Wikipedia:Featured picture candidates/John William Waterhouse - The Lady of Shalott - Google Art Project.jpg

John William Waterhouse - The Lady of Shalott
Voting period ends on 6 Aug 2014  at 10:05:49 (UTC)
 * Reason:John William Waterhouse's  The Lady of Shalott (1888) was one of Sir Henry Tate's founding gifts to the  Tate Gallery and has always been one of Britain's favourite paintings.
 * Articles in which this image appears:The Lady of Shalott
 * FP category for this image:Featured pictures/Artwork/Paintings
 * Creator:John William Waterhouse


 * Support as nominator – Coat of Many Colours (talk) 10:05, 27 July 2014 (UTC)
 * Oppose much darker than the original.--Mark Miller (talk) 21:09, 27 July 2014 (UTC)
 * It's a Google Art Project work. At 100% it looks fine. Its the image at 100% we should be judging not its thumbnail. If you take the virtual tour in the Google Cultural Institution you can see the painting as hung is naturally a rather sombre one. The thumbnail is actually missing some 20% of the lighter tones in RGB colour space. If you try to adjust it with the naive techniques used in the Romano painting nominated here you get the same sort of hideous result you see there. I've uploaded an Autocontrast edit in LCH space which editors coming after me can use as a thumbnail if they like. Coat of Many Colours (talk) 05:56, 28 July 2014 (UTC)
 * Without being too judgmental, I would rather you stopped accusing people of judging "the thumbnail". I am familiar with this painting. It is way darker than the original and is just not Feature quality to me.--Mark Miller (talk) 06:13, 28 July 2014 (UTC)
 * I'm saying it looks just fine at 100%. The thumbnail is indeed too dark by some 20%. We know that's a problem with some Google Art Project images. I suggested a workround as far as the thumbnail is concerned. I love this painting too. It's one of the most popular at the Tate. I'm glad you're familiar with it. Coat of Many Colours (talk) 07:17, 28 July 2014 (UTC)


 * Oppose- compared with Tate's own photo here, it's darker, has a different colour balance and is slightly cropped, and the stretcher line's not as visible at top. Xanthomelanoussprog (talk) 16:48, 28 July 2014 (UTC)
 * Hello Xanthi. We've already corresponded elsewhere when you wondered whether I was making up my criticism of another image here and it's plain that you're expert in imaging, which makes me wonder why you are commenting here like this. The Tate Gallery image is a scaled down version of its license managed image available for fees typical ranging from £50 to £250 or more depending on how many copies it is proposed to publish. It's a 27.6 Mb file comapred with Google's 6.6 Mb and the first thing that can be said about it is that Wikipedia isn't going to to get to lay its hands on it unless an editor makes a legacy donation of it, unlikely I think. I think it's quite possible it was taken years ago, but the main point is that neither of the two images can be held to be inauthentic in the way I have critised other images appearing in this forum. It's within normal variation for art images.There is nothing in WP:FP?#1 criteria concerning technical standards that precludes it, nor in WP:FP?#8 regarding digital manipulation, the issue in the image we discussed elsewhere. There's no reasonable prospect of a better image appearin soon, and if I may so the feature here isn't its eye-candy wowness, but its utility as a source for students anxious to examine more closely the artist's technique.
 * I have no agenda in promoting this image (maybe there's be a message there, maybe not ). I gather at least one aficionado of this forum are pre-Rapaelite fans, so I present one ... itt van! I shall check its progress with interst when I return (off for a while). Coat of Many Colours (talk) 18:41, 28 July 2014 (UTC)
 * To show Waterhouse's technique (for example his use of heavy textured paint on highlights over a flatter surface) a photo at an angle with raking light is useful. A clearer example is "Bird's Nest" Hunt where his distinctive contribution to watercolour technique isn't shown by flat-on photos. According to Margulis (who devotes a chapter to the subject) LCH or equivalent colour space is the most useful for removing casts or increasing the range of colour in an image- so it may be possible to bring the image closer to the Tate rendition. Xanthomelanoussprog (talk) 20:32, 28 July 2014 (UTC)
 * Yes, well I wouldn't claim to match Dan Margulis's expertise, even read his books, and I don't own Adobe Photoshop or intend ever to. I think it's quite possible there are edits out there which are still superior, but I would say LCH space should be the starting point. The only thing I would say is that the Tate image may not be the reference of choice here. I mean I just don't know, I haven't looked at the painting in more than 20 years. I do recall it's pretty sombre. When I get back I might try to find a forum which discusses these Google images. They're not all wonderful by any means and I don't think articles should blindly link to them. Coat of Many Colours (talk) 21:23, 28 July 2014 (UTC)


 * Oppose - another crappy Google Art rendition. They really should hire someone to do some post-processing. While there may be nothing unauthentic about the rendition, technically a 100% black rectangle is also authentic (image under no light). It would be hugely disappointing to feature such a dark, drab version of the painting. Kaldari (talk) 00:32, 30 July 2014 (UTC)
 * I can't find a source for this image and I don't know enough about their images. It might well be "in-house". It's certainly not the Tate image. Which is more faithful in tone I can't say. Coat of Many Colours (talk) 05:42, 30 July 2014 (UTC)


 * Oppose for now. I'll see if I can get something up. There's a reason why the article on this painting doesn't use the Google scan. — Crisco 1492 (talk) 05:37, 30 July 2014 (UTC)
 * The The Lady of Shalott (painting) image comes from the-athenaeum.org, a blog whose images are uploaded by its members without sourcing. In Wikipedia terms it can't really be regarded as a "reliable source". I shall probably be away when you upload your something, but if it's a derivative work and I see it I shall oppose it. Coat of Many Colours (talk) 05:57, 30 July 2014 (UTC)
 * I'm not adding it here, because I'm not pleased with it. But my oppose on the darkness issue still stands. There are some very interesting color dynamics going on here, but they're lost in the darkness of the scan. — Crisco 1492 (talk) 14:56, 30 July 2014 (UTC)
 * When I get back I'll get in touch with the Tate and with Google about this image. From what I can remember the Google image is more likely to be the more accurate tonally. The painting is really very sombre, as you might expect from its subject (she was after all sailing to her death). Of course it's gorgeously rich in those autumnal hues. Coat of Many Colours (talk) 16:13, 30 July 2014 (UTC)
 * I agree, it's a sombre painting, but that doesn't preclude dynamic colours. The reds and golds of the tapestry vs the whites of her gown, in particular, are a little lacking in the Google scan. — Crisco 1492 (talk) 03:33, 31 July 2014 (UTC)
 * The Google Art Project sheds more light on the Tate. The impression I get is that Google provides its own images of artworks when they do a Virtual Tour of a museum. I can quite readily believe these are often superior to the images a gallery provides. Equally they can't all be successful. It's interesting that when I search on the terms "Google Art Project image too dark" I only find these posts here i.e. to say it seems not to have been an issue debated elsewhere. Whatever the source used, it should be documented and verifiable. I have been editing some Mauritshuis files on Commons recently. It's quite common to see high resolution digital images uploaded that can only have come from their rights managed collection. They are typically sourced as "unknown" or given spurious origins such as geheugenvannederland.nl (Memories of the Netherlands), a Dutch national archive which does provide moderate high resolution Zoomify images of some artworks but certainly not these high resolution images. Now of course one can turn a blind eye to the origin of these images and assume good faith. But the trouble is that they have invariably been tinkered with by the uploaders, lightened or warmed, and even cropped, to their personal satisfaction, and that's when it becomes problematic. But I'm afraid I do think this isn't the place to carry on this debate. I really do wonder that the community can be satisfied with showcasing an image whose source simply can't be verified. Coat of Many Colours (talk) 07:54, 31 July 2014 (UTC)
 * The images from Google Art Project are almost always too dark. I imagine they are trying to avoid specular highlights on the paint texture/cracks at all costs (which is only an issue for extremely high resolution images). Unfortunately, the cost is a dark, drab image that rarely does justice to the actual painting. It's like touring a museum with the lights off. Kaldari (talk) 05:53, 1 August 2014 (UTC)
 * That might well be. I haven't really interested myself in Google Cultural Institute's virtual tours very much. The idea is very pleasant, but in practice I find it quite frustrating positioning myself to look at the paintings. In the case of Commons:File:Auguste Renoir - Dance at Le Moulin de la Galette - Google Art Project.jpg, where it was claimed in this forum that the image was too dark, in fact the image comes from the museum database and when I go into their virtual tour at the Musée d'Orsay I don't find their images too dark. I uploaded yesterday the smaller version of this painting Commons:File:Auguste Renoir - Dance at Le Moulin de la Galette (ex Whitney collection).jpg which hadn't previously been represented on Commons. I shall be away most of August incidentally, and not contributing here.Coat of Many Colours (talk) 09:47, 1 August 2014 (UTC)

--Armbrust The Homunculus 10:19, 6 August 2014 (UTC)