Thursday, February 26, 2009

Wikipedia in academic studies

There are always interesting new things to discover at Wikipedia.  Found this page last night. Definitely worth watchlisting, probably also worth bookmarking.

Wednesday, February 25, 2009

Requests for arbitration

Currently there are two requests for arbitration at English Wikipedia.  This post pertains to both of them, and also to that process generally. When contemplating opening a request for arbitration there are really three questions that need to be asked:
  • Is this an urgent problem that cannot be resolved by anything else?
  • Is this a non-urgent problem that cannot be resolved by anything else?
  • Is it clear to uninvolved people that it cannot resolved by anything else?

The times to request arbitration are where the first condition has been met, or else when both of the other conditions have been met.  One big cause of strife and drama happens when the third condition hasn't been satisfied.

This is a good metric for determining when other formal dispute resolution is appropriate.  Non-arbitration dispute resolution serves two purposes:

  • To resolve a dispute.
  • To provide a fair (if slim) chance at resolving a dispute, while demonstrating to uninvolved people that the appropriate attempts have been made.

It's important to remember what constitutes dispute resolution.  Administrative noticeboards are not dispute resolution.  Other than arbitration, formal dispute resolution comes in six flavors.

A while ago the dispute resolution navigation box used to list these clearly.  It no longer does, which may explain why requests for arbitration have been getting filed recently that cite nothing other than article talk discussions.  Those filings get rejected but waste everyone's time.  On other occasions, people try to list administrative board threads as if they were prior dispute resolution.  

It's a requirement when filing an arbitration to list prior steps at dispute resolution.  It's been my longstanding opinion that non-formal dispute resolution should be removed from that list on any request for arbitration.  The one exception is when the arbitration enforcement board has failed repeatedly, which indicates that a prior arbitration decision was unsuccessful.

If a situation is not urgent, usually two or three attempts at formal dispute resolution should be tried before filing a request for arbitration.  That's enough to demonstrate good faith efforts to resolve the problem.  

A key mistake that many editors make is to fail to open enough formal dispute resolution because they don't think it would work.  They might be right about that, but there's no way for the larger community to see that until it's tried.  

A lot of difficult RFAR discussions occur when the filer is technically right, but hasn't taken enough formal steps to demonstrate that to the community at large. Even if the case opens it could place all of its named parties at the center of a storm for one to three months afterward.  Usually it's better to open another formal dispute resolution process--even if it's mostly to clear the air and make the eventual arbitration more straightforward.  Optimism is worthwhile too with other dispute resolution; pleasant surprises have been known to happen.

Tuesday, February 24, 2009

Wikimedia Commons Picture of the Year


In case you haven't done it already, you can still help choose the Wikimedia Commons picture of the year. Round 1 remains open until February 26. Then the finalists get chosen for the second round.

Due to coding issues, it's possible to see how many votes each candidate is getting. So far the leader appears to be the spectacular fire breathing photo above by Luc Viatour, a volunteer from Belgium. Nearly 500 other featured images are also in the running.

With material like this to choose from, voting is a pleasure.

Monday, February 23, 2009

Congratulations to Xavexgoem

A while back I blogged about an ongoing restoration for the Ottoman surrender of Jerusalem in 1917. Good news: it's completed now and Xavexgoem has his first featured picture credit.

Here's the unanimous candidacy. It's great to see his work come to fruition. And it's a highly encyclopedic subject, well photographed, that counters systemic bias. Wikipedia could use more good material about the Middle Eastern Theater of World War I.

Xav's busy with another restoration now. Here's looking forward to praising his next success soon. Three cheers for a job well done.

Sunday, February 22, 2009

Devious histograms

Jake Wartenberg is working on a wonderful chromolithograph of the Montana state capitol that's nearly done. He sent me a copy for review with a couple of challenges he was facing. And he was also generous enough to let me blog about it. So today we'll discuss more about histograms.

Changing the levels is just about the last step in restoration. It can be very useful to preview a levels adjustment in order to detect subtle problems that need to be addressed, but to do the actual fixes it's important to step back to a pre-levels version. It's a good practice to always save a version from immediately before the levels adjustment, and to keep that pre-levels version under a separate filename. Jake's got good habits and he's making more changes to his pre-levels version right now.

This is very good work and it's nearly complete. The bottom border needs cropping and an information tag that someone pasted into the lower margin needs to be clone stamped out. The data on that tag is useful--it ought to be transcribed for the image hosting page when the work gets uploaded--but it isn't part of the original work. Since histograms are dumb, that tag translates into bum data as far as the histogram is concerned. In turn, that has an effect on what the software wants to do with brightness and color adjustments throughout the image.

Another element to remember with older paper prints is that paper often acquires uneven brightness as it ages. Edges tend to dry out and darken more than the center of the paper. Notice how in this example it's the corners that are darkest of all, while some of the margin in the middle (top and bottom) is closer to white. So Jake's also going over his pre-levels image to create feathered adjustment layers that will even out the brightness and color balance.

While we were discussing this he asked me why that matters. How much difference does it make, really? Here's a demonstration from one of last year's restorations where the difference stands out.

Here's the unrestored version of a Robert Fulton submarine design from 1806.

And here's the final featured version. It took quite a bit of work to get to this point. The hardest part to fix was Fulton's description, because he had rubbed out the original description and written a second one in smaller lettering in its place. In order to get an even tone through that section I actually went in at 700% resolution, working with tool settings 2 and 3 pixels wide along the outside edges of each pen stroke. It took a lot of work, but the end result is a natural paper tone.

The original rubbed-out lettering is easier to see on a preview of the unrestored version.

But obviously, trying to work from this version isn't going to yield the best end results.

Histogram previews can be a wonderful way to glimpse the potential of an unrestored image. A histogram preview gives a hint of how far the restoration might go, and reveals challenges that might be difficult to detect otherwise. But whenever possible, do the other restoration work before a final histogram fix. The end result comes out better that way.

Friday, February 20, 2009

Failure to thumbnail

This is a featured picture on English Wikipedia. But you can't see it because the Wikimedia Foundation software has a problem. The problem is a longstanding one because it's been replaced at the article about sumo wrestling with a non-featured image. And that neglect does not demonstrate respect for the volunteer Victor Rocha who did the restoration work.

In order to see this featured picture, it's necessary to go to the hosting page and click on the link where the thumbnail ought to be. So I did a check on how many people are actually viewing this page. Turns out that for the month of January 2009, the hosting page at English Wikipedia received only 50 page views, and the corresponding hosting page at Wikimedia Commons received only 56 views. So at most--if every one of those views represents a a unique hit and every one of those visitors clicked through to the full version--106 people set eyes on one of the site's best images. And we can be pretty darn sure the actual number was lower.

It's time to prioritize the proper display of images. I'm not certain what bug caused the sumo failure, since it's a JPEG image and JPEGs normally display, but WMF software consistently fails to display PNG images larger than 12.5 MB. What you get instead is this at right, which ought to look like an eighteenth century engraving of American Revolutionary War sea captain John Paul Jones. He cuts a dashing figure topside on the deck of a warship in battle--or he would if you could see him. Since WMF software doesn't accept TIFF files, we're stuck with PNGs if we want to upload uncompressed formats. This is a serious detriment to utility and to collaborative editing, and editors who live in parts of the world that have slow Internet access are effectively shut off by this bug.

These problems inhibit the use and creation of featured content.

For shame.

Tuesday, February 17, 2009

How standards change

Next month marks the fifth anniversary of Wikipedia's featured picture program. So to commemorate that occasion here's a quick look back at the beginning to see how far things have come. The bridge above was Wikipedia's first featured picture. Here was the discussion:

From Tamar Bridge and Royal Albert Bridge. I uploaded this image so I've put it in self-nominations, but it was taken by a non-wikipedian.
  • Nominated by fabiform | talk 14:55, 1 Mar 2004 (UTC)
  • Triffic!! (take that as a second) - Gaz 16:34, 1 Mar 2004 (UTC)
  • Approve. I've never seen a panorama that inclusive. Bevo 17:43, 1 Mar 2004 (UTC)
  • I heartilly approve, I've never seen anything like that... Sam Spade 05:16, 2 Mar 2004 (UTC)
Two years later, its delisting discussion:

Too distorted, blown highlights.

  • Delist. - KFP (talk | contribs) 23:48, 17 October 2006 (UTC)
  • Delist per nom. howcheng {chat} 22:17, 18 October 2006 (UTC)
  • Delist per nom. --Steven 01:11, 19 October 2006 (UTC)
  • Keep. A pity to delist this picture inspite of blown highlights. I don't think distortion is a problem, it reminds me of a M. C. Escher painting! -- Alvesgaspar 11:18, 19 October 2006 (UTC)
  • Delist I always wondered how that got to be a featured picture in the first place. | AndonicO Talk 18:16, 20 October 2006 (UTC)
  • Delist - how did this get through the first time? Yeuch. —Vanderdeckenξφ 10:45, 22 October 2006 (UTC)
  • Keep the Earth revolves around a sun. Deal with it. drumguy8800 C T 01:11, 24 October 2006 (UTC)
  • What does the earth revolving around the sun have to do with it? HighInBC (Need help? Ask me) 01:12, 24 October 2006 (UTC)
  • Delist Per blown highlights and distortion. HighInBC (Need help? Ask me) 01:12, 24 October 2006 (UTC)`
  • Delist per nominator. -- Moondigger 21:37, 24 October 2006 (UTC)
  • Question Are the bridges actually curved like this in reality? -Arad 22:22, 24 October 2006 (UTC)
Not nearly to this extreme, although the approach to the Brunnel bridge is quite curved.Laïka 22:31, 24 October 2006 (UTC)
  • Delist; confusing. Laïka 22:31, 24 October 2006 (UTC)
  • Delist per nom. Yuck. -- CountdownCrispy ( ? 22:33, 24 October 2006 (UTC)
  • Question wheres the original nom? -- Coasttocoast 23:29, 24 October 2006 (UTC)
Here. --KFP (talk | contribs) 23:33, 24 October 2006 (UTC)
Thanks, lol it seems that every picture there has been delisted :D -- Coasttocoast 00:04, 25 October 2006 (UTC)
Delisted. --KFP (talk | contribs) 23:42, 26 October 2006 (UTC)

Wikipedia saw 23 nominations during the featured picture program's first month, of which 14 were promoted. 11 of those have since been delisted. While reviewing the remaining 3 I noticed one that really wasn't up to current standards and nominated that for delisting also. It's a dust storm in Texas, but at 700 x 459 pixels and 68KB it just doesn't meet the minimum expectations.

For comparison, here's the discussion for one of last month's featured candidates: a shot of the harbor at Sydney, Australia. without its famous bridge:

Sydney Skyline at dusk

Original - Sydney's skyline as viewed south across the harbour from Kirribilli at dusk. The Sydney Opera House is visible on the far left.
Reason
Shows the skyline of Sydney in an encyclopaedic and visually pleasing way. It is currently the lead image in the Sydney article.
Articles this image appears in
Sydney
Creator
User:Diliff
  • Support as nominator --Diliff | (Talk) (Contribs) 21:53, 8 January 2009 (UTC)
  • Comment I have to say that it is a nice image, but it would be good form to wait until the active discussion about the choice of top image has formed a consensus, especially as there has been talk about EV. Your image has only in fact been in the article for 24 hours. (I disclose that the previous image in the article was mine, it as been there for some time. Yours is a far better image photographically, I am not offended by it being replaced - mine was no more than a quick shot taken from a ferry that I thought would be useful to WP so I uploaded it, i am not proud of it at all.) But it feels a bit off to self nominate yours for FP at this point, especially in light of recent discussions about self nomination and the length of time images have actually been in the host article. Mfield (talk) 22:20, 8 January 2009 (UTC)
    • Well, fair enough. There didn't seem to actually be any discussion on the image on the talk page. The person who initially seemed against the image admitted they preferred the replacement once they saw it in the infobox and there was no other discussion since then. I'm happy to mothball the nomination for a while if desired. To be honest though, the EV of the image as a definitive skyline of Sydney in the article is clear. Even if it were decided to keep your image as the lead image, I'm fairly certain it would, as I mentioned on the Sydney talk page, be a good candidate to replace the skyline as viewed from Balmain further down the article. I wouldn't have nominated the image if I didn't think it stood a good chance of sticking there. As we discussed about self-noms and time periods, I thought we basically agreed that while it is generally safer to wait to confirm EV, there is no mandate for it and we should judge each case individually. As I said above, I don't think there could be much argument on the basis of EV. Diliff | (Talk) (Contribs) 22:32, 8 January 2009 (UTC)
  • Comment I don't see the appropriateness of having this picture in the economy section of the Sydney article. For one it is not supported by text, none of the buildings mentioned in the text can be seen in the image and its features are to small. I also think that the previous image would be more suitable for the info box. Perhaps this would be more suited in the tourism section replacing the opera house pic and Just because it is a featured photo (soon) doesn't make it appropriate for that section. The photo that I have added to that section is supported by text and some of the institutions and a lot of the subject matter can clearly be seen in it. Adam (talk) 02:34, 9 January 2009 (UTC)
    • What makes you think the previous image is better in the infobox? Also, I concede that this image doesn't show the logos on the buildings as well as the Balmain image, but it is a far more complete skyline of Sydney. I don't know if I would say that the text of the section relates to the buildings visible in the Balmain image. The only one I can see is Westpac, and the caption makes no mention of the economy. Anyway, I think if you want to discuss it further, we'd best take it to a talk page as this is getting beyond the reaches of the nomination. For disclosure, both Adam J.W.C. and Mfield are the authors of two images that this image has replaced or has been suggested to replace, respectively. Diliff | (Talk) (Contribs) 02:59, 9 January 2009 (UTC)
      • I have already disclosed that if you'd care to have read my original comment, I have no bone to pick with the image being replaced, I have an issue with an image being proposed for FP whilst a discussion is still underway as to whether it should even be included in the only article it is now in. Mfield (talk) 03:30, 9 January 2009 (UTC)
        • I know you did, and I did read your original comment. It was just a summary of disclosure regarding both you and Adam together. No need to get narky - it wasn't an attempt to undermine your disclosure.. As I said though, there actually was no discussion underway at the time I submitted this. The only objection raised on the Sydney talk page was withdrawn. Diliff | (Talk) (Contribs) 13:24, 9 January 2009 (UTC)

Comment I'm supportive of the image, but I tend to agree there should not be ongoing dispute about it among contributors to the article (either here or on the talk page). Hard to assess its EV until we know it is going to be stable in the article. Tentatively I favor it for the infobox rather than the Economy section, as the Balmain image places the emphasis on the financial/corporate landscape; a broad overview of Sydney's skyline is less useful there. Fletcher (talk) 04:07, 9 January 2009 (UTC)

  • To be fair, there actually wasn't a dispute on the article's talk page at the time I submitted this. I raised the issue on the talk page prior to adding it to the article, and one person had initial reservations about it, but when another contributor added it to the article of his own accord, they rescinded their objections and supported its inclusion. For all of the debate here, as far as I can tell Adam J.W.C is the only person who objects to the image in the article. Diliff | (Talk) (Contribs) 13:20, 9 January 2009 (UTC)
  • Weak Oppose While visually it's pretty much perfect for a city nightscape, I'd have like the composition to also include the harbour bridge as in this pano you previously took. --Fir0002 04:08, 9 January 2009 (UTC)
    • Wouldn't it be an almost identical picture to that previous one, then? :-) I actually intended to take the same shot as the previous FP when I was in Sydney, but because of the NYE fireworks, they had begun to scaffold the bridge up and it looked pretty awful so I delibrately excluded the bridge. I can appreciate that you might want a slightly different composition, but given that I cannot simply go back and re-shoot (I've been back in the UK for a month now), this is the image I managed to take. Are you opposing because you wish it were different, or are you opposing because it is really unsuitable to be an FP in its current form? Diliff | (Talk) (Contribs) 11:52, 9 January 2009 (UTC)
      • Well I took it as a HDR upgrade to your existing one :) While I very much appreciate this being very difficult for you to redo, I do think that the bridge should be part of a lead panorama of Sydney as it's such an iconic part of that city (and at any other time of the year including it would have made perfect sense). So for that reason I think your existing FP has greater EV and thus I couldn't support this one nice as it is. --Fir0002 10:35, 10 January 2009 (UTC)
        • I think the existing FP has a different purpose though - this image is suitable for the infobox and the existing FP works well as a panorama at the bottom. Yes, it would have been nice to include the bridge, but if I did, it wouldn't make a good thumbnail for the infobox and therefore would merely have been a candidate to replace the existing FP, rather than complement it. Besides, as Durova mentioned, not every image of New York City needs the Brooklyn Bridge. It is very difficult to get photo with a good composition, with non-panoramic proportions and the Opera House, skyline and the Harbour Bridge all in the frame. As an analogy, you already have a good FP of a kangaroo but you've nominated a new one with different characteristics. Some things about the new one are better (no man made elements in the background, shows feeding) but some things are worse (angle, less aesthetically pleasing composition/pose), but because it describes a slightly different aspect of the same subject, you thought it worth nominating. Same situation here IMO. This image wouldn't be fit for purpose if it included the bridge. Diliff | (Talk) (Contribs) 11:52, 10 January 2009 (UTC)
          • I've never been to Sydney but is it possible to shoot from an angle similar to this? That would be one way to get the bridge and the opera house whilst retaining acceptable dimensions for the lead image. If it's not possible then I'll change my vote to Neutral because I can understand what you're saying with them illustrating different aspects. But just to let you know that other FPC wasn't actually successful so I haven't yet got a kangaroo FP :) --Fir0002 22:12, 10 January 2009 (UTC)
            • Oh really? I was sure that last one went through.. Okay, nevermind, but you got my point at least! As for that image, well it might be possible. Sydney is a massive city with lots of coastline, but quite often it is difficult to get a good shot from public property, as all the good vantage points are obscured by multi-million dollar houses on the harbour. :-) Anyway, you can't actually see the skyline in it, so it may be cut off by other houses, I'm really not sure. It would take a lot of exploring to find the hidden gem shooting locations, I'd say. And as I said to Mfield below, some of them require access to people's balconies! As a mere tourist with limited scouting time, I was more limited to where the ferries happened to take me! Even this shot required a 10 minute walk through residential streets from the nearest ferry stop, or else a 20 minute hike across the Harbour Bridge and then down a few side streets. You sort of have to know what you're doing, to cut a long story short! Diliff | (Talk) (Contribs) 23:55, 10 January 2009 (UTC)
            • On that note, I am waiting until next time to shoot one from here[10] (or better yet of possible from the land in the lower half of the frame) as it gives a nice broad view. Mfield (talk) 22:53, 10 January 2009 (UTC)
              • That actually looks like a good spot to shoot from. Sydney is blessed with so much harbour that there is practically an unlimited number of good locations to shoot from (if you can get access to the shoreline, a decent vantage point or in an ideal world, a friendly resident with a balcony!), although I think that would probably also result in a panoramic composition. I can't quite tell from a map exactly where it was shot from. That little peninsula is either Cremorne Point or Taronga (from what I can see from Google Maps). Actually given the title of the image (given it is on your site, did you take it? If so, why are you waiting until next time to reshoot it?), it must be from Taronga. The ferries go to both locations so it would be worth the trip. I'd be very interested to see what you can shoot from there. Diliff | (Talk) (Contribs) 23:55, 10 January 2009 (UTC)
                • Yes I took it and that one is shot from the zoo, but it's a crop from a two frame handheld pano that I shot a few years back when we jumped on the ferry to go there for the afternoon with some friends and i just had the camera on my shoulder. No tripod and only the one lens. I have meant to get over there at dawn last time i was in Sydney to shoot it properly but had no time. Mfield (talk) 00:11, 11 January 2009 (UTC)
  • Comment I think this image would be appropriate for the lead image in the Tourism in Sydney article instead of the Opera House image and in the same section in the Sydney article. Also I am not sure if this image looks natural especially the colour of the Sky. If I were there at that time would it look exactly like that. I wouldn't mind seeing a version of this image that hasn't been tone blended. Adam (talk) 06:43, 9 January 2009 (UTC)
    • A couple of things.. I agree it could work in the tourism article, but I actually think that if it were to, the Opera House picture should probably be moved to replace your image from Balmain in it too. I'm not saying that to spite you, I'm saying it because it is actually located right in the middle of the Sydney Opera House section! ;-) As for the colour of the sky, the sky can and does turn pinky purply just after sunset. It doesn't happen at every sunset, but it does happen. It has been enhanced slightly with a minor saturation boost and contrast enhancement, but I maintain it does still look natural. Besides, would your image of the Blue Mountains look like this if I were there? Diliff | (Talk) (Contribs) 11:52, 9 January 2009 (UTC)
        • It would look like this if you were there. Also I never objected to the image being in the Sydney article but I did suggest that it would be good for the tourism section, but I admit that that it does look good in the info box as well. As for the tourism article (in Syd) I think all images would be appropriate as they all show Sydney from different perspectives. Adam (talk) 22:02, 9 January 2009 (UTC)
  • Strong support This picture is a great composition and shows a very impressive view of Sydney. Wladyslaw (talk) 08:10, 9 January 2009 (UTC)
  • Support--Mbz1 (talk) 16:17, 9 January 2009 (UTC)
  • Support Good quality, aesthetically pleasing. IMO adding the harbour bridge would probably make the image lose its appeal as seen in thumbnail in the article. Muhammad(talk) 16:49, 9 January 2009 (UTC)
    • I agree. I think it would also make the proportions of the image too extreme for the infobox. Diliff | (Talk) (Contribs) 17:29, 9 January 2009 (UTC)
  • Support excellent composition. No comment on the harbor bridge dispute--looks fine to a silly Yank from the other side of the world. Not every panorama of NYC needs the Brooklyn Bridge and/or Statue of Liberty. The possibility of different vistas adds character to a city. DurovaCharge! 17:49, 9 January 2009 (UTC)
  • Support I said above I was supportive of the image so I should give an official !vote. I think it is definitely a great image for the article, and whatever its final placement, I agree no one seems to be trying to get rid of it. Fletcher (talk) 00:26, 10 January 2009 (UTC)
  • Support Though I wonder if there isn't a larger resolution available. The colors in the sky are very impressive! --Massimo Catarinella (talk) 01:01, 12 January 2009 (UTC)
  • Support - really good.--Avala (talk) 20:41, 12 January 2009 (UTC)

Promoted File:Sydney skyline at dusk - Dec 2008.jpg --Wronkiew (talk) 04:25, 17 January 2009 (UTC)

I'm on the fence about one of those other first month promotions--a map of the Philippines--but the third remaining example of our original featured pictures holds up pretty well. It's since been featured at Commons and at the German Wikipedia: an overhead view of the USS Iowa firing its guns.

Anchors aweigh.

Sunday, February 15, 2009

Wikipedia Art and media restoration

Wikipedia had one of its more interesting deletion discussions overnight.  A page called Wikipedia Art lasted about a day. By site standards the deletion was mundane, but the editors who created it were not. There's an untapped opportunity here and I'm reaching out to them. The artists Scott Kildall, Nathaniel Stern, and Brian Sherwin were active in it, apparently with a measure of support from the academic art world.

There are untapped synergies between Wikipedia and professional artists. One of them is illustrated here: a portrait of actor Mark Harmon by professional photographer Jerry Avenaim. This photo is scheduled to run on Wikipedia's main page tomorrow.

Mr. Avenaim himself didn't nominate the portrait for featured picture. Another volunteer noticed its high quality and put it up as a candidate where it nearly failed the minimum resolution requirements until I noticed the photographer was already an active Wikipedian and contacted him. He was surprised and delighted to learn his work was under consideration, and supplied a larger version.

Now here's good news for Jerry Avenaim: as Picture of the Day for February 17, 2009 the portrait will receive an estimated 6 to 7 million page views as a main feature, plus about 30,000 direct views to the image hosting page. That's more attention than his work would get from a day on the front page of The New York Times. Thank you, Jerry Avenaim, for doing well by doing good. Here's a link to his blog.

I would love to establish contact with the Wikipedia Art participants and help them direct their considerable talents into productive endeavors. Posted to the Village Pump discussion about this. Let's hope it yields fruitful results.

Saturday, February 14, 2009

Kudos to Jake Wartenberg

It's been wonderful collaborating with one of our newest restorationists, Jake Wartenberg. He's shown exceptional enthusiasm and versatility. The image above is one of his restorations, a photochrom of the quays at Waterford, Ireland during the 1890s. One of our Irish Wikipedians, Alison, helped select the image so now, in addition to being a featured picture candidate at English Wikipedia, it's also a featured picture candidate at the Irish Gaelic Wikipedia. Alison even sounded excited to try a restoration herself.

But more about Jake's fine work. His first restoration was a panorama of Jerusalem from the early twentieth century. Now he's working on a Japanese woodcut and an early World War II Kodachrome of the liberation of Paris.

One thing Jake could use help with, though, is translation from Japanese. He's mostly done with a print about the Russo-Japanese War, a comical depiction of the Tsar having a nightmare about his wounded forces returning. The original is available at the Library of Congress here. It would be very useful to have a full translation of the caption.

Friday, February 13, 2009

Update on Wounded Knee


The Library of Congress has confirmed my discovery of four human remains in a photograph of the Wounded Knee Massacre aftermath. An excerpt from their reply follows:

Upon viewing the high-res TIFF file we made of the file, the human remains are quite visible, indeed.

Thank you very much for contacting us regarding this image, and for your interest in our collections. You can imagine that among a collection of 14 million items here, there are a lot of secrets waiting to be uncovered!

So there really is something to be said for scanning historic images into high resolution files, and digitally restoring them.

Am writing to historians to communicate the discovery. Will update as new developments occur.

Many thanks to the Library of Congress and the Wikimedia Foundation for making this possible.

Tuesday, February 10, 2009

Saturday, February 07, 2009

Serendipity

The request for help on the Tel Aviv construction site has turned up a few nibbles and one wonderful bite. What you're looking at is a new restoration of Dizengoff Square, available in full resolution here. Thanks go to Ynhockey for the assistance.

A key element of restoring historic photography for encyclopedic purposes is the ability to recognize the significance and historic nature of a scene. I've never been to Israel and fortunately for Wikipedia, Israel's copyright law is relatively generous. So the request for input on the Tel Aviv construction site prompted him to contact me. He didn't have an answer on the original question, but we went through the Library of Congress archives together and he identified an important location at the heart of Tel Aviv that has undergone significant changes since the photograph was taken 65 years ago. Without his help I wouldn't have understood its significance since I've never been to Israel.

So now, if all goes well, Wikipedia might gain its first featured picture about Israeli history. That's important both for encyclopedic purposes and for balance. Last year, working with several Palestinian editors, I helped raise the article about Palestinian costumes to good article status and restored a few photographs about Palestinian culture. In collaboration with another Israeli editor, Jaakobou, I had helped a few new articles about Israeli culture reach Wikipedia's main page at the 'Did you know?' section, but hadn't contributed to any featured content about Israel yet. With the current situation it's important to be evenhanded.

Shalom.
Salaam.
Peace.

Friday, February 06, 2009

A milestone

Last night I reached 200 featured credits. Three featured picture candidates at English Wikipedia got promoted, bringing the total to 201.

Number two hundred was the Wounded Knee aftermath. Currently I am corresponding with the staff at the Library of Congress about the find: restoration revealed four human bodies in the foreground which had been not been mentioned in bibliographic notes and were nearly invisible in the faded original print. Am looking to make contact with a historian who specializes in the area.

Wednesday, February 04, 2009

How to cheat a crop


Part of the art of restoration is to generate data when it isn't there but you know it ought to be. A lot of historic images have damage at the edges. This one is a glass plate negative, which was a popular technology from the mid-nineteenth through early twentieth centuries. Glass colloidon plates were an improvement over daguerrotypes because getting a negative makes it possible to create high quality reproductions; daguerrotypes couldn't do that. The trouble was glass plates are very fragile. They tended to collect scratches and often the emulsion would simply peel away. That's what happened at the top of the photograph here.

It's tempting to just crop that damage out of frame; someone even tried that. Here's their result.

This gets rid of the damage, but it leaves the portrait with an odd composition. The subject's head wasn't intended to be so close to the top of the frame. That gives it an oppressive feel, and so much of the tree in the background is lost by the crop that the leaves no longer stretch continuously from right to left across the image. That loss enhances a compositional flaw in the original: a branch appears to be growing out of her head. The more context is present in terms of how much of that tree appears in the background, the less distracting that branch will appear. Ideally I'd like to grab the camera, take two paces to the left and reshoot. But I don't own a time machine and this photo is about a hundred years old.

The next best solution is to cheat. Yes, cheat. I cropped the image a little bit higher than the actual remaining data allowed, then patched in a small blank area at extreme upper right.

Cheating isn't something to try recklessly, of course. Here it was reasonably certain that the only missing data were a few leaves. There's no unique historical value to those leaves and it was already clear from the unrestored image that a higher crop had been the original artistic intent. I wasn't going to be so bold as to cheat data all the way to the very top of the original crop, but a small careful cheat improved the composition.

The key to cheating a crop is to do it only on small areas that are obvious to intuit and where no essential data is likely to have existed: blue skies, dirt, or in this case tree leaves.

Tuesday, February 03, 2009

Something's wrong with this redirect

WP:DRAMA redirects to Wikipedia:Administrators' noticeboard/Incidents.

Either we take administration seriously or we don't. Sure, ANI gets misused. But it's inappropriate to validate that disrespect with this sort of joke.

Sunday, February 01, 2009

Levels of challenge

It's always a pleasure to see people react with delight when they first encounter a restored image of their own culture. Yet it's rare to find a feature-worthy image of my own local area. Although San Diego is the oldest city in California, it was far from the gold rush. So the Library of Congress collection has far more historic images of San Francisco and not much quality material about the southern end of the state. Yet last night a high resolution photochrom turned up of our most famous local landmarks: the Hotel del Coronado. Ooh, this was exciting.

The bibliographic notes described this as a color slide. Well, that's right and it's wrong. The image itself is a photochrom c. 1900. Photochrom is a hybrid between black and white photography and chromolithography. Basically, before real color films were invented this was a type of colorization, nineteenth century style. Several decades later the damaged print was photographed for archival purposes on Kodak slide film. So although this is a color slide in one sense, in another sense it isn't.

When reviewing an archival image it's good to know what type of image it is, and also very important to assess the challenges to restoration. This one had two levels of challenge. One was more obvious than the other.

This is easy to spot in the original: the original panorama was photographed in two parts, then later the print got creased from careless folding. So up and down through the center of this image there are two white lines: one straight, one jagged. The image on the right doesn't quite match up with the one on the left either: a few pixels are missing. So it's necessary to go up and down patching in missing data. That's easy to do on the sky at top and the lawn at bottom, but to fill in the architectural details takes work. The more boundaries of any sort that a crease or a seam or a scratch crosses, the harder it will be to repair. This one crosses many boundaries in an important part of the image.

Basically what I do is first move the right portion into proper alignment, then take the healing brush and clone stamp to fill in data from surrounding material. A way to repair a damaged window on the print is to borrow data from a similar window; to fill in a missing portion of the eave, borrow from another part of the same eave. Then blend in so the stretch looks natural. By starting with the easiest parts and moving in to the complex areas, eventually the entire seam disappears.

It's not exactly easy, except in a relative sense. There's a tougher problem on this image that's less apparent at first glance.

Here's a demonstration of the problem. It's the sky. The trouble is easier to recognize after a preview selection and auto levels on the unrestored version. We don't do the restoration work on a levels preview, of course. It's just a good way to see whether something subtle would become a bigger problem than it seemed at first glance.

The natural first step in restoring the sky is to remove dirt and other marks. That's normal for any historic restoration. This is an unusual sky, though, and it has three additional problems.
  1. Extreme brightness differences from the center to the left and right edges.
  2. Uneven color: the left side has retained some cyan, but the right looks more magenta.
  3. Excessive contrast at the bright center.
It's normal for paper to dry out and darken at the edges. That makes sense if we remember that even though the digitization comes from a color slide, that slide must have been a photograph of an older print. If the original were shot and printed around 1900, then it would have waited four decades or more before technology developed to create the archival slide copy. So we have an explanation for the edge fade. The solution is relatively simple: patch and blend.

The bright area at center is harder to correct, though. And so is the uneven color loss and uneven contrast. Without getting too much into details, the solution involved a series of adjustment layers. The end result isn't quite natural blue sky, but then photochrom isn't a natural color process. It's colorization. So a slightly artificial appearance is expected.

The important thing about this restoration is to recognize that there can be multiple challenges: obvious problems sometimes distract from non-obvious ones. It could be discouraging to a novice restorationist to finish the seam at the center and then get mired in the tougher challenges of the sky. If you're starting out at this sort of thing, get to know a few people who already do it. It's better to start off on material that's really within your skill set or to get a helping hand with the tough parts, than to put days into an image for sentimental reasons and then push it aside because you've gotten stuck.