Commons:Featured picture candidates/File:York Minster Rood Screen, Nth Yorkshire, UK - Diliff.jpg
Jump to navigation
Jump to search
File:York Minster Rood Screen, Nth Yorkshire, UK - Diliff.jpg, featured[edit]
Voting period is over. Please don't add any new votes.Voting period ends on 21 Feb 2015 at 10:15:49 (UTC)
Visit the nomination page to add or modify image notes.
Info created by Diliff - uploaded by Diliff - nominated by Diliff -- Diliff (talk) 10:15, 12 February 2015 (UTC)
Support -- Diliff (talk) 10:15, 12 February 2015 (UTC)- Support Excellent. --Uoaei1 (talk) 14:24, 12 February 2015 (UTC)
- Support -- Pofka (talk) 15:57, 12 February 2015 (UTC)
- Support Daniel Case (talk) 18:13, 12 February 2015 (UTC)
- Support --Martin Falbisoner (talk) 06:41, 13 February 2015 (UTC)
- Support Another masterpiece for my Diliff personal churches book --The_Photographer (talk) 10:27, 13 February 2015 (UTC)
- Support If I may... you look sometimes on the verge of overprocessing (I give it to you that I don't crosscheck all the times). - Benh (talk) 12:26, 13 February 2015 (UTC)
- I know you feel that way generally. What aspect of it do you think is overprocessed? The strong micro-contrast is an unavoidable (as far as I know, anyway) side effect of Lightroom's tone mapping algorithms, but different scenes seem to be affected by it to different degrees. Sometimes I wish I could turn the effect down a bit, but I can't. I certainly don't go out of my way to overprocess it because I share your dislike for the overprocessed HDR look. I've just uploaded a new version - I've tried to minimise anything that you might take issue with - the red colour of the wood and the gilded gold are toned down slightly. I've sent the clarity slider into the negative (usually a bad idea!) to try to reduce the micro-contrast. I'm not sure if this is an improvement in your mind but after reflecting on the image, I see the changes as minor but a slight improvement. Diliff (talk) 17:14, 13 February 2015 (UTC)
- Sometimes only ;) (I insist). I don't really know how to describe it, but yes, might be a little too contrasty and saturated (this is what I'd tone down). Strangely this is more obvious on the thumbnail, but it looks quite nice when browsed at 100%. The very fine details of the (beautiful !) roodscreen scream HDR a little bit too much. Have you tried processing with only a single exposure and check at the difference? As sidenote, I can understand you can not fine tune your pipeline for each picture given your (huge! do you outsource? ;) ) ouput of pictures. But as we discussed, just nitpicking from me ! - Benh (talk) 17:44, 13 February 2015 (UTC)
- Yes, single exposures always have a slightly different tonality. I've just tried processing a single exposure and it's similar, but with less micro-contrast (and more global contrast, because it's not HDR). Here it is, it's probably more to your tastes but the overexposed and underexposed areas make it inferior IMO. Diliff (talk) 18:55, 13 February 2015 (UTC)
- Benh re your comment about thumbnail vs 100%. Someone pointed out a month or so ago that the thumbnailing code for MediaWiki was performing badly due to a software change (they were creating thumbnails by progressively reducing+sharpening from original to smaller to smaller and so on, rather than always from the the original). I don't know if this "feature" has been reverted yet, but something to bear in mind. -- Colin (talk) 19:44, 13 February 2015 (UTC)
- @Colin: That might have been me. It has been reverted in the code repository and I'd assume that the change has gone live on Commons since December. --El Grafo (talk) 18:12, 17 February 2015 (UTC)
- El Grafo, thanks. Good news. -- Colin (talk) 18:55, 17 February 2015 (UTC)
- @Colin: That might have been me. It has been reverted in the code repository and I'd assume that the change has gone live on Commons since December. --El Grafo (talk) 18:12, 17 February 2015 (UTC)
- Benh re your comment about thumbnail vs 100%. Someone pointed out a month or so ago that the thumbnailing code for MediaWiki was performing badly due to a software change (they were creating thumbnails by progressively reducing+sharpening from original to smaller to smaller and so on, rather than always from the the original). I don't know if this "feature" has been reverted yet, but something to bear in mind. -- Colin (talk) 19:44, 13 February 2015 (UTC)
- Yes, single exposures always have a slightly different tonality. I've just tried processing a single exposure and it's similar, but with less micro-contrast (and more global contrast, because it's not HDR). Here it is, it's probably more to your tastes but the overexposed and underexposed areas make it inferior IMO. Diliff (talk) 18:55, 13 February 2015 (UTC)
- Sometimes only ;) (I insist). I don't really know how to describe it, but yes, might be a little too contrasty and saturated (this is what I'd tone down). Strangely this is more obvious on the thumbnail, but it looks quite nice when browsed at 100%. The very fine details of the (beautiful !) roodscreen scream HDR a little bit too much. Have you tried processing with only a single exposure and check at the difference? As sidenote, I can understand you can not fine tune your pipeline for each picture given your (huge! do you outsource? ;) ) ouput of pictures. But as we discussed, just nitpicking from me ! - Benh (talk) 17:44, 13 February 2015 (UTC)
- I know you feel that way generally. What aspect of it do you think is overprocessed? The strong micro-contrast is an unavoidable (as far as I know, anyway) side effect of Lightroom's tone mapping algorithms, but different scenes seem to be affected by it to different degrees. Sometimes I wish I could turn the effect down a bit, but I can't. I certainly don't go out of my way to overprocess it because I share your dislike for the overprocessed HDR look. I've just uploaded a new version - I've tried to minimise anything that you might take issue with - the red colour of the wood and the gilded gold are toned down slightly. I've sent the clarity slider into the negative (usually a bad idea!) to try to reduce the micro-contrast. I'm not sure if this is an improvement in your mind but after reflecting on the image, I see the changes as minor but a slight improvement. Diliff (talk) 17:14, 13 February 2015 (UTC)
- Just showed the picture to Nicolas (so he knows a bit about HDR and stitching) and his first reaction was "C'est pas une photo !" (let you check your French lessons ;) ). So it's not only me. You guessed right, i like the single exposure one better overall. It doesn't seem to trade off that much compared to the HDR, and looks more photograph. Colin, Good to know. Last time (10 years ago?) I played with MediaWiki, they provided interface to ImageMagick's conv, and if I remember right, the downsampling was done from the original picture and cached somewhere. Don't remember they played with sharpening and all! - Benh (talk) 21:15, 13 February 2015 (UTC)
- There will always be 'purists' out there who expect a photo to have a particular look. It's still a photo as far as I'm concerned. :-) Diliff (talk) 22:13, 13 February 2015 (UTC)
- I beg to differ. It's not about particular look, it's about something is out of place when looking at an overprocessed picture (like my brain anticipate how reality should like and the picture doesn't match that). To me HDR is for compensating shortcoming of a sensor's dynamic range, not to necessarily recover ALL highlights and shadows to the point reality is broken and it doesn't match any more what the eye saw then. But yes, this one is still (more than) acceptable as a photo. Otherwise, I would happily oppose ;-). - Benh (talk) 23:11, 13 February 2015 (UTC)
- My point was that images from a DSLR frequently don't resemble what a human eye sees either, but many people are familiar with this look and expect photography to look like it, just as many people don't like fisheye or other uncommon projections because it represents a scene in an unexpected way, even though there's nothing inherently more 'correct' about rectilinear. Our eyes don't see a scene in true rectilinear projection but because we scan a scene in small areas at a time, our brain builds up a picture that approximates rectilinear. I agree with you about how HDR should be used and that's also how I try to use it. I never tried to render highlights and shadows as midtones. Dark things are still dark, bright things are still bright, but our eyes do generally see most detail in shadows and highlights so it makes sense to render them viewable at the extreme shadows and highlights of the image. I'm not trying to argue that HDR is perfect or that my use of HDR here is perfect, only that it can represent reality reasonably faithfully (when it's not done badly - full of haloes, oversaturated colours and lacking contrast). Diliff (talk) 11:53, 14 February 2015 (UTC)
- It is interesting that the HDR image is being compared with the single-exposure image as though the latter might be more faithful or the former a deviance from reality. The "looks (or doesn't look) like a photograph" comment says more about expectations than what matters does it not? An artist can choose his interpretation of reality so why not a photographer? While it is useful, if one is discussing how to make an image, to compare different exposures and processing techniques, the comparison of the two falls into a similar mistake as to compare b&w with colour and assume the only faithful transition from colour to b&w is desaturation. My books on b&w photography comment that we are less tolerant of gross alternations of colour images than of b&w images, partly because we have already accepted the latter's disconnect with what the eye sees. But colour has a history too of a conscious choice for one interpretation such as Kodachrome vs Fuji Velvia. There are those who claim to appreciate "Minolta colours" or "Zeiss micro contrast" yet software now lets us fiddle in ways no glass engineer could. What Lightroom Camera Calibration Preset you use can have a marked change on the contrast and colours. The look of this scene will also change dramatically depending on the daylight (strength, temperature, angle) and any artificial lighting. Without being there, it is hard to judge and I don't feel a single-exposure represents anything much more than what Adobe's Lightroom engineers feel is a pleasing starting point for the average photo. That said, there's always the temptation to bring out detail in shadows that really should remain hidden or to tone down the full glare of a sunlit window. Sometimes were are tempted to judge an image by "how I'd have taken/processed it" rather than simply weigh up its strengths and weaknesses for what it is. -- Colin (talk) 12:40, 14 February 2015 (UTC)
- Colin this left me speechless as I had to decipher first. I won't go into this many considerations, I lack knowledge, no one cares, and anyways my point was just that to me (and at least to my friend) the picture ressembles a painting more than a photo on the thumbnails, and in some of the finest parts of the roodscreen, hence my comments. Yes where an image ends looking like a photo is debatable. As for me any BW, velvia film and whatever look like photo. Some badly overdone HDR look like computer generated and because I did and saw many of them I can tell the symptoms quickly. But maybe I was educated like that and it's fine with others. - Benh (talk) 22:22, 16 February 2015 (UTC)
- I don't want to distract you from this discussion with Colin, but do you see any improvement in the tonality of the rood screen detail now? I've uploaded a number of updates to it that try to tone down the microcontrast, the initial issue that we both agreed was less than ideal. Diliff (talk) 23:41, 16 February 2015 (UTC)
- Apologies for the over-long comment :-) -- Colin (talk) 18:55, 17 February 2015 (UTC)
- Colin don't know if you apologize because I said I had to decipher. If so, it was only because I don't read english so well. Diliff I think it does look better, though I seem to always see HDR signature. You've listened to enough of my
complaintsfeedbacks I think, which I thank you for. - Benh (talk) 22:45, 17 February 2015 (UTC)
- Colin don't know if you apologize because I said I had to decipher. If so, it was only because I don't read english so well. Diliff I think it does look better, though I seem to always see HDR signature. You've listened to enough of my
- Apologies for the over-long comment :-) -- Colin (talk) 18:55, 17 February 2015 (UTC)
- I don't want to distract you from this discussion with Colin, but do you see any improvement in the tonality of the rood screen detail now? I've uploaded a number of updates to it that try to tone down the microcontrast, the initial issue that we both agreed was less than ideal. Diliff (talk) 23:41, 16 February 2015 (UTC)
- Colin this left me speechless as I had to decipher first. I won't go into this many considerations, I lack knowledge, no one cares, and anyways my point was just that to me (and at least to my friend) the picture ressembles a painting more than a photo on the thumbnails, and in some of the finest parts of the roodscreen, hence my comments. Yes where an image ends looking like a photo is debatable. As for me any BW, velvia film and whatever look like photo. Some badly overdone HDR look like computer generated and because I did and saw many of them I can tell the symptoms quickly. But maybe I was educated like that and it's fine with others. - Benh (talk) 22:22, 16 February 2015 (UTC)
- It is interesting that the HDR image is being compared with the single-exposure image as though the latter might be more faithful or the former a deviance from reality. The "looks (or doesn't look) like a photograph" comment says more about expectations than what matters does it not? An artist can choose his interpretation of reality so why not a photographer? While it is useful, if one is discussing how to make an image, to compare different exposures and processing techniques, the comparison of the two falls into a similar mistake as to compare b&w with colour and assume the only faithful transition from colour to b&w is desaturation. My books on b&w photography comment that we are less tolerant of gross alternations of colour images than of b&w images, partly because we have already accepted the latter's disconnect with what the eye sees. But colour has a history too of a conscious choice for one interpretation such as Kodachrome vs Fuji Velvia. There are those who claim to appreciate "Minolta colours" or "Zeiss micro contrast" yet software now lets us fiddle in ways no glass engineer could. What Lightroom Camera Calibration Preset you use can have a marked change on the contrast and colours. The look of this scene will also change dramatically depending on the daylight (strength, temperature, angle) and any artificial lighting. Without being there, it is hard to judge and I don't feel a single-exposure represents anything much more than what Adobe's Lightroom engineers feel is a pleasing starting point for the average photo. That said, there's always the temptation to bring out detail in shadows that really should remain hidden or to tone down the full glare of a sunlit window. Sometimes were are tempted to judge an image by "how I'd have taken/processed it" rather than simply weigh up its strengths and weaknesses for what it is. -- Colin (talk) 12:40, 14 February 2015 (UTC)
- My point was that images from a DSLR frequently don't resemble what a human eye sees either, but many people are familiar with this look and expect photography to look like it, just as many people don't like fisheye or other uncommon projections because it represents a scene in an unexpected way, even though there's nothing inherently more 'correct' about rectilinear. Our eyes don't see a scene in true rectilinear projection but because we scan a scene in small areas at a time, our brain builds up a picture that approximates rectilinear. I agree with you about how HDR should be used and that's also how I try to use it. I never tried to render highlights and shadows as midtones. Dark things are still dark, bright things are still bright, but our eyes do generally see most detail in shadows and highlights so it makes sense to render them viewable at the extreme shadows and highlights of the image. I'm not trying to argue that HDR is perfect or that my use of HDR here is perfect, only that it can represent reality reasonably faithfully (when it's not done badly - full of haloes, oversaturated colours and lacking contrast). Diliff (talk) 11:53, 14 February 2015 (UTC)
- I beg to differ. It's not about particular look, it's about something is out of place when looking at an overprocessed picture (like my brain anticipate how reality should like and the picture doesn't match that). To me HDR is for compensating shortcoming of a sensor's dynamic range, not to necessarily recover ALL highlights and shadows to the point reality is broken and it doesn't match any more what the eye saw then. But yes, this one is still (more than) acceptable as a photo. Otherwise, I would happily oppose ;-). - Benh (talk) 23:11, 13 February 2015 (UTC)
- There will always be 'purists' out there who expect a photo to have a particular look. It's still a photo as far as I'm concerned. :-) Diliff (talk) 22:13, 13 February 2015 (UTC)
- Support -- Colin (talk) 12:40, 14 February 2015 (UTC)
- Support --Code (talk) 19:06, 14 February 2015 (UTC)
- Support --Wladyslaw (talk) 06:56, 19 February 2015 (UTC)
- Support ///EuroCarGT 03:34, 20 February 2015 (UTC)
Confirmed results:
This image will be added to the FP gallery: Places/Interiors
