Commons:Featured picture candidates/File:Abbey of St. Jean des Vignes, Soissons, Picardy, France - Diliff.jpg

File:Abbey of St. Jean des Vignes, Soissons, Picardy, France - Diliff.jpg, not featured

Voting period is over. Please don't add any new votes.Voting period ends on 18 May 2015 at 15:48:17 (UTC)
Visit the nomination page to add or modify image notes.

Featured picture candidates/File:Abbey of St. Jean des Vignes, Soissons, Picardy, France - Diliff.jpg Commons:Featured picture candidates/File:Abbey of St. Jean des Vignes, Soissons, Picardy, France - Diliff.jpg

• Category: Commons:Featured pictures/Places/Architecture/Religious buildings
•  Info created by Diliff - uploaded by Diliff - nominated by Paris 16 -- Paris 16 (talk) 15:48, 9 May 2015 (UTC)
•  Support -- Paris 16 (talk) 15:48, 9 May 2015 (UTC)
  • Ah, you found my collection of new images from France... ;-) Diliff (talk) 15:56, 9 May 2015 (UTC)
•  Support. Thanks for the nomination. Diliff (talk) 15:56, 9 May 2015 (UTC)
•  Oppose Sorry, although decent quality I dont really like the composition or crop. For me, it is a poor compromise between focusing on the tower or on the whole Church. The typical spring green color of the trees is pleasant, but at the same time distracting (the tree is too prominent imo). It is obviously no problem to photograph the church without shadows and trees. I do not think the light and color either produce any WOW imo.--ArildV (talk) 17:43, 9 May 2015 (UTC)
•  Oppose I start off saying that the photograph is almost flawless. Great dof, great detail throughout and great exposure and dynamic range. What bother me is the perspective correction. I consider perspective correction an abused tool, giving at first a "little something" but in the end the images look unnatural. Perspective correction is no substitute for rise/fall in the view camera which gives a more "correct" perspective. I think that software perspective really gives an unnatural look, for it tends to put the subject, as in this case, on a single parallel plane to the viewer, when in reality the subject is in a relatively inclined plane. That is the difference between view cameras and digital perspective control. It actually deforms the subject while keeping verticals perpendicular to the ground, not allowing for the natural fall off of the distance between the viewer and different areas that are not equidistant from the point of view. --Tomascastelazo (talk) 18:10, 9 May 2015 (UTC)

Discussion about the digital vs optical perspective correction

You're showing your ignorance here Tomas. There's functionally no difference whatsoever between the perspective achieved digitally and that achieved with a view camera. Any perspective correction that you can achieve with a view camera (or a tilt-shift lens) can be achieved digitally too. The one difference is that you can also shift the plane of focus with view/tilt-shift cameras, and obviously that is not possible in post production. Diliff (talk) 18:36, 9 May 2015 (UTC) What is obvious is your ignorance of photography and physics, there is a difference, whether you call if functional or not. In view cameras the surfaces remain parallel all the time, whereas in digital correction the photograph is inclined to achieve the perspective correction and thus distorts the image, basically stretching the image, whereas that does no occur in view cameras. Granted, the view camera provides another type of distortion, but maintains a diminishing perspective because it retains the relationship of the different points whithin the subject and the focal plane, thus imprinting on the brain a more natural look. That, photography ignorant (at least), is called a difference. --Tomascastelazo (talk) 19:07, 9 May 2015 (UTC) In the old days, we inclined the photo paper to correct parallel lines, and we called it the poor man´s view camera. Saqme happenes with digital PC. Read a little about perspective control and distortion... #REDIRECT[[1]]--Tomascastelazo (talk) 19:17, 9 May 2015 (UTC) You're talking absolute rubbish. I've already done plenty of reading on perspective control and distortion, and I suspect that unlike you, I've actually got 10 years of experience in dealing with digital perspective distortion and correction as it relates to stitching. You know the terminology but not the physics. If you are taking a photo from a specific position, the perspective cannot be different no matter what you do with your lens. Even with tilt-shift, you cannot change the distance from your lens to an object. You cannot see around a corner to give you a different view. The view by definition is based on the point where the image converges in the lens. You still have to trace exactly the same lines from that point to the subject regardless of whether you're using a tilt-shift/view camera or a regular camera. Everything else is simply distortion of some kind or another, and view cameras distort the subject exactly the same way as digital perspective correction does. There's no getting around physics. Find one respectable source that explains how or why view cameras provide a different perspective that cannot be replicated exactly with digital perspective correction (sharpness excluded, because sharpness is just a function of the distortion, it doesn't affect the actual perspective of the subject) and I will eat my words. You won't. Physics is on my side here. And please, don't just reply with more 'you're wrong, I'm right, blah blah'. Actually find me a source that explains how it is different. I'm not interested in discussing it further if you're not interested in finding actual evidence. Diliff (talk) 20:25, 9 May 2015 (UTC) And a rethorical question, if the perspective control is the same, why would Canon, Nikon and many other make perspective control lenses? I would think that smart photographers would not buy them, or are they stupid too? Do you know something they don´t? --Tomascastelazo (talk) 19:20, 9 May 2015 (UTC) Is it really a rhetorical question do you actually care to know the answer? The answer seems obvious to me. 1. Because not everyone wants to correct perspective digitally. 2. Perspective control lenses were designed before the age of digital photography. 3. There are some advantages to the sharpness of images when corrected optically when compared to a single photo which is digitally perspective corrected. However, those advantages disappear completely when comparing to high resolution stitched digitally corrected images because even after taking into consideration those detrimental effects on the sharpness of an image, the additional resolution afforded by stitching outweighs it. Diliff (talk) 20:25, 9 May 2015 (UTC) Look here, it is basically "perspective for idiots! #REDIRECT[[2]]. There are many more, but they basically get it, and I quote "However these techniques do not enable the recovery of lost spatial resolution in the more distant areas of the subject, or the recovery of lost depth of field due to the angle of the film/sensor plane to the subject. These transforms involve interpolation, as in image scaling, which degrades the image quality, in particular blurring high-frequency detail. How significant this is depends on the original image resolution, degree of manipulation, print/display size, and viewing distance, and perspective correction must be traded off against preserving high-frequency detail." So you got 10 years experience? Hmmm, well here in Mexico we have a saying for that kind of presumtuos attitude... "When you are on your way to the store to get the milk, I{m already on the way back with the cheese." --Tomascastelazo (talk) 20:34, 9 May 2015 (UTC) Tomas, I couldn't give a toss about the sayings you have in Mexico, despite your repeated insistence on giving me these 'pearls of wisdom'. The quote you provided explains precisely what I've already been saying. The spatial resolution that it refers to is the sharpness loss that I refer to. The depth of field difference is the other difference that I explained right from the start. Neither of these points speak of a perspective difference that you claim there is - this 'unnatural look' that differentiates digitally corrected perspective. Does it? Diliff (talk) 20:38, 9 May 2015 (UTC) Dear David, you throw strong words about fellow contributors, like rubbish, ignorance, etc., and you put yourself in a pedestal of knowledge with the absolute centainty that there is no one more capable of your technical and digital savvy. I do admit the possibility that you are more proficient in those areas, but those areas are not what make up for good photography only. You are an above crafstman photographer, flawless, but a craftsman nonetheless. Your images are nearly perfect, as perfect as many beautiful, look alike, sometimes even better than real artificial flowers, but they miss something, the aroma. You obviously have a need to have no one oppose your images, or a need for perfect approval. I expressed my honest opinion and if you do not like it, so be it. The emperor has no clothes on. --Tomascastelazo (talk) 20:52, 9 May 2015 (UTC) Tomascastelazo, this is the second time you have, in my opinion, failed to accurately express your issues with an image. I'm sure the language and cultural differences we face on Commons do not help. I respect you, as an artist, for your artistic opinion on what images you think are great and what fail to achieve greatness. But I don't think you have successfully argued those opinions at all times. Here, I think you are just plain wrong about the reason the image makes you feel uncomfortable. Let's just leave it that you don't like the perspective on this image, rather than both of you fighting over why. I'm sure that David can accept that is a perfectly rational reason to oppose, and probably one you share with da Vinci. -- Colin (talk) 21:14, 9 May 2015 (UTC) My oppose is very clear, if you care to read it, the controversy starts after David did not like my reasoning... it is up there, read it. And I stand by it. --Tomascastelazo (talk) 21:27, 9 May 2015 (UTC) I have read it. Sorry, a view camera creating a large rectilinear projection on the back of the camera (which one then moves the film/sensor around), produces exactly the same shape as the one above. You may indeed find the effect unnatural, for one's eye only really sees and concentrates on the very centre of our vision, and you would not be the only great artist to object to such a projection/scene. Spend a few hours reading the source I give below. This is an issue of projection, not software vs optics. I fully agree that the top of the tower is further from the eye than the bottom. There are several ways the eye perceives distance, two of which are that the objects get smaller, and they get less clear due to atmospheric effects. But the eye has a lens that changes focus as you move around the scene, whereas a static image cannot change what is in and out of focus. And the eye has very distorted and out-of-focus capture outside of this central portion, which we ignore and shift our eyes when we want to look there. The fact that one can move one's eye round a wide-field-of-view-captured-image and it doesn't actually change is really disconcerting. It is unnatural. But it's nothing to do with optics vs software. -- Colin (talk) 21:46, 9 May 2015 (UTC) When I throw strong words around, I mean them. They are not there simply for verbal theatrics. You make the mistake to think that I consider that there is no one more knowledgeable or capable than myself. I'm open minded to new ideas. I evaluate arguments as they're presented to me, as anyone should do. If they're logical and persuasive, I accept them, if they're obviously rubbish, I treat the ideas with the contempt they deserve. It has nothing to do with an ego beyond comparison or a desperate need for approval. If you merely expressed your opinion, I'd have little to say except "you're entitled to it", but you expressed your 'opinion' about digital perspective correction as as a fact-based claim, and I corrected you. You tried to justify your claim, and I disputed it further and challenged you to prove you're right. You failed to do so - your quote only explained what I'd already mentioned as the differences between digital and tilt-shift correctsions, and had absolutely nothing to do with the claim you were making in your first point. I'm still waiting for you to prove you're right. This is not just my opinion vs your opinion, this is fact vs fiction, and I'd like to settle this. You might actually learn something in the process and recognise that your preconceived ideas about perspective are mistaken. I'm not expecting to change your vote. I don't actually care if the basis of your vote is that you don't like the composition. That's fine, I've never denied you an opinion. But please don't dress an opinion up as a fact-based claim. If you do and you're wrong, I'm going to correct you. I'd expect you to do the same, as long as you can back it up. Diliff (talk) 21:30, 9 May 2015 (UTC) Tomascastelazo is right that the sensor/film plane is parallel to the subject in a camera with tilt-shift lens. At these distances this will have a fairly small effect on DoF/focus-plane compared to angling the camera when taking a stitched photo. However, that assumes the tilt-shift lens or view camera lens has a perfectly flat focal plane, something that lens designers struggle with. As the WP article points out, many software tools are very naive about "perspective correction". But serious tools like Hugin and PtGui can correct in much the same way as a lens designer can correct when producing an extreme wide-angle lens. Which is what a tilt-shift lens is -- a (generally) wide angle lens that has a projected image-circle far larger than the sensor, which is then moved about as the lens is shifted. The "stretching" occurs in the lens towards the edge of its projected image-circle, rather than in software. It is no more magic than the effect of changing from a crop-sensor camera to a full-frame camera, while keeping the same lens. Extreme wide-angle rectilinear images have distortion problems, whether generated by a tilt-shift lens or view camera, or the output of computer software. The apparent deformation in a very wide angle scene has troubled artists from Leonardo da Vinci and earlier, leading many of them to recommend narrow field-of-view or else to "fiddle with" the image to "correct" the distortions while then altering the relationship between objects -- something only a painter can achieve. This in-depth online book tells you everything you might possible want to know about the subject, and more. The only solution, should one find the altered proportions in an ultra-wide-angle image unpleasant (and many do), is to take the photograph from further back and further up and thus narrow the field-of-view. This is, unfortunately, not always possible, or introduces other compromises in the scene. The other alternative, of unconventional projections, has its own problem with curved lines and is generally inappropriate for architecture. Fancy cameras and lenses will make not one iota of a difference to the overall shape of the projected image -- merely potential very small differences in sharpness. And this is assuming the lens has a flat focal plane and maintains sharpness towards the edges, which again is a battle against physics and something that lens designers mostly fail to achieve at less than Zeiss Otus levels of cost. Given that Dillif can capture the image at essentially whatever high megapixel resolution takes his fancy, and then downsize, the issue of sharpness is utterly irrelevant. A rectilinear projection, as we see here, is a rectlinear projection, whether created in software, with an expensive camera/lens, or by an artist with a paintbrush obeying the rules of linear perspective. It has a large field-of-view, and the issues are a consequence of that alone. -- Colin (talk) 20:50, 9 May 2015 (UTC) Indeed, Tomas is right about some of the points he made. But they were not the central points of his original claim, they were points made in support of and to justify the claim - that perspective correction made by a view camera or tilt-shift lens looks fundamentally different. It's no good being right about 1+1=2 and 2+2=4 if you're going to take the two calculations and use them to justify why 1+1+2+2=5. Diliff (talk) 21:30, 9 May 2015 (UTC) I quote my reason to oposse: "What bother me is the perspective correction. I consider perspective correction an abused tool, giving at first a "little something" but in the end the images look unnatural. Perspective correction is no substitute for rise/fall in the view camera which gives a more "correct" perspective. I think that software perspective really gives an unnatural look, for it tends to put the subject, as in this case, on a single parallel plane to the viewer, when in reality the subject is in a relatively inclined plane. That is the difference between view cameras and digital perspective control. It actually deforms the subject while keeping verticals perpendicular to the ground, not allowing for the natural fall off of the distance between the viewer and different areas that are not equidistant from the point of view." That said (again) I can tell an image that has been digitally perspective corrected, I distinguish the effect, and I do not like it. Can it be any clearer? --Tomascastelazo (talk) 21:36, 9 May 2015 (UTC) The one thing a digital image can achieve is a far wider field-of-view than most lens designers have ever commercially offered. This is indeed often over-used to get a church steeple "vertically correct" in a photo taken a few metres from the church. A view camera or tilt-shift simply wouldn't be able to create such an enormous and ridiculous field-of-view. That's the difference. For a given field-of-view, your opinion that you can "tell the difference" is as much a fairy-story as those who buy ridiculously expensive speaker cables and think they can hear the difference. -- Colin (talk) 21:46, 9 May 2015 (UTC) Colin has pretty much expressed the same thing I would have. If you think you can distinguish the difference, you are deluded. I could set up a blind test to prove you cannot tell the difference if I had the tilt-shift lens and the patience, but I don't have either. If you want to continue to believe you can distinguish the difference, go ahead. You've shown yourself to be completely disinterested in understanding perspective. We've got a saying in English too: "You can lead a horse to water but you can't make him drink". I'm going to collapse this thread and leave the discussion at that. Diliff (talk) 22:28, 9 May 2015 (UTC) Do not be so sure about what you think are my delutions. Arguing with you, judging from your comments in general as as effective as asking a rock to think. You are absolutely right all the time, in everything, and you have 10 years experience!!!, so we can leave it at that. You do not own to your own words and contradict yourself, a fact proven by reading the thread, chronologically, not cherry picking. So be right all you want, but let me tell you again, the emperor has no clothes. --Tomascastelazo (talk) 05:20, 10 May 2015 (UTC)

•  Support --Code (talk) 19:47, 9 May 2015 (UTC)
•  Oppose Sorry, but the lighting doesn't work for me; too much is in shadow. --King of ♥ ♦ ♣ ♠ 20:45, 9 May 2015 (UTC)
•  Comment This kind of photo is hard. First, its tall, doing PD correction it get unrealistic size (shortened), except doing big matrix so can be rescaled. What i dont like it extracting shadows to such extent it get noise - texture there is weird. I suppose you did best from what you can get to represent object is some "normal" way, but still lacks something more for FP. --Mile (talk) 21:26, 9 May 2015 (UTC)
  • There seems to be the misapprehension that the angle of view is very large. It's really not that large. Ignore the focal length in the EXIF data, it's not reliable. My estimate is that it's the rough equivalent of 24mm on a full frame camera, there are many buildings with a much wider focal length than this that have been perspective corrected. And also, the texture in the shadows is the way the building actually looks. The shadows have not really been boosted, this is a HDR image taken at ISO 100, and the bracketed image which has not had any shadow lifting looking exactly the same in the dark areas. The texture of the dark areas of the tower indeed just looks that way and it has nothing to do with processing. Diliff (talk) 22:47, 9 May 2015 (UTC)
•  Oppose I am not convinced by the right crop (maybe better no tree than half of it) but the fact that the subject is in shadow definitely spoils it to me. I can imagine that your time there was limited and that was the best you could make out of it, but the timing was just not convenient. Furthermore I also agree with Mile's comment that the texture of the facade doesn't look realistic to me, sorry. Poco² 08:15, 10 May 2015 (UTC)
  • I don't mind criticism of the composition but I already explained to Mile that the facade actually looks like that. I took the photo at ISO 100 so it is not noise. The image is HDR but I can confirm that the bracketed image that is correctly exposed for the facade has the same speckled texture. It's not noise, it's not a HDR artefact, it's just (I think) lichen growing on the stone. If reality doesn't look realistic, what can I do? :-) Diliff (talk) 10:35, 10 May 2015 (UTC)
    • See screenshot from Lightroom here. All values are set at zero, this is the 'neutral' processing settings of the dark bracket image. The texture of the dark areas are exactly the same, only slightly darker. Diliff (talk) 11:39, 10 May 2015 (UTC)
      • David, I believe you, you don't to show any LR screenshots, but the result is still strange. I just looked into some detail pictures (e.g. this one or this one) and the texture looks fine. By the way, looking at your picture (I didn't know the place) I wasn't aware that I was missing the probably most interesting of this Abbey: the openings. Having that lighting, why didn't you try it like this from the other side showing also those nice openings in the facade?. Btw, I took a note about the place when I travel in France :) Poco² 16:24, 10 May 2015 (UTC)
        • Well the thing is, you originally said it looks unrealistic, not strange, so I wanted to point out that it isn't unrealistic, it's as real as the camera can output and isn't the result of manipulation in post. I think you can see the same texture in the second image you referenced, although because it is in sunlight, it and doesn't accentuate the texture as much. I did actually take a photo from the side in your third link, but I wasn't as impressed. I thought this view was the best, personally, but I guess everyone has their own opinion. I'll upload my image from that angle though, it's probably still the best image we have from there. Diliff (talk) 16:54, 10 May 2015 (UTC)
          • In case you're interested, here's my image from the view you suggested. I'm still not convinced that it's a particularly aesthetic view though, but I suppose you can see the tower a bit better. It looks rather unbalanced from an (almost) straight on angle. I think asymmetry of the tower makes it better to photograph from the side. Diliff (talk) 02:30, 11 May 2015 (UTC)
• Weak  Oppose for others --Σπάρτακος (talk) 12:11, 10 May 2015 (UTC)
•  Oppose Per others. -- Pofka (talk) 18:05, 10 May 2015 (UTC)
•  Neutral Has composition problems. --Tremonist (talk) 12:56, 12 May 2015 (UTC)

Confirmed results:

Result: 3 support, 6 oppose, 1 neutral → not featured. /Yann (talk) 21:21, 18 May 2015 (UTC)