Wikipedia:Featured picture candidates/William Grant Still

Voting period is over. Please don't add any new votes. Voting period ends on 22 Jan 2016 at 18:22:25 (UTC)

Reason: A fine photograph, by a notable photographer, of an important African-American classical composer
Articles in which this image appears: William Grant Still+8
FP category for this image: Wikipedia:Featured pictures/People/Entertainment
Creator: Carl Van Vechten, restored by Adam Cuerden

Support as nominator – Adam Cuerden ^(talk) 18:22, 12 January 2016 (UTC)[reply]
Comment I prefer the original monochrome version. It doesn't look odd to me and it seems like it is a matter of personal preference where the original should be favoured per WP:VERIFY - Wolftick (talk) 19:22, 12 January 2016 (UTC)[reply]

The original is a very light scan from a source (the LoC) known to have poor colour fidelity and contrast. Further, it's a monochrome file, meaning that the subtle shades even black and white film has are lost. Adam Cuerden ^(talk) 19:30, 12 January 2016 (UTC)[reply]

I am not opposed to sympathetic restoration of source where it is clearly lacking, but I cannot support going from this: [1] to this: [2]. That is just adding yellow to a perfectly acceptable greyscale image that is most likely and to the extent that it can be verified a fairly accurate depiction of the original. - Wolftick (talk) 20:06, 12 January 2016 (UTC)[reply]

That may be, but unless you have a verifiable source in this specific case the monochrome version should be favoured per WP:VERIFY: "Even if you're sure something is true, it must be verifiable before you can add it.". The LoC would generally be considered a reliable source - Wolftick (talk) 20:28, 12 January 2016 (UTC)[reply]

The file is explicitly a monochrome file. It ain't a source for nothin' colour-balance-wise; It literally can't be. The information's been stripped. Adam Cuerden ^(talk) 20:32, 12 January 2016 (UTC)[reply]

Again, this may be true, but I don't think this means you can just add it back in without any additional source. - Wolftick (talk) 20:37, 12 January 2016 (UTC)[reply]

...Show me one non-monochrome file of Van Vechten's that isn't slightly yellow, and you have a point. Otherwise, you're basically stating this should be considered an exception. Here's the complete set! [3] Adam Cuerden ^(talk) 20:40, 12 January 2016 (UTC)[reply]

My point is not that you are incorrect. My point is that you do not have a source for the changes specific to this image other than your own understanding of what the image should look like according to other, different images. My reading is that adding colour where there is none is different to "colour/exposure correction" per WP:FP? #8 and so this falls foul of WP:V and WP:OR irrespective. - Wolftick (talk) 21:54, 12 January 2016 (UTC)[reply]

...I think... that's... I don't even know what to say to that. I can't follow your logic here, where you seem to think that an image known to have no information as to warmness or coolness of the greys - it was reduced to greyscale after all, which cannot store tone, greyscale has no information but lightness and darkness of each pixel, there's no tone information - is more correct of a guide than images that actually contain the information, by the same photographer. Adam Cuerden ^(talk) 22:07, 12 January 2016 (UTC)[reply]

It's fairly simple. I'm saying there is no suitable guide that doesn't require original research and lack verifiability. The original, while lacking colour information, is at least from a reliable source - Wolftick (talk) 22:12, 12 January 2016 (UTC)[reply]

So... you're saying almost certainly wrong is preferable to... never mind, I'm done here. Adam Cuerden ^(talk) 22:53, 12 January 2016 (UTC)[reply]

It is not unthinkable for what we'd consider to be "reliable" sources to be wrong. Both scans of Tobias and the Angel are unaltered from the National Gallery of Art (albeit each scan is from a different part of their website), and that's at a stalemate now: we know that one of the two is more accurate than the other, but we don't know which. Then there's the Yorck project scans, which were originally taken at face value as reliable but are now so infamously inaccurate that they have been widely replaced. If sources are known to have issues, even if they are ostensibly accurate in some cases, we have the right to exercise editorial discretion. — Chris Woodrich (talk) 05:04, 14 January 2016 (UTC)[reply]

Using editorial discretion to weight conflicting sources per Tobias and the Angel is very much part of Wikipedia's core content policy. However the source image in question is not "wrong". It is instead lacking information, namely colour. As far as I know there is no additional colour source for this image, so any addition of colour is against WP:VERIFY: "Even if you're sure something is true, it must be verifiable before you can add it".
...However, I've since looked at WP:GL/PHOTO... and, well, I'm out. - Wolftick (talk) 13:26, 14 January 2016 (UTC)[reply]

Support – Yann (talk) 20:11, 12 January 2016 (UTC)[reply]
Support - Surprised that Vechten chose such a close crop. — Chris Woodrich (talk) 23:38, 12 January 2016 (UTC)[reply]
Support – In general I agree with Wolftick's point. In this particular case, the color addition is subtle, not excessive, it is based on similar images by the same photographer which is a guide or template, even though it is not a reliable source. In this particular case I don't see any harm. Bammesk (talk) 17:37, 16 January 2016 (UTC)[reply]
Support – Jobas (talk) 14:28, 19 January 2016 (UTC)[reply]

Promoted File:William Grant Still by Carl Van Vechten.jpg --Armbrust ^{The Homunculus} 00:18, 23 January 2016 (UTC)[reply]