image offer more flexibility in restoration: although it wouldn't matter much if a few pixels altered a crease on her garment, human facial expressions convey a great deal of information. The original image looks pretty good at thumbnail, but this is her right eye, unedited, at 200% resolution.

That's a vertical scratch from her eyebrow which passes down the left edge of her eye. Then a second scratch curves with the lower edge of her eyelid and ends in a patch of corrosion, coupled with a second corrosion spot near the orbital bone at lower right. The image also has several dust specks that are most noticeable as light marks on her eyebrow. And if you look very carefully there's a series of faint horizontal scratches at the right edge of her eye. Those horizontal scratches continue farther right in the image.

The goal is to correct all of this damage without altering her expression. There's a hint of challenge to her face--not really hostile, but confident. She looks like she's studying the photographer as intently as the camera studies her.

Beginners usually go for the largest areas of damage immediately. It works better to begin with the small issues instead.

The smallest spots and scratches are out of the way now. One needs to be careful with this and not go so far that the work damages skin tone. Also, the shape of those knitted eyebrows forms part of her expression. Our aim is realism rather than airbrushed perfection. The next edit will work on that scratch at the lower eyelid.

This was the hardest part of the edit sequence because the scratch ran along the border between light and shadow. A slight tension in the orbital muscle that surrounds the eye gives the expression its sense of challenge. if the bright part of the lower eyelid comes out too wide, then that orbital muscle looks relaxed and the expression turns bland. If the dark part of the lower eyelid gets too wide then the expression hardens into a scowl. Working with the healing tool at a four pixel diameter setting managed a good balance between those extremes.

Here's the same closeup after the rest of the dirt and scratch removal is finished. Haven't changed the histogram here; that will come later. But the signs of photographic age don't distract from the look in her eye anymore. Pulling back a bit, and viewing in the context of other unrestored areas of her face, have another look at the work in progress.

She would be about 110 years old if she were still alive today. Perhaps the museum staff knows her name.

High tech for 1850

2010-01-14T20:22:00.000-08:00

It's been quite a week. Welcome to the new followers! Received a sudden influx of comments. A friend figured out that this got selected for Blogger's "Blogs of note". As luck would have it, the selection happened during a break (the Blogger staff don't notify in advance). So thank you very much for the attention and kind comments. I'll try to keep supplying material.

The latest project is a portrait of a Swedish opera singer named Jenny Lind. She's such an important name in Sweden that they put her portrait onto their money; she's on the front of the 50 kronor bill. In 1850 she toured the United States and sat for a photograph. Unfortunately for my poor self, high tech for 1850 was a process called a daguerrotype. I've never seen a daguerrotype that's been preserved in good condition. As you can see from the detail above, hers is no exception. I've already spent more than twenty hours on it and expect to spend twenty more. So here's a look at the work in progress.

First I worked on the background and the really obvious scratches. As of this screen shot I was starting to get to the scratch that crosses the bridge of her nose.

One question I asked along the way was whether those spots on her face might be freckles. A close examination at 300% resolution indicated that the spot pattern on her face is continued through the background. She was thirty years old when this was taken, which is past the age when freckles are most noticeable. So I worked from the assumption that those spots were degradation. Here's how far this segment has progressed.

The collar and shoulder are looking better. Still need to do detail work at her eyes, eyebrows, hair, and nostril. Large numbers of parallel scratches are difficult to repair. In this instance that repair has additional complications: the version of this image that has been digitized isn't the original daguerrotype. Sometime during the twentieth century the daguerrotype was rephotographed with a film camera for archival purposes. That second camera was at soft focus. So unlike direct daguerrotype scans, which have good resolution, this version gets muddy at close view.

Still, she's one of the most famous women in Swedish history. So here's a look at the whole thing, reduced in size from the version I'm actually working on.

Before:

Current status:

Holiday spirit for the Scrooge in us all

2009-12-22T15:46:00.000-08:00

The holiday season is upon us so this post combines image editing comments with links to my favorite Christmas music. Overall I've got a low tolerance for canned Christmas cheer so today's selections skew toward the offbeat and cynical.

First musical selection: "The Twelve Days After Christmas"

The image at top is a 1904 drawing in pastel and gouache by Everett Shinn, called "Fourteenth Street at Christmas Time". This is what we'll be discussing today. Usually the Library of Congress hosts unedited files of its scans. I can tell someone has fiddled with this, though. the histogram gives it away.

See those vertical lines across the graph? That's a dead giveaway the image has been digitally edited. The lines are evenly spaced, too, so I know what that person did was a straightforward levels adjustment. This is now it appeared when I downloaded the image and cropped out the border. It's a shame, really. Because the histogram wasn't asjusted correctly. See the gap at the right end of the graph where there's no data at all? The white point is too far to the right, which gives the image a muddied appearance.

Normally the histogram would be the final edit anybody makes to an image. It's a point of no return; subsequent editing can't recover the data that gets changed this way. I played around with this one though, and didn't do anything else to the image. You'll see how this brings a little more life into it.

And to go with that, here's "Mrs. Claus Wants Some Lovin'"

The difference between this and the earlier version is most noticeable at the shop window and in the snow at the lower right corner. The downside is that we've lost more data in the meantime. Here's the new histogram; each of the vertical bars represents a brightness level between 0 and 255 where the image contains no data.

It's sad to encounter a good image that has already been edited in ways that make it impossible to get optimal results from later edits. This is a good demonstration of why it's best practice to maintain archival copies of digital images without any histogram alteration. Although I'm not against histogram adjustments, I do upload pre-histogram versions of edited files for recordkeeping purposes whenever possible. That way, if anyone wants to do additional work they don't have to choose between quality loss or completely starting over.

To go with this we have the most somber of today's selections: Simon and Garfunkel's "The Seven O' Clock News/Silent Night"

Just to see what else could happen, I also worked on the histogram with a couple of other tools. Curves and shadow/highlight allow for more sophisticated alteration than the levels tool. There are many possible ways to apply them, and I actually don't know what the artist's original looked like or what his intention was. So these are speculative changes. Yet it can be a useful exercise to work with different tools and see what one can do with them. First, the curves tool.

In spite of the somber tone of this scene, I do like the difference. So here's Shirley Temple singing "I Want A Hippopotamus for Christmas" while we look over the histogram for this edit.

The jagged appearance is a giveaway that this has been edited after a histogram change. I enhanced contrast toward the darker part of this image, which accounts for the change in its overall curvature.

One other thing we could try is the shadow/highlight tool. It performs sophisticated enhancements on the brightest and darkest ends of the histogram.

If you're interested in studying the artist's technique this version might be useful, but it brings out distracting information.

The emotional impact of this image seems to be in the contrast between the bright shop window and the poor peddler. A mass of shoppers cluster near the established stores while he struggles with his small cart of merchandise ankle deep in icy sludge. If the store window isn't bright enough it doesn't set off his misery. More than anything else, that hunch was what prompted me to look at the histogram in the first place. And the straightforward levels adjustment of my first edit seems to confirm that opinion. The curves and shadow/highlight tool edits were attempts to refine the idea. I didn't want either the snow or the cart to look too cheerful, but I wanted to keep that bright window looming above his head. Does it suggest a halo? Maybe he endures this to support a family.

Of course these are speculations that I wouldn't articulate on Wikipedia. Yet they're elements I do consider when working with an image. There's also an argument to be made that fine art shouldn't be restored at all. Yet this image was already edited before it was put online, and if my opinion is worth anything that edit wasn't done very carefully.

So since we're looking at histograms today, here's the histogram for the fourth version. I happen to like The Bobs (they're an underrated a capella group), so linking "Christmas in L.A." along with this.

Overall I prefer the third edit for this one, but am not quite sure which way to go. Seeking opinions and input. Your voice is welcome.

Making light of troubles

2009-12-14T16:06:00.000-08:00

While Garrondo and I were discussing the Photoshop curves tool last night he asked me to blog this restoration as an example. It depicts a pair of Balinese dancers and dates from 1929. The photograph also has a lot of other distracting elements, so let's fast forward to the completed restoration and then concentrate on a few points. Here's wishing the actual editing went so quickly; it actually took a day and a half.

The most difficult part of this restoration was not the obvious marks and streaks. It was the tall flower on the right dancer's headdress. The dust specks weren't so hard to remove; the problem was what to do with all that streaky low contrast bands which intersected it. Compare it to another flower at far left: the unrestored versions appear so different that side by side, they hardly look like they both come from the same photograph.

Regardless of what caused the streaking (I've got several guesses), the first question was whether to correct for it. So work began by getting rid of the spots specks, while leaving the streaks unaltered. Sat back for a view of that version and decided the streaks really needed to go; they served no visible function other than distraction from the primary subjects.

The next phase of editing didn't try to create a totally even background. I didn't want to alter much adjacent to the dancers themselves. One of the few areas where background absolutely had to be patched near a costume element, though, was that darn flower. Used the clone stamp at a low hardness setting to get a natural effect. Afterward several brightness adjustment masks reduced the starkness of the differences between portions of the image. Also enhanced contrast at the rose, which ultimately meant re-cloning the background in that area.

The final result on that flower might be a smidgeon too contrasty. Yet considering what there was to work with, it was a pleasure to see it come this far.

Another aspect of this restoration was really gratifying at the end, although it's subtle enough that it's scarcely the first thing to catch anyone's eye. In the original photograph the face on the dancer at right is slightly overexposed.

To see the difference a curves adjustment makes, here's a view of her as she looked after all the edits except for the curves adjustment had been completed. There's a blandness and flatness about her; it's hard to catch her expression.

Now have a look at how the curves tool brings out the midtones. Her face is still youthful but less like putty; more personality shows. And the intricate contours of her headdress and shoulder band become more noticeable.

Restoration can be tedious work. An image such as this one has so many obvious flaws that one runs the risk, out of exhaustion, of shortchanging the final adjustments. I wish I knew the name of the somber young dancer in this photograph. Those eyes look like she saw a lot of life at an early age.

Showing off one's curves

2009-12-13T16:50:00.000-08:00

December has been a very busy month so far. After publishing an editorial in the Wikipedia Signpost, several people approached me with an interest in undertaking digital restorations. Some of them had never tried it before. So this post is dedicated to one of those adventuresome and talented people. Garrondo expressed an interest in science and in Spanish history. So after his first effort, which went quite well, he asked for advice about choosing a second project. The Jay I. Kislak collection of the Library of Congress rare book division was a good place to go: it includes a 1790 publication of early Spanish colonial archaeology by Antonio León y Gama. The scanned book is 145 pages of text about Aztec science and art with one glorious fold-out illustration, reproduced above. And thanks to excellent digitization practicesthe LoC the illustration is also available in a 37 MB TIFF version.

Garrondo liked the selection--who can blame him? But it's a very ambitious project for a beginner. The scan suffers from bleed-through text and the original has smudges with a distracting crease through the center. It was also made on laid paper, which carries its own distinct pattern that makes restoration bothersome.

He came back today with an impressive first effort.

One has to view these things at full resolution to see how good his work is: Garrondo took care of the dirt and creases and bleed-through text while preserving the laid paper pattern. The result is worthy of featured picture recognition and highly encyclopedic. So bless him--he took my criticisms very well.

It looked like he had adjusted the histogram. He provided a pre-histogram TIFF file for me to play with, which became a demonstration of the difference between Photoshop's auto levels feature and its curves tool.

At left is the auto levels version; at right is a manual curves adjustment. Notice how the paper seems less dirty and the contrast is better on the snake's scales and the sun's face. Both adjustments deal with the image's histogram, which assigns a brightness value of 0 to 255 to every pixel on an image.

Physical documents lose contrast as they age: nearly every newly produced image will have values across the entire 0 to 255 range, but older documents often have no data at the extreme ends of the histogram range. In this instance the extreme points were 7 and 236, with minimal data toward both ends.

Auto levels chops the histogram toward the extreme points of its range--but may lose a little bit on both sides--and does a straightforward average of the remaining data. So I suspected the auto levels had lost information on paper texture. The simplest way to address that would be to redo levels manually, but there was more to be accomplished here. Curves allows a nonlinear change in the histogram. Or to express it without jargon, a curves adjustment lets the editor enhance contrast in the darker end of the range where those scales are and enhance contrast again in the midrange for the sun's face and the figures on the dial while reducing contrast at the bright end to make the paper appear cleaner. For a closer look, here's the curves adjusted version on its own.

The result is brighter, clearer, more vivid. Garrondo can get even better results if he puts more work into that irritating laid paper background to minimize smudge marks. Notice how good the lower right corner looks: a similar quality on the other three corners will improve the final result. This isn't just a matter of background detail. Software isn't intelligent; Photoshop can't tell the difference between a historic archaeological report and a dirty smudge. It's the human editor's task to prepare an image with informed choices so that the data is as good as possible. Takes a lot of patience, but when it's done right the final edit feels like magic. Like a window back in time.

Garrondo agreed to put more work into the smudge removal. It will take a little practice for him to get used to the curves tool, but the result is worth the effort of learning.

Arbitration voting guide

2009-11-30T13:51:00.000-08:00

People have been asking for recommendations on this year's Arbitration Committee candidates. So here's a rundown of principles and useful links.

The Arbitration Committee's core function is to weigh evidence against policies in conduct disputes. The Committee is also a final venue for community sanctions appeals and they do a few other things which privacy issues require keeping offsite, such as ensuring that checkusers and oversighters are trustworthy.

It's equally important to remember what the Arbitration Committee is not: its mandate does not include content disputes. In 2003-2004 ArbCom was created in tandem with the Mediation Committee and the specific mandate of content resolution was designated to MedCom. The Arbitration Committee does not write Wikipedia's policies or guidelines. It is not a governance body.

Arbitrators have to solve Wikipedia's toughest disputes. These include several long running conflicts which involve a dozen or more editors and have lasted for years: ethnic, nationalist, and religious disputes, as well as several others. The level of interaction within those disputes sinks so low that in some of them serious offsite harassment occurs.

A good candidate for arbitrator should come to the position with formal dispute resolution experience, so they're aware firsthand of the problems that happen in dispute resolution. There are six types of formal dispute resolution at Wikipedia:

Half of those processes have designated volunteers:

Other relevant processes include oversight and checkuser:

Script-generated measurements of a candidate's contributions can undervalue the people who are active in these areas. Mediation, clerking, and checkuser all require careful attention so the people who do them spend a lot of time reading and evaluating. This yields a high percentage of Wikipedia namespace edits and, often, a relatively low edit count. So when checking out a candidate, look to see what they actually do.

An ideal arbitrator is someone who puts doing the right thing ahead of politics. Good arbitrators make hard decisions. When mud is being thrown they wash it off, even if they're pressured to apply adhesive. Good arbitrators don't twist logic into knots to avoid offending someone who has political connections. As a result, the best candidates for arbitrator usually have made a few people unhappy.

"Drama" is a word that gets tossed around too carelessly. The important thing is to see whether a candidate who has been in a controversial area has been making it better or making it worse. One useful approach is to survey a candidate's actions during several controversies where most of the circumstances are different but some of the same names appear. If the candidate invokes opposite principles while consistently supporting the same people, that's a big red flag.

Another consideration is content experience. Wikipedia is an encyclopedia and it's good to have arbitrators who understand that firsthand. Yet it's possible to take the content emphasis too far. Featured articles are the palaces of Wikipedia; arbitration occurs in its slums. It certainly is worthwhile for arbitrators to have contributed featured content, but once a person has done it a couple of times, they get the point. A person who has written twenty featured articles but not done much else could be a babe in the woods when they get to arbitration. Or worse, they might try to expand the Committee's mandate into content where it doesn't belong. Balance of power is a good thing. Let's elect a Committee that keeps it that way.

Parsing statistics

2009-11-29T14:09:00.000-08:00

The BBC has published Wikipedia's rebuttal to statistics about a dropoff in editors. If one uses different criteria and defines "editor" at a minimum of 5 edits a month instead of 1 edit, then the departure rate drops from 49,000 to 4,900 and the actual inflow/outflow of editors is at stasis.

Since those numbers differ by an order of magnitude, I'm curious to see a more in-depth approach. What percentage of posts by 1-4 edit editors were reverted as vandalism? And how does that compare against those who edit more frequently?

Possibilities to explore:

Good faith newcomers are getting driven off by aggressive reversion?
Breaching experiments are on the decline?

I have a hunch that reversion is somewhat more vigilant toward editors who haven't registered, or who have redlinked userspaces. Also have a hunch that there is a large but limited pool of people who conduct breaching experiments without actually intending to contribute. In the latter scenario, most of the people whose main intention is to test the edit function have already written "hi there", gotten reverted, and left.

It might be possible to measure those behaviors in relative terms. One approach would involve tracking how many edits the average new account makes before the user talk and userpage get created, and by comparing the reversion rates of edits from accounts that are fully bluelinked, partially bluelinked (talk page but no user page), and fully redlinked. Another approach would be to parse the rate of short-history editors whose posts include the words "hi", "hello", or "test".

Open Progress

2009-11-28T16:28:00.000-08:00

Expressing special thanks today to Gerard Meijssen and the Open Progress Foundation for outstanding work opening global access to digitized material from great cultural institutions in The Netherlands. This week 35,000 images were uploaded to Wikimedia Commons, and due to the cooperative relations he has developed Wikimedian volunteers can obtain high resolution versions of selected highlights upon request, for restoration purposes.

Several months ago I expressed gratitude for these efforts by restoring an 1890s photochrom print of the Amsterdam Centraal railway station. Today it's Wikipedia's picture of the day being highlighted on the site's main page.

Congratulations, Gerard. This is for you. Couldn't have happened to a nicer person.

Giving thanks

2009-11-23T21:03:00.000-08:00

Most of the active Wikimedians are aware that the English language Wikipedia is largest in terms of total articles and the German Wikipedia is second, but it it comes as a surprise that the second largest featured program is not German but Turkish.

The English language Wikipedia currently has 2,111 featured pictures. German has 747 featured pictures. Yet the Turkish Wikipedia has 976 featured pictures. That sets the Turkish language site in second place if one counts by Wikipedia editions. To set this in perspective, the Turkish Wikipedia is nineteenth in overall size with 138,000 articles. The German Wikipedia is on the verge of crossing the one million article milestone.

This means roughly 1% of the articles in the Turkish Wikipedia have featured pictures. One featured picture can illustrate multiple articles. That comes out as a far higher ratio than at any other Wikipedia of substantial size.

What's really interesting is how one medium sized Wikipedia developed a featured picture program that's 25% larger than the German program. A core of perhaps half a dozen Turkish editors have been scouting other language projects' featured picture programs, translating the captions into Turkish, and adding the images to articles on their Wikipedia. Featured picture content at the other projects is different enough from one site to another that the Turkish editors could amass three or four thousand featured pictures if they just keep doing what they're already doing.

Of course I hope they also knock on the doors of museums in their home country. Also very curious about whether this generates synergies with text edits and improvements at their Wikipedia.

Two things happen universally when people from other parts of the world see restored featured pictures about their own culture:

They're delighted.
They want to share information.

Now and then I do restorations about Turkish history and culture, in gratitude and to give their project an extra boost. The most recent of these is the 1917 Ottoman heliograph crew at Huj pictured above. Its featured candidacy is underway at the Turkish Wikipedia. Last week the Turkish Wikipedia passed Wikimedia Commons as the project where I have the second most featured content credits. Darned if I understand the discussion other than the succession of green light icons.

The Turkish editors have created a model that can be emulated. And I'm very interested in trying a pilot project with another Wikipedia to see whether a caption translation and featured picture drive would provide a shot in the arm in terms of editor participation and article growth.

One place where I'm proposing a pilot program is the Irish language Wikipedia. They already have a small featured picture program (two dozen images, a quarter of which I restored) and I know one of their administrators (hi there Alison) and have done a few restorations specific to her country and culture. A few days ago President Kennedy came up in conversation and I asked whether she knew en:wiki has a featured picture of his brother.

There's Robert F. Kennedy at a CORE rally in 1963: the Attorney General of the United States speaking from the steps of the Justice Department in favor of racial equality.

Very cool.

Right now the Irish Wikipedia is the ninety-second largest Wikipedia with 9,274 total articles. If their editor community is willing I'd like to help them emulate the Turkish featured picture program.

Although I can't speak a word of either language, if a picture is worth a thousand words we can hold a conversation.

Letter perfect

2009-11-22T11:57:00.000-08:00

Thumbnail previews and reduced size views don't always reveal how much work a restoration is going to require. This World War I era poster is in very good condition. At the web-optimized reduction for this blog post it hardly seems to need any restoration at all. A closer look at the full size file, though, shows that this won't be a walk in the park.

It's a mostly good image with several tricky problems including a crease that runs vertically through the Statue of Liberty's hand and torch. Today let's look at the ink smears. It isn't unusual to encounter smeared lettering on historic posters. Often the source of the problem is water damage. In this instance it only affects the line that was printed in black ink. Most of the caption is gray and doesn't have this problem, but the entire line of black lettering has ink smudges.

In case you're wondering, this poster translates to say, "Food will win the war - You came here seeking freedom, now you must help to preserve it - Wheat is needed for the allies - waste nothing." It was printed in 1917.

Here's the most heavily damaged word at full resolution. I actually perform the restoration at twice that resolution, but this is enough to convey what the work will be. The basic idea is to trace the outline of each letter and substitute undamaged paper texture in place of the ink smears. Two factors will make the difference between a mediocre repair and a good one:

Paper texture in historic images is not created equal. Slight differences will occur in brightness, color balance, and roughness. So the source area has to be chosen with an eye for those subtle distinctions or else the result will look patchy.
The letters have to look like they exist naturally within the cloned area. This means the cloning has to mimic the aliasing that occurs in undamaged regions.

My first passes worked mostly with large areas and a tool setting of 12 to 15 pixels in diameter. The source area for this cloning comes from an area that looks a bit rougher and more textured than most of the poster. Our goal here is not to create something that's digitally perfect, but that fits seamlessly with the surrounding image. The aim is to mimic good printing for 1917.

The narrowest parts between letters have to be done at a tool diameter of five or six pixels. Unfortunately this is a situation where Photoshop has a big advantage over the current version of GIMP. The Photoshop clone stamp tool has a sliding option that allows the user to select any percentage hardness.

Hardness affects how much a cloned area blends with surrounding data. One hundred percent hardness looks like the cloned area was cut with scissors and pasted in. Zero percent hardness is really soft and smudgy. Smudgy is what we're trying to get away from, so we do want some hardness here. But we don't want the text to seem like a ransom note or an old punk rock poster. So what's needed is something in between. A static setting is going to lose its subtlety as this work progresses from wide spaces to narrow gaps between letters. I do most clone stamping at thirty-five percent hardness. Getting down to half a dozen pixels, though, it helps to be able to drop that to twenty percent.

This is where GIMP users get caught between a rock and a hard place. Hardness in GIMP is a simple on/off toggle. If there's a plugin to make that more nuanced I'd love to know about it because the GIMP editors who work with me have real trouble with this sort of challenge. They can get an acceptable result with the default program if they work hard enough, but it takes them several times as long as it takes me in Photoshop. GIMP is open source, so if you happen to be a motivated programmer who likes to see this work spread free culture you could do something to help solve this problem.

I fixed the smudges on this row of text in about two hours. If discussions with good GIMP editors are accurate, multiply that by a factor of three to five for them to get a result of comparable quality.

The full version of the completed restoration can be viewed here. Below is a glimpse of it.

Warm welcomes

2009-11-21T14:27:00.000-08:00

More good news: another editor has gotten into image restoration and doing really good work. The image above is a historic isothermal chart from 1823. Jujutacular ran it as his first restoration about a week ago. His first effort was impressive and it was a treat to see that he had gotten it from the New York Public Library website. The pool of resources for source material is broadening.

The version Jujutacular ran as his first nomination is better than my early work.

That represents quite a lot of cleanup. His effort really shows on the upper margin and far left. I was on the fence about the nomination--didn't want to rain on his parade yet thought the restoration could go even farther. Took a chance and offered to collaborate. It turns out he's a really good sport, eager to learn, and a joy to work with.

He had a version saved without the histogram adjustment. Smart fellow! We traded off on additional dirt and smudge removal; both of us applied masks to correct the uneven brightness. I added a perspective crop, patched in a margin at the lower right edge, and did the final tweaks with curves and color balance. Sometimes it's magical when the final work feels like a time machine. This was one of those occasions.

There's nothing quite like the moment of enjoying an editor's reaction for the first time when he realizes, "I did this." So cheers to Jujutacular. Looking forward to seeing his next project.

----
Per request, adding links to the full versions:

What is a stain?

2009-11-20T15:59:00.000-08:00

The Tropenmuseum has provided a very special challenge. This post is probably the first glimpse of it to become visible to the public. Roughly speaking, it's something like the Bayeux Tapestry of Indonesia. The museum had it professionally photographed in twenty-nine parts. Then they discovered that none of the digital segments quite matched together: the lighting was slightly uneven from left to right and the thin tapestry fabric had shifted so that the fibers and images no longer matched neatly. These problems affected every one of the twenty-nine segments; the museum staff thought the image couldn't be stitched. We've had very enthusiastic responses from the museum staff now that I've pieced it into a single file. It's a huge job: 992.4 MB. Now it gets interesting.

There's a lot more to be said about the digital restoration on an image that's nearly a gigabyte than one blog post can encapsulate. One part that's easy to recognize is the large brown stain that runs vertically along this segment. Today's post takes a different approach to repair than the last post because this time the original design remains visible within the damaged area.

In terms of digital image data, this stain consists of three elements:

Brightness
Color
Contrast

If one thinks of a stain as too dark and deficient in blue and lacking contrast, then the way to eliminate the stain is to brighten, add blue, and restore contrast. The concept is simple. It's mainly a matter of creating enough masks to make the corrections in small increments. Stains tend to be uneven so different portions require individual adjustments. Here's an interim save as it looks on my monitor right now.

The proportions on this are reduced substantially from the original file, which was 13MB on this cropped area. Yet it can be informative to see the work before the stain removal is complete. No cloning has been done at all in this stain correction.

President of Suriname visits WMF partnered museum exhibit

2009-11-18T10:20:00.000-08:00

Good news from The Netherlands. President Ronald Venetiaan of Suriname has visited the Tropenmuseum exhibit about the cultural history of Suriname. First Lady Lisbeth Venetiaan saw the exhibit last Saturday and was so impressed that she returned with her husband the next day. The Suriname exhibit was developed in partnership with WMF Netherlands and the Open Progress Foundation.

Difficult repairs

2009-11-15T17:54:00.000-08:00

The image above is an interim save on one of my current projects: an early nineteenth century Japanese woodblock print of the ram from the Chinese zodiac. The artist is Gogaku Yajima, who doesn't have a Wikipedia biography yet. It's an attractive print, but as you can see from the lower right corner this isn't in pristine condition. Have already spent about seven or eight hours to bring the restoration this far. What follows is a close-up of the upper border before starting work on the hardest part of this project. Notice the odd mark that branches downward from the upper edge.

At a glance it looks like really sloppy click-and-drag clone stamping, but close up the problem isn't consistent with post-digital alteration. My best guess is that the rice paper print developed a tear and that the mark represents an attempt at physical repair. Someone might have added a reinforcement to the reverse side of the paper and used an adhesive. Then, since the repair could have occurred generations ago, over time the adhesive had a chemical reaction with the paper which resulted in that branching stain. It would be interesting to contact the library staff and see whether this guess is accurate. But for today the challenge is how to make it disappear.

For starters, here's a closer look. The border is a repeating pattern of cherry blossoms. This takes careful study because reconstruction of the design involves inferring the shapes of three flowers and a bud.

Repairs to complex areas such as this are much more difficult than repairs to solid colors. A photograph print that gets torn at a cloudless sky could be fixed in a minute. Damage to near-random features such as a sandy beach or a grassy field are also fairly simple. This is a different matter.

When I look at this type of image I estimate the difficulty in terms of five elements:

How many color borders or line demarcations does the damage cross?
Does enough information remain to infer the shapes of the missing areas?
Is the damaged pattern repetitive or unique?
Is there enough similar data elsewhere in the image to patch a repair?
How much importance will the viewer attach to the damaged region?

This bit of damage interacts with the woodcut design in complex ways. If I want the repair to look seamless, and I do, it's going to take a few hours to get this right. Here are a few close-ups from screen shots as the work progressed. You'll see how, progressively, I fixed areas and inferred colors and shapes.

In the far right frame I actually altered one petal, which is something I don't like to do. That particular area had sustained so much damage that I had difficulty understanding what the shape was intended to be. If I'm a stickler I might go back and recreate that petal again, because after a day I had better ideas. These are the kinds of decisions that make me bite my lip. If the feature were a face instead of a flower I wouldn't have attempted to alter it.

After a bit more cleanup around the edges, here's how that area looks now.

Wiki witches

2009-11-13T19:16:00.001-08:00

Just because you're all so wonderful.

We are all Berliners

2009-11-09T11:09:00.000-08:00

Not that it was likely to kill me, but I happen to be one of the last people who crossed the Berlin Wall illegally.

Twenty years ago it was delightful to follow the news about Czechoslovakia opening its borders, and East Germans crossing that way and seeking asylum. Then when it seemed that political pressure was closing the opportunity, one magical day the people from both sides of Germany stormed the Wall and the guards laid down their rifles.

But the Berlin Wall didn't absolutely disappear overnight.

That fall I was a taking a course on the history of German unification in the nineteenth century. Took a seminar in Germany since 1945 the following spring. An opportunity like this doesn't happen twice so I also enrolled in a conversation course to brush up my German and spent a month in the country after graduation. Reached Berlin in June three days before the authorities dismantled Checkpoint Charlie.

By that time Germany was still two separate countries on paper, but they had already worked out most of the details for reunification. Germans from either side of the internal border could legally cross anywhere, and European Union members had additional crossing points. Not everyone was so lucky.

I was visiting the Brandenburg Gate, pictured above, and there was the wall in the way. Most of it was still standing. On the other side was Unter den Linden--the famous avenue with the city's best museums. One of the new checkpoints was right there. "Reisepass, bitte." The official waved me away after one glance at my United States passport.

The city was changing so quickly that year that travel guides were obsolete as soon as they were published. The only thing do to was follow the news and use one's wits. Berlin was so filled with tourists that I nearly resorted to sleeping on the floor of a convent. Fortunately my language skills were good enough to use the local Mitwohnzentrale. Negotiated a short sublet on a comfortable apartment.

So there I was at the Brandenburg Gate, unable to get to the museums. There was one other thing to see on the western side: a line of crosses. The citizens of West Berlin had erected informal memorials to each person who had died trying to get across. Most of the memorial crosses had names and birth dates. A few only listed the date and time of death. They were all in chronological order; the very last was dated February 1989. "If only that person had waited a few more months," I thought.

Right then I decided I was going to break that law while it was still possible.

The question was how to do it. The subways had resumed operation between the two halves of the city. So the first plan was to blend into the crowd. Wearing Nike running shoes, no one would mistake me for an easterner. But a good accent had gotten me through the Inner German Border on the way to Berlin without absolutely having to undergo a passport check. "Sind Sie Deutsch oder Ausländerin?" Tough choice, but an East German border crossing stamp would make a good enough souvenir. Had told the truth, but the border guard didn't bother to stamp the passport. So a day later, standing at the Brandenburg Gate, it seemed possible that maybe the Hamburg accent I had learned from my grandmother would bring another opportunity to pass for a German. In order to look more local I carried a newspaper under one arm.

It was easy to tell when the train got to the Eastern side: the track suddenly got about three times as noisy. Disembarked to the platform, but then--an amazingly strange sight--every single person pulled out a passport to obtain permission to leave the station.

Dangit.

So this wouldn't be easy. After several hours of playing musical subways and other subterfuges I gave up and went to Checkpoint Charlie. The outer guard was smiling broadly and posing for tourists. The guidebook said to expect about a ten hour wait, but the spot was nearly empty. Walked up to the counter and paid five marks to The Man for a day's visa. The System had won.

By then it was nearly sunset so Unter den Linden was out of the question. The museums had already closed. Wandered a few blocks past one of the first commercial billboards to appear in a Communist country, and walked over to a patch of Wall.

Back during the Cold War, when they were serious about it, I wouldn't have been able to approach within several meters of the Wall from the eastern side. Not if I'd expected to survive. But by June 1990 sharpshooters had quit their jobs and it was possible to walk right up. If I couldn't cross that wall illegally I could at least help to destroy it. That hour did more damage to my Swiss Army knife than to the wall, but it broke loose several handfuls of pebbles that made superb presents for friends back home.

Had to quit before twilight ended.

Normally I have a good sense of direction, but this was a strange city and nearly nightfall. So although I thought I was retracing my steps to Checkpoint Charlie somehow I took a wrong turn. Wound up staring down a dead end and wondering how I had gotten there.

Technically the Berlin Wall was two walls--an outer segment and an inner segment. They used to rake several meters of sand in between every day in order to make it easier to track footprints and shoot people. Above and to the left was a watch tower, but the paint was peeling and it hadn't been manned in months. In a few places the walls had been broken with gaps in between. One of the gaps was there. The sand hadn't been raked in months and wildflowers were growing in it. Wild rabbits were nibbling on the flowers.

A woman came in from the West Berlin side and walked her dog through that stretch. Then a man on a bicycle came up from behind me and walked his bicycle across. They were Germans. It was legal for them to cross anywhere.

Hm.

Looked left, looked right, and thought "What are they going to do, shoot me?"

So I escaped from East Berlin.
----
If you want to honor the memory of the people represented by that line of crosses you can restore the image at the start of this post, the photographer Toni Frissell has donated it to the public domain. The uncompressed TIFF version is available for download here.

Bringing down the house

2009-11-08T17:56:00.000-08:00

Let's face it: really old disasters are fascinating. Pictured above is the aftermath from the the Galveston Hurricane of 1900. If any of those children are still alive the youngest would be about 117 years old now. Galveston has long since been cleaned up and rebuilt. We can call this sort of photograph educational because it probably is--sure makes me want to live somewhere else than Galveston...as a wiki witch I avoid places where houses fall. But heck, this photo also serves the human desire for guilt-free rubbernecking.

Ready for more? Another house, smaller disaster. Published in April 1919. (Yes there's a point to this; continue reading after the requisite gawk).

The Library of Congress had an unconfirmed record which called this a French biplane. Two Military History Project members have identified this as a Curtiss JN-4 "Jenny", which was the best known United States biplane of World War I. So thank you, Monstrelet and Eurocopter! The Curtiss JN-4 "Jenny"...it'll be easier to remember that name now.

Palenque revisited

2009-11-06T10:58:00.000-08:00

The bas relief at Palenque has deteriorated a lot since 1787 when Ricardo Almendáriz sketched it. Thanks go to Infrogmation for spotting this 2008 photograph of the same panel.

And here's another look at the digitally restored version of the 1787 drawing.

Goals

2009-11-05T10:43:00.000-08:00

Awadewit is among the Wikipedians I most admire: a prolific contributor of high quality articles on scholarly topics, and a joy to work with. She's completing her dissertation in eighteenth century children's literature this year, so as a gesture of thanks for her hard work I've done a little restoration for her today. She surprised me last month by saying she's actually used my restorations as examples in the introductory undergraduate classes she teaches. Well here's something closer to her real world scholarship: a William Blake illustration for Mary Wollstonecraft's Original Stories from Real Life.

"Wow - they look so different from the Blake versions - so much softer" was her first reaction. After a moment of poking through the bibliographic notes I confirmed a suspicion: this is indeed the Blake version, scanned from his original sketch for the frontispiece.

Here's wishing the digital file were higher resolution. The Library of Congress has been digitizing its collection for fifteen years; some of the older digitization isn't quite at the technical excellence of their recent work. Equipment and formats have changed, and their staff has gotten better at it. So this might be a viable featured picture candidate at English Wikipedia but the higher Wikimedia Commons technical requirements would exclude it from serious candidacy.

Nonetheless, a medium resolution file is usually less work to restore--the smaller damage just doesn't show up at this resolution. So I did a quick restoration for Awadewit to cheer her on through the final months of the dissertation. First, let's have a peek at the version of this image which appeared in the printed edition.

This is basically a mirror image of the sketch. Maybe the Library of Congress has been spoiling me: the file that was in use at the Wikipedia article was less than 50 kilobytes. It's stained, faded, yellowed, and suffers from distortion, plus the crop is so tight that it would be difficult to correct the problems if the filesize were large enough to work with.

Really, it's best to restore files that are at least 10 megabytes in uncompressed format. I don't call an image large unless it's more than 100 megabytes. On a fairly frequent basis other Wikipedians ask me to restore material that is only a few dozen kilobytes. I can't produce a featured picture from that; the magic has its limits.

Getting back to this project, though, one issue worth mentioning is how to deal with staining. Stains can be a problem when they occur over important parts of an image. In this instance a large brown stain occurs on the central figure's knee. I've blown this up to about double the original resolution.

This stain consists of two parts: a small deep brown spot surrounded by a much larger faint discoloration. First I got rid of the small spot. The easiest way to do that is with a clone stamp at 35% hardness, but with a careful and steady hand it's possible to get a more natural effect with the healing brush. The healing tool algorithm usually works poorly in the middle of narrow stripes, but if the sample point is the same distance from a line as the destination point the tool can give a pleasant blended effect. It's quite tricky to do this and often takes several tries because if it isn't done just right the tool will create an ugly smear.

Unnatural healing brush smears are a hallmark of poor image editing; that's something to really avoid. If you feel like trying this trick the way to get the feel for it is by backing up through the history when a smear occurs (and you will make smears--a lot of them--so don't feel bad). Don't try to re-heal a smear; it'll only smear worse. Just choose a new source point, try to line it up just so, and eventually the spot will vanish perfectly.

Of course there's still the larger stain. Low intensity stains are basically localized color problems. So I treat this like any other local color correction: choose the selection lasso, set feathering to 10 pixels (which is not very much, but appropriate for a small image), and create an adjustment layer to manually adjust the colors. Think of "brown stain" as "deficient in blue" and nudge the RGB sliders until the colors appear consistent. Don't try to be hasty and copy the same tweaks to the light, medium, and dark ranges: often the best results come from slightly different settings.

With additional work the final version comes out looking like this.

Three cheers for Awadewit and her Dissertation! (Now Awadewit, go finish it). ;)

Spinal curvature

2009-11-04T20:59:00.000-08:00

One problem with book scans occurs when the printing goes close to the spine and the page fails to lie flat on the scanner bed. The subject here is a detail from the Mayan ruins at Palenque and the illustration carries particular historic importance: it was drawn by Ricardo Almendáriz in 1787 during the initial excavation of Palenque and his illustrations formed part of the first significant archaeological report in the Americas.

The lower left corner on this image makes it a bit more challenging than the other perspective crops of recent posts because the others could all be corrected with a single perspective crop. This requires multiple perspective fixes. After the first set of changes the lower left corner still curls upward and inward. The effect is most apparent at the horizontal border line. That distortion affects the entire far left column of the image.

We're lucky in one way: other than border, only the lower half of this drawing contains information at far left. So the solution is to select the distorted segment from the seated figure at left and create a new image file solely for that partial figure, then perform a second perspective crop on that segment only. After that, copy and paste the segment to the original image as a new layer and align it so the correction becomes invisible. It's easier said than done.

Here's how it appears after the correction. The change is subtle but it alters the proportions of the figure's hip and spine. Alignment and blending require judicious erasure, partial redrawing of border lines, and minor edits at high resolution.

Other than that the restoration was routine. The drawing is in better shape than most comparable material of the same age, but it still took several patient hours of dirt and stain removal to complete the work.

One of the interesting things about the Almendáriz drawings is that he recorded features which have since been destroyed by exposure to the elements. Comparison against the remaining undamaged bas reliefs shows that his work was unusually accurate for its era. So this remains useful to modern scholars.

Heritage

2009-11-03T15:33:00.000-08:00

At right is a spectacular portrait of Willy Brandt, former German chancellor and Nobel Peace Price winner. He was mayor of West Berlin when the Berlin Wall was built. His likeness was nominated for featured picture at English Wikipedia earlier this year, but failed on technical issues. Brandt's portrait was part of a very generous donation from the German National Archive. Yet at 568 × 800 pixels, file size: 47 KB it just didn't meet the minimum criteria.

Several months ago I performed a courtesy restoration on a higher resolution portrait of Konrad Adenauer, which unfortunately can't be uploaded to Commons because the Bundesarchiv retained full copyright over its high resolution digitized files.

So if it's possible, I'd like to renew the request to Bundesarchiv to seek selected copies of higher resolution material. The lighting and expression on this Brandt portrait is brilliant; he deserves attention on the site's main page.

There's another reason to make this request. There's another image that was nominated for delisting nearly two years ago, and it was retained because a cultural institution donated a higher resolution version. Featured content criteria tend to rise over time, and the lower resolution version had been nominated and promoted before the standards changed. It's quite famous.

The final days of the Warsaw Ghetto Uprising. Many thanks to the United States Holocaust Memorial Museum for their assistance. The lower resolution version had been on the route to delisting. Since then the photograph has been featured on several projects including the Turkish Language Wikipedia. It's amazing how far one donation can go toward raising awareness.

One reason I'd love to be able to upload a better resolution version of Adenauer or Brandt is for balance; not everybody in Germany during World War II agreed with the government's policies at that time. Adenauer and Brandt were among the people who helped rebuild Germany into a respectable country again. It takes a special sort of leadership to steer a country that had gone off on such a bad course back to stability and democracy.
----
Image credit:
http://en.wikipedia.org/wiki/File:Bundesarchiv_B_145_Bild-F057884-0009,_Willy_Brandt.jpg

When you've seen one Edwardian field marshal, you've seen 'em all

2009-11-02T10:41:00.000-08:00

Props to Roger Davies! The Library of Congress staff has emailed to confirm that they will be correcting their records regarding the identity of the man in the recently promoted featured picture at right. This is not Herbert Kitchener, 1st Earl Kitchener but Frederick Roberts, 1st Earl Roberts. Roger Davies noted differences in age and facial structure, and in particular the Victoria Cross at lower right which Roberts received, but not Kitchener. Roger Davies also found two reliable sources for Earl Roberts as the subject of this poster: the Smithsonian Institution and VADS (a joint project of the Imperial War Museum, the British Arts & Humanities Research Council, and Manchester Metropolitan University).

The updated information should be available at the Library of Congress website within a month.

In the meantime we'll need to do a little bit of housekeeping at the wikis because the featured picture candidacies and the relevant filenames all contained Lord Kitchener's name.

This goes to show that reliable sources aren't always perfect. With good reference works it is possible to improve the data that major institutions can share regarding their collections. Here's another recent example: a portrait of John Surratt that the Library of Congress is updating per my feedback.

John Surratt was a Confederate spy who conspired to kidnap Abraham Lincoln in 1864 and who was suspected of conspiring in Lincoln's assassination. His mother Mary Surratt was hanged by the United States Federal Government for her part in the Lincoln assassination plot. John Surratt survived by fleeing to Europe and serving as a Papal zouave; the statute of limitations on most of his activities had expired before he was identified and extradited.

The Library of Congress had misidentified him as the son of "Annie Surratt" and had misidentified his uniform as an Egyptian Army zouave rather than a Papal zouave. The LoC record also dated this photograph "between 1860 and 1875". The source Lincoln's Assassins: Their Trial and Execution by James L. Swanson and Daniel R. Weinberg narrows that time frame to 1866-1867 because the photograph was taken by an American photographer and was sold during Surratt's 1867 trial. Swanson and Weinberg also confirm the other corrections. The Library of Congress staff recently wrote with thanks for the corrections; their website should be updated within a few weeks.

I've started a restoration on Surratt's portrait but gotten sidetracked. Interested in completing the project? Welcoming collaborators; coaching available.

In the bigger picture, this sort of assistance constitutes a significant motivation for cultural institutions to digitize their collections for the public. For example, a Swedish archive called Regionarkivet recently uploaded several dozen high resolution public domain photographs by major nineteenth century photographers to Wikimedia Commons in the hope that volunteers would locate more information about the material. When this type of effort succeeds it achieves the best goals of the free culture movement. I encourage everyone who can provide assistance to join the effort. Digital editing skill isn't required; all it takes is a little research.

Not that it matters, but I've given Roger an award for his research. If people like this I could submit it for formal barnstar approval.

Bright ideas

2009-11-01T19:18:00.000-08:00

It can be delightful to root through obscure corners of the Library of Congress website and unearth something special such as a high resolution full book scan of a beautifully illustrated volume. This self-portrait of William Blake comes from Songs of Innocence and Experience, published 1794. He's an important painter and poet; Wikipedia ought to have a featured portrait.

Every restoration has its challenges. The main challenge in this one is that the aged paper does not lie flat. Scanner light is highly directional so digitization enhances small buckles in the original image. You'll see this as a series of bright and dark horizontal bands near his forehead, chin, and chest. The example below is an interim save after initial dirt and scratch removal.

Distracting, isn't it? The brightness gradient tool can be used here, but it only helps a little bit. The page itself is slightly brighter at the lower left corner. So one quick gradient adjustment at a very slight setting is a first step toward correcting the problems. In this instance I used about 2% opacity. Most of these issues have to be solved another way.

The solution here is to look at each buckled area in terms of what it changes about the image: brighter above, darker below. This can be fixed incrementally with multiple feathered mask selections in a series of brightness adjustment layers: brighten the dark sections and darken the light sections. The trick is to make plenty of layers, draw each one separately, and to change just a little bit at a time. This restoration went through 30 adjustment layers, with feathering staring at 40 pixels and decreasing to 25 and 10 pixels, nudging the brightness about two notches at a time (sometimes one and never more than three). The direct result of that follows:

That's more like it: almost like taking a gentle iron and pressing the paper flat (which of course we don't dare do to the actual physical copy). The final restoration added only moderate changes to curves and color balance.

Images that have nuanced brightness or staining issues can be an exception to the usual time allocation in digital restoration. Usually the bulk of restoration time goes into dirt and scratch removal. In this instance dirt and scratches were minimal, but layered masks required plenty of attention. The most difficult instances I've restored have needed as many as 200 brightness adjustment layers to correct for this type of problem.

Happy Halloween

2009-10-31T12:55:00.000-07:00

As a Wiki Witch I can't really let the day pass without a note about today's picture of the day. This one was created in 1892 to mark the two hundredth anniversary of the Salem witch trials. The revisionism here is really delightful: an innocent defendant, clad in white, calls down divine lightning to smite her accusers. Manacles spring open in the background while the prosecutor drops to the floor and the judges rise to their feet in amazement.

Possibly next year's Halloween featured picture will be one of the Édouard Manet illustrations for "The Raven."

Happy Halloween.

A matter of perspective

2009-10-30T11:08:00.000-07:00

Today's challenge began when GeraldK asked for help getting a featured picture about the Winter Olympics. He's active with the Olympic Games project and is hoping to have something run on Wikipedia's main page in time for the 2010 Olympic Games. The Bundesarchiv material he suggested was far too small for consideration: ranging from 21 KB to 51 KB. Normally 10 MB would be the minimum for my work.

So Staxringold found a 32 MB poster from the Works Progress Administration about the bobsled run at Lake Placid, New York. Lake Placid hosted the Winter Olympics in 1932 and 1938; its bobsled run was converted to public use after the 1932 Olympics.

So we've got a technically adequate file that fits the topic. It comes with a problem: the dimensions aren't rectangular like they ought to be. This is how it looks after 0.7 degree counterclockwise rotation.

This demonstration crop leaves a little background to help illustrate the problem. If you look carefully enough, the caption itself is slanted in conjunction with the distortion. There's an elegant solution to this in the form of a software option which mainly serves the different purpose of perspective control.

When perspective distortion is a problem it usually appears in architectural photography. Back in the days of film photography people corrected for it with a view camera or a perspective control lens. In the digital era this usually gets corrected with software. The same software controls can help the Lake Placid poster too.

In Photoshop this is called perspective cropping. Perspective cropping allows the editor to drag corners into non-rectangular shapes. Then the software performs calculations to generate a rectangular result. To access the perspective crop, select the regular cropping tool and drag a crop selection. A toolbar option will appear above with a check box for "perspective". Check that box and the perspective adjustment becomes possible. In our situation that means aligning cropping tool with the corners of the poster. Just place one corner of crop selection directly onto each corner of poster, then crop. GIMP software users can get a similar effect from the Perspective Tool. Below is the result after additional edits.

Much more satisfactory and probably fairly close to how the poster looked when it was new. If the distortion had been more severe we might have supplemented the perspective crop by separately rotating the poster caption. But this was a fairly mild instance and the perspective crop did the job.

What causes this distortion, you may wonder? Many of the items from the Library of Congress poster collection were photographed for preservation purposes during the film era. These posters were already old and fragile at the time of the photography and they may not have laid perfectly flat. A slight rise or buckling resulted in the distortion. So if the digital file was made from the slide film rather than from the original, the same problems get translated into digital format.

----

Photo credits:

http://commons.wikimedia.org/wiki/File:Cathedral_Notre-Dame_de_Reims,_France.jpg

http://commons.wikimedia.org/wiki/File:Cathedral_Notre-Dame_de_Reims,_France-PerCorr.jpg

Holidays and special occasions

2009-10-29T09:59:00.000-07:00

"You're doing Valentine's Day in October?" A friend commented in amazement when I shared the half-finished work. Well yes, actually it's best to work on these things months in advance. Valentine's Day 2008 taught us that. From the Wikipedia Signpost report of March 13, 2008:

After being featured and de-featured in February, Image:Love or dutyb.jpg, an 1871 chromolithograph by Gabriele Castagnola, has been re-featured. A suggestion by Spikebrennan for a Valentine's Day featured picture led Durova to locate the image at the Library of Congress archives, restoring it as Image:Love or dutya.jpg, where it passed featured picture candidacy on a unanimous vote; however, some users complained that the vote violated usual procedure. In order to feature the picture on the Main Page on Valentine's Day, the vote ran for just over two days, while most nominations run for about seven days. After being featured on the main page, these procedural issues led to a nomination for delisting, which resulted in the image's de-featuring. Afterward, the image was renominated as Image:Love or dutyb.jpg, where it passed on a second unanimous vote. This is believed to be the first time that a featured picture has been nominated, promoted, highlighted on the main page, delisted, and repromoted in less than 30 days.

That's quite a journey for an innocent little depiction of a nun and a painter...well, almost innocent...

The featured picture regulars reacted in good spirit. It was a hoot, but it was also a learning experience. Holidays and special occasions really do need advance preparation.

Other editors are welcome to contribute holiday and special occasion featured picture candidates. Chinese New Year, Mardi Gras, and April Fool's Day 2010 are wide open.

Good news about Booker T. Washington

2009-10-28T13:21:00.000-07:00

Early this year I blogged about the unsuccessful search for a quality portrait of Booker T. Washington. It's a good thing to retry searches occasionally because the Library of Congress keeps adding to its digital selections. Now they provide a wonderful image of him. It has good composition, has been well curated, and is technically better than Wikimedia's software can accommodate. This cropped version is 126MB in 16 bit TIFF format. Wikimedia uploads are limited to 100MB and 8 bit TIFF files. We'll have to reduce the technicals a little bit to get this uploaded. So the media editors will need to start tugging on developer sleeves.

This is a good problem.

At full resolution quite a bit of work remains to be done. Glass plate photography is a very delicate medium and almost always needs attention. The technical condition of this one is better than average. Here's a closeup from part of the chair.

Cheers to Xavexgoem; he's working on this now. I had gotten a start on the Washington restoration and then been pulled away by other projects. Had been feeling guilty about not returning to this sooner. Xav was looking for something to do last night and loves the portrait as much as I do. We decided to trade off and share the work.

This is what Wikipedian collaboration is all about. Now does somebody have a talent for tugging at developers' sleeves? It would be a shame if the full version of this restoration existed on only two people's computers.

Mountains and molehills

2009-10-27T13:28:00.000-07:00

There's nothing like a rare photograph of an important historic event. The one above is the eruption of Mount Vesuvius in 1872. The shot is magnificent and the scan is superb: 217MB at full resolution. After rotation and cropping the uncompressed version comes to 93MB--just within the upload limit for Wikimedia Commons.

As you can probably tell, though, this image needs a lot of work. At full resolution it's really daunting: thousands of spots plus substantial areas of chemical damage at the slopes and coastline. So today's post is about what to do with images that have a lot of problems.

First, this isn't a beginner project. Be realistic about your time and skills. This is the sort of restoration one works up to doing, and even after hundreds of others I feel like I'm on a fool's errand for trying it. A closer look may convey that more clearly.

This is one small bit of coastline. The full version as it sits on my computer now is 6684 x 4857 pixels; I've been working on it with a tool 9 pixels in diameter. That means a whole lot of labor and some of the choices will be difficult, such a what to do about the patch of chemical damage at lower right. Most novices approach this sort of problem by prioritizing the large areas; I do the opposite and work on the small problems first. See below.

Eventually the large problems such as the chemical stain will have to be patched in with repairs from other parts of the image. So it makes perfect sense to start by resolving the less serious problems at those other areas. I'm lazy; I'd rather resolve a problem once than copy it somewhere and fix it twice.

Also, clearing up the little stuff gives time to think about the hard part. This particular stain covers coastline, part of a town, part of the Vesuvius slope, and also--at its very bottom--part of a sailboat. That's going to be a monster to resolve.

So the first question to ask before even deciding whether to do a restoration is to look for the hardest problems and evaluate whether it's within one's skill set to address them. I think this is resolvable, although not necessarily sure exactly how. One of the advantages of clearing out the minor problems first is that it eliminates distractions and makes it easier to focus on the difficult issue. Another advantage of this approach is that it gives the subconscious time to work out a resolution.

My subconscious is much brighter than I am. It makes me look smart. I do the simple routine fixes, the subconscious goes to work on the hard problems. Then once the subconscious has pondered the matter for days I steal its solution and take credit. It's a nice arrangement for me; the poor exploited subconscious has nowhere to complain.

On a more serious note, here's a closer view of the unaltered and partially restored slope. You can tell that I haven't resolved everything yet--just working conservatively from the smaller and simpler problems up to the larger and more complex ones. I usually take a particular tool and adjust it to particular settings, then sweep through the image in a grid pattern doing everything that tool setting is appropriate for. After several sweeps and several different tools and settings the image will approach completion, and be ready for specialized attention to the larger problems. The tool used here was the healing brush at 9 pixel radius, 100% hardness.

One last thought before parting: Vesuvius is a gorgeous location where humans really are better off visiting than living. You've heard of Pompeii? That was Vesuvius's doing. Maybe a great find for archaeologists, but I'd rather contribute to posterity by digitally restoring images than by curling into a fetal position, succumbing to poisonous fumes, and getting buried in volcanic ash.

Vesuvius wakes up and gets temperamental on a regular basis, but not quite frequently enough for the human lifespan. Its last significant eruption was 1944. Have a look at its slopes today: dunno about you, but I sure 'n heck wouldn't want to live there.

Maybe the utility of restoration work is to serve as a reminder.

Chromolithography

2009-10-26T13:51:00.000-07:00

For digital restoration work it is important to develop a working understanding of historic media. A given medium behaves in predictable ways even when applied in very different styles. Two restorations I worked on yesterday are examples of that. The medium is chromolithography. The example above is a World War II propaganda poster, which an editor who goes by Staxringold selected in part because of its racial overtones. Below is an example nearly a century older: an 1847 print about Irish political leader Daniel O'Connell.

The Library of Congress doesn't specify which medium O'Connell's portrait is. I can't specify it onsite because that would violate Wikipedia's policy regarding original research, but to my eye it's very clear that this is either a hand painted lithograph or a chromolithograph. One gets to know these media after a while; one learns how they behave. Limestone produces a recognizable grain upon an image. Despite the vast differences in content and style, both of these files bear technical similarities.

There's a lot to be written about how chromolithographs behave during digital restoration. One place to start is the slightly uneven color application which is inherent to this process. Because of that, the two principal tools of digital restoration--the clone stamp and the healing brush--have to be used judiciously. One problem that's important to avoid is the tile effect. With the example below the same two-part motif appears almost in a diamond pattern.

The problem here is caused by repetitive use of the same source point upon multiple destination points. Fortunately it can be corrected in a large open space by resampling from different source points and reapplication over the affected area. The ideal to aim for is a natural randomized pattern like that of real limestone grain.

Renomination

2009-10-26T09:53:00.000-07:00

One of the hard things about the consensus process at the wiki is that sometimes one comes away thinking "this one went the wrong way". This particular restoration went up last January and got unanimous support, but only three people reviewed it which wasn't enough for promotion.

Two days' hard work went down that way, and although the subject, Edward M. Thomas, was a war hero there wasn't quite enough information to start a separate biography for him; he was one of the Tuskegee Airmen. So the image remained at photographer Toni Frissell's biography and a couple of other pages, and I moved on. Which wasn't easy after having read the text of Captain Thomas's Distinguished Flying Cross award; he's the sort of person who deserves to be remembered.

Which was why it felt so good yesterday when a Wikipedian who goes by the username NuclearWarfare found it and decided to give it a renomination. Very pleasant surprise to see him take an interest in it, and even more so to see the editor reviews. Thank you very much!

Susan Sarandon doesn't know what Wikipedia is

2009-10-23T10:38:00.000-07:00

Andrew Sullivan of The Atlantic Monthly has picked up on David Shankbone's recent blog post about an encounter with Susan Sarandon, where Sarandon said she didn't know what Wikipedia is.

Well good for her! It's refreshing to find a celebrity who is not fussed about her Wikipedia biography and not vain about the portrait that runs at the lead. David took the shot at right last year at the Tribeca Film Festival.

She's one of the best actors in the business, she's been making films since 1970, and she looks fantastic in more ways than one.

Kudos, Susan. Keep it real.

Anniversary

2009-10-22T07:54:00.000-07:00

Today marks my four year anniversary as a Wikipedian. Another featured picture promotion came in last night, so I have 333 featured credits now. Perhaps in another four years I'll have 666 featured credits and sell Wikipedia to the devil... ;)

Raising the color bar

2009-10-19T09:15:00.000-07:00

During the twentieth century one technique for archival preservation was photography on high quality film This could be invaluable in case the original document later gets lost or destroyed. Eastman Kodak, which made the film, provided a grayscale and color bars. And that is the source of today's dilemma.

It's a reasonable first reaction to the presence of color bars and grayscales to apply them mechanically and call the result original. But looks can be deceiving: in digital restoration it's always important to think about which sort of data is useful and which isn't.

The key thing to bear in mind is that the color bars represent data at the time the photograph was taken. They do not yield data on how this poster looked when it was new.

This World War I poster was created in 1916. The earliest color Kodak film was produced in the 1930s. Color film rephotography remained in use until digital technology replaced it. So any World War I poster that got preserved by this technique would have undergone at least two decades of decomposition before the photograph was made.

An unknown amount of distortion occurred prior to photo reproduction: if you've ever spent a summer on the Eastern Seaboard, consider the damage that twenty humid summers could inflict. Especially without air conditioning.

But twenty years is the bare minimum here: the date of this particular reproduction was not recorded. The bars and grayscale might be somewhat useful if we knew what year this was taken: 1946? 1966? 1986? That time frame is too large to be useful, which is why I crop out these bars and ignore them. That may sound radical, but yes I ignore them.

The reason becomes obvious when working with this material: substantial decomposition had already taken place. The lower half of the poster is more deficient in blue than the top half, with additional localized problems such as the pronounced vertical yellow band at far left. Chromolithography from this era can behave that way as it ages. The edges had dried and darkened more than the center, which required radial gradient masking to correct, and the brittle poster did not lie flat for reproduction: it needed a perspective crop, then half of the caption had to be rotated separately in order to recreate the original alignment, and then separate corrections had to be applied for the effects of uneven lighting.

So it wasn't until today when another editor at Wikipedia's featured picture candidates insisted that the gray scale bar gives definitive data on how this image looked when it was new, that I located a more specific demonstration of why it does not. First, here's the gray scale bar.

Do you notice something odd about that gray scale and color bar? Let's look in closer.

Copyright Eastman Kodak Company, 1977. That's right: this poster was at minimum 61 years old when the photographic copy was made.

A sidelong glance

2009-09-24T22:39:00.000-07:00

Wikipedia's internal manual of style says portraits shouldn't look away from the text. At the actual biography this one looks away. A lot of them do. In this instance that choice especially apt; this is George Atzerodt. The night Abraham Lincoln was assassinated, Atzerodt had plans to assassinate Vice President Johnson. Atzerodt was an associate of John Wilkes Booth. Except after Atzerodt arrived at the hotel where Johnson was staying he didn't go upstairs; he stopped at the bar.

And got drunk instead.

Atzerodt could have notified the authorities and perhaps saved Lincoln's life; he didn't. Atzerodt could have backed out of the conspiracy; he didn't do that either. He could have tried to kill Johnson. Instead he went on a bender. This portrait was taken after his arrest. The irons on his wrists are slightly out of frame. He went to the gallows after having earned the respect of nobody.

The look on his face seems to summarize the man: intensely avoiding eye contact.

The Atzerodt portrait was a restoration I did several months ago. There's another set of portraits from the same group that's more challenging. Alexander Gardner's portraits of Lewis Powell have an informality that was rare for the era. Powell was twenty-one years old. He has the build of a college athlete, except the manacles are visible and he leans against the gun turret of a Monitor class warship. Powell had stabbed Secretary of State William Seward repeatedly in an attempted assassination. It was a lucky chance that Seward survived. When Powell assaulted Seward, Seward was bedridden and wore a neck brace. The brace prevented Powell's knife from reaching the jugular vein.

Two years later Seward organized the purchase of Alaska.

A neck brace and some drinks at a hotel bar changed history.

Powell is hard to crop. It's also difficult to select just one portrait; he's very natural in front of a camera. So I haven't worked on him yet. Maybe someone else will beat me to it. There are a lot of worthwhile images out there.

Let's have a TIFF

2009-09-11T12:13:00.000-07:00

Wonderful news from the Wikimedia Technical Blog today: Wikimedia Germany is developing full TIFF support for Wikimedia Commons with implementation scheduled for November.

Earlier this year the WMF developers enabled TIFF uploads to Commons for the first time, mostly at the request of the growing restoration community. TIFF is a file format that's widely used for many real world applications including archival digitization, but most web browsers do not display it and WMF software is unable to render it. The new development will make WMF more TIFF-friendly.

TIFF support facilitates file uploads directly from museums, libraries, and other cultural institutions. It also aids wiki-style collaboration among image editors. The most popular format for online display (JPEG) is a compressed format that undergoes progressive deterioration from repeated saves and edits. TIFF support helps media editors improve on each others' work without compromising quality.

Although TIFF is not the only uncompressed file format hosted at Wikimedia Commons (PNG is another), TIFF's dominance in off-wiki uses means that full TIFF support adds long term scalability to our acquisition of encyclopedic media files.

Kudos to WMF Deutschland!

Lurk-o-meter

2009-09-10T17:34:00.000-07:00

Wikipedian MZMcBride recently launched a new tool that reports how many people watchlist a given page. This has interesting uses. Word first reached this direction last night. Today Jehochman's page update served as a reminder (because, of course, I watchlist it). Wikipedians being what they are, I soon ran a search on Jehochman whose userspace is being watched by 263 people. Comparison to Jimbo Wales's user talk (1893 watchers) shows that Jehochman's user space receives 13.9% as much attention.

So to coin a unit of measurement, the centijimbo is 1/100th the number of watchers that Jimbo Wales has. One centijimbo currently equals approximately 19 watchers. It's a relative measurement so this could alter as people add and drop watchlist pages.

Here's how several prominent Wikipedians measure up:

GiacomoReturned (featured article contributor): 5.4 centijimbos
Roger Davies (arbitrator): 6.8 centijimbos
Elonka (administrator): 15.6 centijimbos
YellowMonkey (functionary, featured article review coordinator): 19.0 centijimbos
JzG (administrator): 27.7 centijimbos

The two most lurked Wikipedians other than Jimbo himself appear to be JzG and Raul654, who have 525 and 524 watchers respectively.

One might hold interesting debates over what these numbers mean, which factors affect them, or why the Wikipedian with the highest centijimbo rating has a deleted userpage.
----
Juliancolton (13.8 cj) has noted that SlimVirgin measures 32.9 centijimbos.

A touch of class

2009-09-08T21:30:00.000-07:00

One of the most fascinating photographic exhibits I've attended was a show of George Hurrell's portraits at the University of Southern California library. Hurrell was a master of Old Hollywood glamor. It's hard to find material of his under free license so here's a link to his iconic portrait of Jane Russell. Hurrell was a very talented photographer whose legacy is augmented by two very fortunate circumstances: his subjects were among the most beautiful human beings on the planet and Hurrell employed a stellar team of retouchers. During an era when plate photography was getting supplanted by film for most purposes, Hurrell shot on oversized glass plates; his retouchers would have half an hour to retouch the original negative before it dried. The results require viewing at close range to fully appreciate: they don't feel artificial. Errol Flynn looks genuine down to the pores on his cheeks, except it isn't a close shave. Those pores look like he never grew stubble at all.

Hurell's work came to mind tonight as I was dabbling with a very different undertaking. Although it wouldn't be right to beautify a portrait subject for encyclopedic restoration, reconstructing data on photographic prints does bear similarities. Bad image editing takes on an airbrushed appearance similar to a weak attempt at physical retouching. Good restoration work on high resolution digital files includes reconstructing photographic grain at damaged areas. The finished image needs to appear credible. For Hurrell's studio it wasn't enough to take out stubble; they had to rebuild Flynn's pores. It isn't enough to take out dirt and scratches; I rebuild grain.

Images that are at the margins of restorability fascinate me. This one of Cy Young from 1905, for instance. It's shot through with problems: faded, stained, smeared, dirty, and creased across the middle. At full resolution it also has fingerprint marks (remember from Honus Wagner how difficult fingerprints can be?) and the Cy Young print even carries a mirror image imprint from a ball point pen. The print was probably stored with a sheet of handwritten paper facing it, and over years the ink transferred. It'd be a very tall order to repair the damage. If an editor gets that far, there's also the dilemma of whether to crop away part of the stadium to get rid of two distracting figures at far right. Plus it's contrasty and Young's posture is awkward. But then it's also a wonderful little scene of a professional ballpark in use at the start of the twentieth century. And it is Cy Young.

This type of image is digital quicksand: alluring enough to want to try, damaged enough to be monumentally hard, and no matter how good the restoration is there's a serious chance that reviewers will reject it at featured picture candidacy due to the underlying compositional flaws. I specifically asked Staxringold to stay away from it. And then, being a sucker, I dabbled at it myself.

Here's a detail of one of the more heavily damaged areas. The wispy horizontal stripes with a bluish hue are surface glare from the reproduction. The print buckled when it got creased, so digitization picked up undesirable details of the print texture. Accompanying these lines are a set of fainter horizontal dark bands that received too little light during reproduction. A fairly large stain is also visible at left. The challenge is figuring out what to do with this. It's editable: most of the important data hasn't been harmed directly. I've resisted the temptation to drag a clone stamp across it because it's important to retain the impression of a crowd scene.

This is where it's come after more effort than I'd care to admit. Obviously nowhere near completed. It's been more like an exercise in reconstructing grain and texture: rebuilding the wooden beam, melding the brown stain, reducing highlights on the folds. Most of this work was done six pixels wide with a healing brush at 200% resolution. It builds skills to toy around with heavily damaged material. Like Hurrell's retouchers, the art is to reconstruct something believable in place of the problems. It's time to set this one aside and think of alternate solutions. Possibly next I'll separate out the blue channel. The subconscious can be a wonderful assistant; the subconscious can chew on this task for a while.

Blue sky speculation

2009-09-06T12:28:00.000-07:00

Occasionally a restoration project goes onto the back burner for a while. Most of this image was routine. It's a photographic print from the end of the American Civil War that shows the ruins of a railway roundhouse after Atlanta was burned. The serious problems were in the sky.

The project started as a scan from a high resolution album. A fascinating historic document: at the time when this war happened photographic technology was just reaching a level where serious visual documentation was becoming possible. Quite a few of the prints from this album may become featured pictures eventually. Roundhouses were sturdy structures. To see this one almost completely demolished is a powerful reminder of the destructive force of that war.

Relatively speaking, skies are easier to fix than other elements on a photograph. The data beneath the horizon was basically sound so I selected this print as a project. But this isn't a simple blue sky; those cumulus cloud formations turned out to be quite frustrating.

This sky has several problems: brightness differences, creases, and stains. It looks like a combination of chemical and physical damage affected the upper center. A lot of work went into correcting these features and I wasn't confident about the result.

It's a matter of how high to set one's standards. On every subject there will be a few people who spot inconsistency. When a television program switches props between scenes, ninety-nine people out of a hundred may notice nothing, but there will also be one person who says, "Wait! He was wearing a three-quarter helmet when he parked the motorcycle. But now he's walking into the house carrying a full face helmet. That makes no sense."

A pure blue sky is easy to repair, but cloud formations are meteorological phenomena. Would this edit insult the intelligence of a person who understands meteorology?

In true wiki fashion I asked another amateur editor. Juliancolton isn't a meteorologist, but he has written about twenty featured articles about hurricanes. That shows either very good judgment or very bad judgment on my part; take your pick.

When Julian received both images he was slow to reply. At least he wasn't indignant about the edit. After discussion he agreed it looked credible.

Which is a relief. But since Wikimedia Commons is a wiki I'll also be uploading a partial edit (without hisotgram or color adjustment) in case someone disagrees.

Lifting fingerprints

2009-09-05T11:42:00.000-07:00

Fingerprints pose challenges to restoration. Above is a photograph of Honus Wagner, circa 1911, which is available in a 55MB digitized file from the Library of Congress. It is a pleasure to see institutions provide such high quality copies, yet with material of this age the odds are pretty good that someone has been sloppy about handling the document. In this instance the biggest problem is a fingerprint beside the right elbow.

Getting in a bit closer reveals what a difficult issue this is: the fingerprint whorls are roughly perpendicular to the seams and creases of Wagner's uniform and they continue past his uniform to the grass behind him.

What we want to do is retain the meaningful data while getting rid of the fingerprint. A lazy approach would be to remove the brightest fingerprint highlights. That won't be good enough: the human brain is excellent at pattern recognition. Subtle sets of parallel lines will still be apparent. So it's going to be necessary to recreate the pinstripes on the sleeve, the creases on the pants leg, and the relevant shadows.

Whenever a restoration task seems daunting a good approach is to do the easiest part first. A little time and familiarity makes it simpler to work out the rest. So we'll start at far right. I used the Photoshop healing brush tool at 100% hardness and 9 pixels. An important trick to getting good results is to change the source sampling point frequently.

Moving left to the edge of the leg, the healing brush becomes more tricky. The tool's algorithm draws data from outside the selection area, which tends to smudge borders. Most people switch to the clone stamp in this type of situation. I recommend around 35% hardness for clone stamping in most restorations, including this one. The reduced hardness usually avoids the cookie-cutter appearance of novice image editing.

Editors who are more ambitious could try what I did here: it is possible to use the healing brush effectively at borders if the source area and destination area are close matches. I reset the tool diameter to 5 pixels and sampled from a border section lower on the leg, then applied the tool to the fingerprint marks at the edge of the bright area. The advantage of this approach is that when it works it yields a very good blend. If the results aren't desirable at the first attempt, just back out through the history and start again.

Although the clone stamp tool is good for restoration at low hardness, I nearly always keep the healing brush set to 100% hardness. Anything less gives muddy results. Here's how the area appears after a few minutes' work.

Now it's time to handle the toughest sections. Most of the work on this photograph has been done at 200% resolution. For this part we'll go in at 300% resolution and use both the clone stamp and the healing brush. The aim is to infer meaningful data about the arm, elbow, shadows, and uniform from the areas that haven't been marred by the fingerprint. The results won't be perfectly sharp because the photo is no longer sharp at such high resolution. Part of the goal is to generate a slightly mottled appearance that looks compatible with undamaged areas elsewhere in the photograph.

There's the result at high resolution. Now we'll pull out and see how this blends with the surrounding area.

This is not a complete restoration, just a partial screenshot of the image as it exists on my computer as of this writing. The fingerprint itself was rather large and this demonstration didn't erase all of it. Notice the subtle grooves above and below the area that was worked on? That's what I meant about the human brain being an excellent at pattern recognition. Other parts of the fingerprint still remain. This work goes more slowly than ordinary scratch and dirt removal.

Fingerprints are chemically interesting. From a preservation standpoint that means there's a strong likelihood that oils and dirt will cause chemical reactions on a photographic print. Those changes can't be brushed or washed away. Attempts at cleaning often cause additional damage. When you need to scan a damaged print, might as well leave bad enough alone on the physical item and correct the problem digitally.

So if you have a family member with bad habits at handling old prints, sit them down and show them this post. Better yet, fire up an image editing program and task them with cleaning up their own mess. Whether the goal is to preserve baseball history or family memories, it's much less effort to wash one's hands and hold photographic prints by the edges.

Which way is up?

2009-09-04T09:09:00.000-07:00

It only takes a moment to sense something off kilter with this shot. Notice how all the buildings slant to the right? This image needs counterclockwise rotation. Many rotations are easy; this one isn't. The photograph has mild perspective distortion. No matter how one rotates it doesn't feel correct because the verticals on these buildings aren't quite parallel. Also the sidewalk slopes upward from left to right. Decreasing the slant on the structures means increasing the slant on the sidewalk. So unless an editor is very careful about which cues to follow, something is going to come out wrong.

It helps to know that this is the Brooklyn Daily Eagle storefront, photographed in 1916. Most of Brooklyn development occurred between the American Civil War and World War I. This was relatively new construction when it was photographed so we can discount the possibility that significant shifting or settling would have altered the buildings themselves. Let's look for trustworthy points of reference within the image.

The man descending the stoop uses canes, so he might not stand upright. But what we can depend on is that the steps would be level. Looking closely at the base of this stoop confirms my suspicions about the sidewalk: notice how the base at the left edge of the stoop is nearly twice as long on its left side as on the right side. This was compensation for sloping terrain.

Here's how that stoop appears after rotation and spot removal. These weren't the only lines of reference when I rotated; I also went by the building verticals and roughly split the difference in terms of angle distortion.

Here's a glance of the whole image after rotation, cropping, and several lengthy rounds of dirt and scratch removal.

It's better than things started, but the distortion still leaves an uncomfortable sense that the lamppost is going to keel over onto something and the French restaurant at right looks as if it is leaning into the Eagle offices. This keystone effect is subtle, but it's there. The solution is to perform a mild perspective crop.

Here's the end result. Added a touch of curves adjustment to bring out contrast in the midtones, plus localized brightness and contrast corrections. The automobile at far right no longer fades toward the edge of the frame. Let's hope those ancient cars had good brakes because the driver didn't angle the wheels toward the curb. Otherwise, nothing looks askew.

Synergies

2009-09-03T10:14:00.001-07:00

Surprising news arrived today. The discovery of human remains from the Wounded Knee Massacre aftermath restoration has been incorporated into the official program notes of an exhibit at the Montreal Museum of Fine Arts, item number 76. The Library of Congress also credits me in its updated bibliographic record. The museum is exhibiting an unedited version of the image; possibly the presence of the bodies was a factor in its selection.

The individual who shared the news does not appear to have a connection to the museum. Would love to make contact with Quebecois Wikimedians who could follow up on this and establish ties with the museum.

Facing facts

2009-09-01T11:04:00.000-07:00

A day later the portrait of Ed Walsh is finished. Staxringold and I conominated it at the end of the evening, and overnight it became relevant at Jimbo's user talk. Here's a link to the ongoing discussion as of this writing.

An editor has found something objectionable about an attempted restoration (not one of mine, and not a featured picture) and followed up a complaint to the photography wikiproject with a complaint to Jimbo. The first I heard about this was a ping from a friend while I brewed the morning coffee. Jimbo had already brought my name into discussion and his description needed followup. That context and followup have been posted to his user page; another thought deserves space at the blog.

Occasionally someone who doesn't do restoration at all comes along and attempts to dictate parameters that wouldn't work. This doesn't occur very often, and it's been a while since the last episode. In this particular instance the editor wants to restrict the editing of historic images to levels adjustment in order to prevent what that editor perceives as original research. It's worth mentioning here that the Ed Walsh example is particularly telling in that regard.

Despite all the time that went into scratch removal, the most significant change between the original and the restoration is correction for the overexposure on parts of his face. That correction was done by using the shadow/highlight tool on a selection of Walsh's face and collar. This is a sophisticated type of levels adjustment and if I had wanted to misuse it the result would have made him look like a wrinkled old man. Compared to the histogram work (which didn't take long but was actually quite delicate) all the preceding hours of dirt, scratch, and stain removal amounted to very conservative editing. Walsh already has bags underneath his deep set eyes. A bad histogram adjustment could make him into a zombie.

Back when Johann Wolfgang von Goethe completed the second part of Faust, a publisher objected to some of the language in the Walpurgisnacht scene. Faust obligingly edited per the instructions although there was nothing all that offensive about the particular words; it was likely the pagan context that might disturb devout Christians. A later publisher commented that the Walpurgisnacht scene actually becomes less offensive with the original wording restored. You see, Goethe not only did as he was told; he also censored things that hadn't been part of the publisher's instruction. So "The witch farts" became "The witch f---s", which carries exactly the same more obscene implication in German as it does in English.

That can be the result when an artist is compelled to operate within restrictions which he regards as dunderheaded.

A hairy situation

2009-08-31T11:44:00.000-07:00

One of my biggest worries with image restoration is that somebody will come along and assume the whole thing was done in fifteen minutes with a couple of filters and plugins. That's not how I work: all of this dust and scratch removal is done by hand. So why not use plugins? For lack of a better word, because software is stupid. A plugin doesn't understand context. Sure it'd save time; for people who care more about time than about losing data that has its place. So Staxringold is working on another baseball restoration and I'm helping him out with the hard part. Above, that's Ed Walsh in 1911. The hairy situation is--of course--his hair.

This is a part of Walsh's hair as it looked when Stax reached the end of his tether and asked for assistance. Dirt, creases, and scratches are not necessarily created equal: junk in a clear sky is easy to fix; degradation that intersects significant image data is harder to resolve. Lots of scratches and hairline cracks that interact with a mop of tousled hair is a nightmare for restoration. I don't trust software to know the difference: it's a thing to reconstruct by hand.

After several hours' work on the portrait this segment has gotten this far. With these matters it's often a question of how detailed to get. Do you go with the quick solution? Do you call it 'good enough' and quit? Or do you do it right because it ought to be right?

At 300% resolution, here's where this is right now. The biggest problem remaining is a scratch on the right side that runs horizontally: it's almost parallel to a nearby lock of hair. Also at lower left there are several faint vertical cracks which haven't been eliminated completely yet, plus a few other things to do.

I don't like to put my name behind slapdash work, so I'll be tweaking this a while yet. Walsh has a lot of hair (and the rest of the portrait still needs tweaks also). Stax is a good guy: when I told him how much labor this was turning into and asked for a favor in return, he accepted three choices:

1. Start a biography for Mignon Nevada. She's a soprano whose portrait I restored yesterday. The daughter of Emma Nevada.
2. Improve the biography of Duke Snider.
3. Restore a photograph from the first successful transit of the Northwest Passage.

Stax selected the third option. So now he has a view of the Gjoa weather deck when it was at Nome, Alaska. The transit took three years; Roald Amundsen and his crew were iced in for two of them. Amundsen succeeded after everybody else had failed because he took a small ship and a small crew and lived off the land. "Cool!" was Stax's reaction when he saw the photo.

This stuff is infectious.

Friends and memories

2009-08-30T09:32:00.001-07:00

In response to JoshuaZ's comment, following up with tips for the situations where readers want to remember friends and relatives. Above is a portrait of Gerard Meijssen. Not something I shot (he's in a country where I've never set foot), but a typical snapshot portrait. Today's tips are about things to do for friends and family.

Gerard stands on a balcony, probably because the decision to take the portrait was made indoors. This was the nearest location which had enough natural light to shoot without a flash. Overcast days are good for portraits: diffuse light flatters the skin. It's best to frame the shot so the gray sky remains outside the picture. This is one reason why family photographs that mix scenery with portraiture seldom turn out well: when the light is good for the setting it's usually far too harsh on faces.

Amateur portraits tend to be contrasty. Even though the light is diffuse Gerard's face still has harsh shadows. This happens because no camera responds as well to differences in brightness as the human eye does. For this informal balcony portrait, taken on the spur of the moment, I would have directed Gerard to stand at the far end of the balcony. That would permit the use of a longer lens setting to eliminate the slight wide angle distortion and allow a framing that would have gotten more of the softer tones from the trees and the red brick building. The thing to do would be frame the shot to minimize the gray sky and eliminate the cellular phone store completely (it's colorful but it's distracting). Even a gray sky will be very bright in early afternoon; reframe to minimize sky and the camera would handle the face better.

Also this shot would work better as more of a close-up to minimize the wrinkles on his shirt. This photographer is significantly taller than Gerard. It's unlikely I'd tower over him, but if that happened I'd bend my knees to bring the shot to eye level. Photographers who are vertically challenged do well to remember this tip: it often yields charming results when photographing children.

But...this is what we've actually got to work with. It isn't bad for an amateur portrait. Let's tweak it a little. The first thing I did when I saw it (Gerard was running for the WMF board) was bring out the shadows in his face using the Photoshop shadow/highlight tool. It's a manual adjustment that helps get the facial shadows back to what the human eye would perceive. Also added a mild linear brightness gradient to reduce the shadows globally. Here's what resulted.

This type of minor change usually isn't a problem; this version ran at Meta during the latter part of the recent board election where Gerard was a candidate. Suppose we were editing for a private album? Well, for family and friends one can go a little further.

This adds a second gradient mask to put a little blue into the sky. Also brightened the whites of his eyes very slightly. Brightening the eyes is a retouching trick that makes people appear healthier. Also I'd still rather have a closer crop.

So here's what we get after ten minutes of editing. It's a little more processed than would be ideal for article space, but fine for friends and family.
----
Source:
http://meta.wikimedia.org/wiki/File:GerardM.JPG

Directing subjects

2009-08-28T11:34:00.002-07:00

It's a pleasure when people share news about their work with images. Igor Berger pinged me yesterday that he had gotten a new camera and used it on a trip to Japan. Good for him! He uploaded several examples to Flickr and asked about adding them to articles. So this post is for Igor and others like him who want to take better photographs and help the encyclopedia.

The Walter Johnson/Calvin Coolidge handshake is provided again as an example where the photographer did not direct the subjects. The most obvious problem is that President Coolidge stands behind a barrier. Most amateur models have no idea where to stand for a camera. Yes, this includes famous baseball players and presidents. So ideally the photographer would move three paces to the right in order to get that barrier out of the way. That might not have been possible: the type of camera needed a tripod and there may have been stairs to the right, or perhaps other photographers had already taken the best positions for the publicity shot. So suppose you were operating that camera and couldn't move, but you could talk to the subjects. That would sound like this.

"Excuse me, Mr. President? The photograph would look better if you took three steps forward. Mr. Johnson, would you step to the side three paces with him? Yes, very good. Excuse me sir, in the pale suit? Would you make room for these two gentlemen please? Yes, and face toward them. Everybody! All eyes on the President and Mr. Johnson please. Pretend I'm not here. May we have a round of applause? A big round of applause for our great pitcher! Wonderful." (snap) "Great, let's try another..."

Whether you're shooting a birthday party or a presidential visit, it always helps to shoot more more photographs. Don't be shy. If you sound like you know what you're doing people cooperate. This used to be expensive in the days of film photography; not anymore. Memory chips are cheap (and you don't have to pay a developer). Best of all, you can preview the shots and delete the obvious duds. Take advantage of that; experiment with different ideas.

Before the photography actually begins, look over the location and decide which parts would make the best settings. You'd be surprised how many people go to a picnic and line up to pose in front of the dumpster instead of beside the flowers. Take charge of that situation. Your subjects will thank you.

The one exception to that rule is vacation photography. I've never understood the logic of plopping Uncle Joe in front of the Grand Canyon. What's the purpose? Are people afraid they'll be accused of faking the trip, so they document that they really did drag Joe past Barstow over to that big gorge in the desert? Might as well hand him the day's newspaper so he can hold that up too and prove when you were there. Great scenery looks better without family obstructing it, no matter how much you love them. Ansel Adams didn't drag his mother in front of Yosemite; he knew what he was doing. When you want to shoot a portrait, shoot a portrait. Half Dome looks a lot better without Mom in the way.

And please don't make them say 'cheese'. Teeth may appear but nobody really looks happy like that. Most amateur models have been taught to put on a big fake frozen grin when they get in front of the camera. Here's one of my secrets: to capture a good expression, talk to your subject and fiddle with the camera. Have the shot all set up as you start a conversation on another topic (something interesting and fun) and pretend to fine tune the controls. The aim is to distract them into a genuine expression--at the point where that 'cheese' smile finally drops away and they're really reacting, then snap the shutter. You may have to dodge for cover under the wrath of "But I wasn't ready!" Used to be, I'd lay low for days until the prints were in hand. Now things are easier: there's a preview screen. Most people aren't aware they look better when they aren't frozen in pose until they actually see the result; the appreciation afterward makes the risk worthwhile.
----
Credits:
http://commons.wikimedia.org/wiki/File:LookingSouthRim.JPG
http://commons.wikimedia.org/wiki/File:Half_Dome_at_sunset.jpg

Throwing curves

2009-08-27T09:36:00.000-07:00

It's always a pleasure to help editors get started with image restoration. Staxringold was looking for baseball-related images and asked for advice regarding copyright and technical issues.

Plate photography is one technological generation advanced from daguerrotype: glass plate photography does use a negative (which daguerrotype doesn't) but the photographic emulsion adheres to glass.

Ever have trouble getting stuff to stick to glass? Same problem here. Those two white spots top and center are places where the emulsion peeled away from the plate. Same with the larger white spots at upper right. At full resolution the image is also marred with hairline scratches. Glass plate negatives are very prone to damage; these are among the reasons film photography supplanted glass plates for most uses during the early twentieth century.

So it's rare that I give the thumbs-up to glass plate work as a first project. Staxringold wanted to do a baseball picture, this is Walter Johnson shaking the hand of Calvin Coolidge, and this image was less badly damaged than most glass plates.

The hardest decision about it was how to crop. The photographer in me wants to take three steps to the right in order to get that foreground barrier out of the way. That isn't possible without a time machine. So either we'll have to accept a major compositional problem or lose data. The issue here is which data is expendable. Staxringold's first edit is at left: he squared and centered the composition by cropping out the woman who wears a white dress.

That wasn't satisfactory so he tried a couple of other ideas including a portrait orientation crop at right. The difficulty with this is that it enhances the distracting presence of a man in full profile who stands behind Walter Johnson. It would be worse to cut off part of Mr. Profile and in order to lose him entirely we would also have to crop out part of Walter Johnson. Plus that darn barricade takes up a third of the image. So although the usual crop for a handshake photo would be vertical, that composition isn't successful here. I suggested a different approach.

We lose the lower half of Walter Johnson with a panorama crop, but this regains the sense of crowd scene including applause from the spectators at far right. The man in profile behind Johnson's left shoulder blends into the scene. This crop still has the handshake and sheds nearly all of the barrier. The hard part is he'll still have to patch and fix all the emulsion damage. Well, no compromise is perfect. Staxringold went away happy and was so enthusiastic that he soon nominated the following version for featured picture.

Staxringold is a great sport about these things. What caught my eye at first about the patching was the firm horizontal line. At far right the heads don't match the shoulders. Started out thinking we could do a little better, then (inevitably) worked on hundreds of overlooked dirt specks and hairline scratches. It would take a close view at 200% resolution on the full version to find those flaws, but they were there. Shoemaker's Holiday and I both helped clean them out.

The final step is to manage the histogram. Couldn't blog about the portrait of a great pitcher without making a pun, so the essential part of today's discussion is the Photoshop curves tool. It's a more sophisticated variant on levels: curves allows an editor finer control over brightness contrasts.

It almost feels like Coolidge and Johnson have come into sharper focus. Midtones dominate both of their clothes; notice how much sharper Johnson's pinstripes appear. The curves tool allows an editor to enhance contrast within particular ranges of brightness. That's why those pinstripes and Coolidge's lapel look more crisp. Now these two principal figures stand out better from the crowd, which is mostly darker toned. Viewers respond intuitively to subtle changes that convey information about what is important in a scene.

Mostly this is a moderate curves adjustment. The white point is moved normally to the brightest point where the original actually had data. The one unusual element about this particular adjustment is that I reduced the contrast slightly in the brightest part of the range. This accomplished two things: it helped prevent that woman in the white dress from becoming a dominant element and it reduced the shine on Johnson's forehead.

And that, in terms of image editing, is how to throw a curve.

A day in the life of a historic image editor

2009-08-26T13:48:00.000-07:00

0:00 Discover high resolution digitized file of a royal Spanish grant of arms.
0:10 Select page for download and restoration.
0:12 Brew coffee.
0:15 Begin restoration.
10:15 Finish restoration. Whew! Feels great.
10:45 Complete notated uploads to Wikimedia Commons, insert at relevant articles with captioning.
11:00 Nominate for featured picture candidacy.
11:17 User:I Love Photography leaves long comment, obviously without having looked at the original file or reading the upload notes or clicking on the source link to the original file. Wonders whether there was text with the original image (there was; nineteen pages of it). Insists that without text the illustration has no encyclopedic value at all except at a nonexistent biography. Never mind its use at the Law of Arms or at Spanish heraldry.
11:18 - 12:00 Mope.
12:15 Upload remaining 18 manuscript pages.
12:29 User:I Love Photography Two demands full translation.
12:30 Brew stronger pot of coffee.
19:45 Complete full translation of calligraphy manuscript from sixteenth century Spanish.
20:00 Add translation to all relevant file hosting pages.
20:57 User:I Love Photography Three opposes because only one of the nineteen pages is used in article space.
21:00 Brew stronger pot of coffee.
21:05 User:I Love Photography Four nominates an image of a dung beetle.
23:00 I Love Photography One, Two, and Three have all supported the dung beetle.
23:30 Talk page conversation:
----
Historic Image Editor: "Why the difference?"
I Love Photography: "We can't promote too many illustrations from manuscript royal grants."
Historic Image Editor: "But we don't have any FPs of royal grants at all."
I Love Photography Two: "This is preventative."
Historic Image Editor: "How many featured pictures of dung beetles do we have?"
I Love Photography Three: "Twenty-seven. But this one is rolling an unusually large dung ball."
----
23:58 Cry self to sleep.

Crafty solutions

2009-08-21T12:26:00.000-07:00

One of the copyright issues Wikimedians seldom think about is the creative copyright of original craft art designs. It's worth considering: in many cases this is the work of artists who operate microbusinesses. The creative design is their livelihood.

This is one reason why Wikipedians who have creative hobbies should be photographing and uploading their work. If you make your own designs then you own the copyright on those designs and you can place them under free license.

Don't be shy. It may seem like vanity at first, but it's actually a responsible way to address the economic pressures that globalization has been having on craft artists in first world countries. While raising the Navajo rug article to GA, one of the things I researched was the impact of imported Asian imitations on the indigenous weaving industry. The cost of living in Thailand or China is so much lower than in North America that the 150 year tradition of handwoven Navajo blankets is in decline.

Traditional Navajo designs are in the public domain, but most craft artists create their own patterns. These are as much their financial lifeblood as a writer's words are the writer's livelihood. Yet there are Wikimedians who wouldn't dream of plagiarizing text who violate copyright on craft work. Usually they do so accidentally; they hadn't thought about this.

And after all, sometimes the crafty hobbyist has an idea that is just plain fun. The number 12 on the clock below is a fossilized tooth from a mosasaur.

The face of history

2009-08-10T20:05:00.000-07:00

John Quincy Adams was the youngest of the statesmen to emerge from the American Revolution. He was just eleven years old when he first accompanied his father on a diplomatic mission to France. At age fourteen he went to Russia, without his father, to assist another mission that sought diplomatic recognition for the United States. Seven decades later this portrait was drawn in pencil during his final hours of life.

After a long career as a diplomat, as Secretary of State under James Monroe, and a single term as president, John Quincy Adams spent seventeen years in the House of Representatives as a congressman from Massachusetts. His last public act was to vote against a bill to honor veterans of the Mexican-American War. After uttering a loud "No!" he collapsed from a stroke. Adams was too weak to be moved from the building; he lingered for two days.

The librarian's notes on the mounting for this image read, "The original sketch of Mr Adams, taken when dying by AJS. In the Rotunda of the Capitol at Washington".
This is not a well known image, probably because the staining and fade are so extensive that the face is barely discernible. Depending on several factors stains in images could be difficult or easy to fix. A medium sized stain in a clear blue sky is easy to correct. Stains that interact with image features are more challenging. This example is particularly difficult: dozens of small stains intersect the subject's facial features. A single careless edit could transform a placid expression into a scowl.

The restoration was several days' work interrupted by about a month's hiatus. Normally I wouldn't attempt work on an image this badly damaged. It isn't great art or even technically correct. The depiction of the pillow is primitive and Adams's neck ends without a shoulder to support it. But there was something about the historic context that made this exceptional. The eyes kept drawing me in: there in 1848, as those lids shut, the living memory of the revolution faded from public life.
It feels like an honor to find those eyes again 161 years later. I wonder why there aren't more people looking for these things.

Speed restoration

2009-07-29T17:51:00.000-07:00

Sometimes preparation doesn't help. WikiVoices was set to a podcast today, and four hours before showtime I learned that the restoration that had been ready for half a week actually wasn't usable. We were planning an article creation episode where several editors collaborate via voice and text to start a new article and get it to DYK. Then we discovered that there actually weren't enough online third party sources for the Aesop fable we had planned to write about.

Quick, time to come up with Plan B.

The Library of Congress has a few dozen vintage childrens' books scanned and available online from cover to cover. There has actually been a fair amount of scholarship about Mother Goose, so Shoemaker's Holiday and I tore through the possibilities. Our conversation sounded something like this:

Me: "There's got to be an article on Humpty Dumpty."
Shoemaker: "There already is."
Me: "'Mistress Mary, quite contrary'"?
Shoemaker: "It's there under a slightly different title."
Me: "'Bye, Baby Bunting'"?
Shoemaker: "Already an article."

Nestled among the selections were also several well known rhymes that had never gotten their own Wikipedia articles. Shoemaker insisted upon "The Queen of Hearts" because Lewis Carroll quotes it in Alice in Wonderland. I had been pushing for "A dillar, a dollar, a ten o'clock scholar" but gave in to his good argument that there would be more sources with the Carroll reference. So he went to look for references and I opened Photoshop.

The advantage to having a good illustration for a new article is that it generally helps a nomination get a good position at Wikipedia's main page in the "Did you know?" section. That makes a good morale boost for a podcast where some of the editors had never written a DYK before. The illustrator William Wallace Denslow had created three images to accompany the poem. They were all in relatively good condition, considering they were more than a century old. Denslow was a fairly important children's book illustrator: he also provided the artwork for the first edition of The Wizard of Oz.
First I selected the smallest of the three images. The Knave of Hearts was the only one of the set that did not occupy a full page. This would be the simplest and quickest to complete, which was an important consideration in case any other unexpected problems arose before the recording.
Although the range of colors is not very broad, at close examination the image has a number of small dirt specks and scratches, as well as fade. The tool to use in this situation is the clone stamp, since the healing tool can result in a mottled appearance when used on this type of printing.
After fixing the scratches and dirt this restoration required a linear dodge and a localizedcolor adjustment to correct for the uneven lighting at far right where the page bent near the spine, plus a few other changes to simulate the colors and tonal range from the original printing. It didn't take long and by the end of the recording all three illustrations were restored.