Wikipedia:Featured picture candidates/The Fall of the Rebel Angels

The Fall of the Rebel Angels

Voting period is over. Please don't add any new votes. Voting period ends on 28 Jul 2014 at 17:10:52 (UTC)

Reason: One of those astonishing paintings, somewhat Boschian (although the contrast and lighting here look better, I hope Google Art didn't mess something again).
Articles in which this image appears: The Fall of the Rebel Angels (Bruegel), List of paintings by Pieter Bruegel the Elder, War in Heaven
FP category for this image: Artwork/Paintings
Creator: Pieter Bruegel the Elder

Support as nominator – Brandmeister^talk 17:10, 18 July 2014 (UTC)[reply]
Support - Coat of Many Colours (talk) 17:58, 18 July 2014 (UTC)[reply]
~~Support~~. Not, per Crisco. Hafspajen (talk) 18:02, 18 July 2014 (UTC)[reply]
~~Support~~ - Although I wonder if this is not overly dark. — Crisco 1492 (talk) 01:17, 19 July 2014 (UTC)[reply]

It looks much better viewed 100% (it's amazingly detailed). But I have to admit it does lack presence. The category image Commons:File:Pieter Bruegel I-Fall of rebel Angels (merge).jpg is very attractive, but is surely too blue and lacks resolution. This thing is 450 years old. I doubt it looks as fresh as that. But I've never seen it. Coat of Many Colours (talk) 04:45, 19 July 2014 (UTC)[reply]
Confirmed as too dark. The 1280px thumbnail has >10% of the histogram completely unused. Samsara (FA • FP) 14:25, 19 July 2014 (UTC)[reply]

Yes, that's right. The same with the Raphael images recently nominated here at at Commons. The trouble is that darkening is the result of aging, not (say) underexposure in a photograph, and can't be corrected digitally (nor in reality can it indeed be in a photograph as, unlike overexposure, the information was never there in the first place). So I would say in a Featured Picture we shouldn't be adjusting, unless we have good reason to suppose it's an artefact of the image rather than the photograph itself. But, picking up on a remark by DCoetzee below, I do think there's a case for doing that so as top provide thumbnails and the like. But these really shouldn't, in my opinion, be nominated as "Featured", and I think better reduced in the first place to avoid that. This evening I'll have a go at that to indicate what I have in mind and post back here. As to whether rather dull images such as we have here for nomination should really be "Featured", I do think that's a question to be debated another day perhaps.Coat of Many Colours (talk) 16:46, 19 July 2014 (UTC)[reply]

I've thumbnailed a version lightened in LCH space using the Nikon Capture X2 AutoContrast tool. The idea is to adjust the contrast without affecting the hues or saturation of the colours. Coat of Many Colours (talk) 19:36, 19 July 2014 (UTC)[reply]

The version on the official site of the Belgian Royal Museums of Fine Arts may be correct, it's closer to the Google Art version rather than that currently in the article on painting. Brandmeister^talk 20:06, 19 July 2014 (UTC)[reply]

Yes, indeed. I noticed that too. Incredibly hard to find that website by the way. I wanted this page in connection with the article start for Little Girl in a Blue Armchair I provided, but it drove me absolutely bonkers finding it. I do think the Google image you nominated is worth Featuring. It's unfortunate that thumbnail size it's pretty unprepossessing. Coat of Many Colours (talk) 21:47, 19 July 2014 (UTC)[reply]

So a very light edit, perhaps. — Crisco 1492 (talk) 00:51, 20 July 2014 (UTC)[reply]

No! Inauthentic WP:FP?#8. There's no reason to suppose the Google Art Project image is anything other than an accurate (faithful representation) of the painting as it is today and on the contrary every reason to suppose that it is because it's hard to imagine Google deliberately providing an image misrepresenting the painting in such an unattractive way. There is, as acknowledged by me above and remarked by DCoetzee below in connection with his image, a place for retouching the image in thumbnails and the like, but not here as expressly envisaged by the guidelines. Coat of Many Colours (talk) 03:43, 20 July 2014 (UTC)[reply]

As with many other Google art scans, it is overly dark, which reduces the EV. This is fixable without the "auto-levels" you claim are manifested in every digital attempt at restoration. I don't think any serious editor is stupid enough to just click "auto-levels". — Crisco 1492 (talk) 03:48, 20 July 2014 (UTC)[reply]

Oppose this scan. Way too dark, even if the resolution is great. — Crisco 1492 (talk) 03:48, 20 July 2014 (UTC)[reply]

Well, I can understand that, but I'm curious to know whether you know for a fact that the image is dark, rather than the painting itself, presumably dark from age I would say. Some paintings of course are dark by design. The Potato Eaters or The Angelus amongst recently discussed here. I don't really see why they should be barred from "Featured Pictures" just for that reason. Likewise paintings, such as here in my estimation, which have darkened with age and cannot be restored by digital processing. Coat of Many Colours (talk) 04:24, 20 July 2014 (UTC)[reply]

And regarding your alternative lightening image, you still haven't equalised the histogram and the result is demonstrably less successful than mine.All AutoLevels does is map the palest tone to pure white and the darkest tone to pure black and interploate linearly mid-tones. If you want to further adjust mid-tones as well (gamma) then a separate edit on the curve will accomplish that. In your image the full tonal range is still not represented. However if you were to attempt this (I haven't tried) in three channel RGB space you would almost certainly interfere with colour balances. In the case of the Romano (school of Raphael) image below, the result was an unpleasant yellow colour cast. The appropriate colour space to use is LCH as used by me above. Coat of Many Colours (talk) 05:53, 20 July 2014 (UTC)[reply]

Not Promoted --Armbrust ^{The Homunculus} 17:13, 28 July 2014 (UTC)[reply]