User talk:Colin

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

mosaic, lighting[edit]

Thanks for your comments at Commons:Featured picture candidates/File:Ceiling mosaic in the Surrogate's Courthouse (32325)a.jpg (also pinging C messier and Ikan Kekek.

I've uploaded a new version (old version, new version).

The question of how much post-processing is acceptable and/or what is "realistic" seems like a question about which people have wildly varied opinions. Should an image strive for what it would look like in ideal conditions, or is it necessary to try to wait for (or arrange for) those ideal conditions for a similar result? The issue here wasn't that the light happened to be bad that day. The light from outside doesn't do much at all for that mosaic other than cast a bit of a blue glow to the lower parts of the ceiling (and most strongly on the side opposite what's in this photo). If this mosaic were spiffed up, shown under bright lights, and photographed up close, it would probably look great. At what point is it ~"cheating" to try to aim for that with post-processing? It's unclear to me at what point that becomes different from if I got permission to simply remove a chunk of the ceiling to clean it, move it to where lighting is more favorable, and then put it back before doing more restrained post-processing for a similar result. :/ If it were always pitch black in that room, would no photo that uses post-processing to actually render the mosaic visible be realistic? — Rhododendrites talk |  04:05, 25 October 2018 (UTC)

My feeling is that falsifying the appearance of something you saw, so that it doesn't look like what you actually saw, is normally a bad thing. I will say that the revised photo looks better than the original, but I think that the thing that kept me from supporting the original wasn't that it required more light, but that it could have been sharper. I think that the duller light tends to accentuate any fuzziness from less than crisp sharpness. I'll be curious to read Colin's thoughts. -- Ikan Kekek (talk) 04:34, 25 October 2018 (UTC)
What about a flash, then? Or bringing light stands/reflectors? or using a shallow depth of field? (when I look at an object, what's behind it isn't in focus, but it doesn't turn into blur like a <f2.0 creates). What winds up in the camera certainly isn't what you're seeing before hitting the shutter in all of these instances. My feeling is that if it couldn't look a certain way, then yes, it's bad, but if it could look good in a particular way, I don't see a problem with using post-processing to accomplish that. When I look at the mosaic, I can see a spectacular, colorful scene, even if I'm also seeing that the light is terrible (i.e. it's clear that if conditions were different, it would be great). — Rhododendrites talk |  04:44, 25 October 2018 (UTC)
I think I agree with you both but the difficulty is that our eye doesn't work like a JPG. Wrt focus, I think actually the object behind really is blurred, and in fact only a tiny central portion of our vision is sharp, but when you want to concentrate your vision on the object behind, you move your eye and shift focus, and so it is really hard to get the same effect as a still image where the focus doesn't follow your eye. A photo is quite a unique form of art that has this focus-blurring effect. It is also unique in being able to freeze motion or blur motion. We accept these forms of photography so don't feel deceived.
Our eye also adjusts to the colour temperature of the lighting, the brightness, and doesn't see colour casts. For example, I didn't spot the green tint on the ceiling of Paisley Abbey (here and here) but I think it does exist. I minimised it a bit in the photos. Lighting can make all the difference. Like between this shaver photo and that shaver photo. The sparkly brilliance in my shaver photo is a result of a single flash light bounced off the my kitchen ceiling with the kitchen lights off. It isn't what the shaver looks like when all the kitchen lights are on. Since nobody uses a shaver in a darkened room lit only by flash lighting, I suppose it is a bit super-realistic. Realty sucks sometimes.
Where I think it most important to be neutral and fair is when photographing other artworks. I think it ok to compensate for bad or uneven lighting, to aim for something you'd get when conditions were more controllable. But not to turn the artwork into something it isn't. Old faded paintings are old and faded. Perhaps the question is whether someone visiting the artwork after seeing your image would feel deceived. Looking at some other pictures of the ceiling mosaic (here and here) and in the photos in this article it does look like the mosaics are low contrast in reality. Is it due to the lighting? How would they look at night under artificial light, or if photographed with flash? Are they dusty? If someone gave them a good clean and then produced before/after photos, would it be embarrassing to now yours should look like the before ones? I recall having problems with my photos of Fitzrovia Chapel which also had mosaic and gold leaf. What does one do if the mosaic is dull and dusty and so nobody at FP goes wow unless you push the processing a bit?
I guess at FP we expect you to bring out the best in what you see, but not go so far that we feel cheated. Perhaps we are over-examining your mosaic pictures now, and if you'd uploaded the current version originally, nobody would have said anything. -- Colin (talk) 08:07, 25 October 2018 (UTC)
Just realized I didn't respond to this, although I'm not sure there's much to add. I suppose I just get frustrated at the lines people draw, often seemingly arbitrary, for what aspects of perception it's acceptable to enhance, negate, fudge, or compensate for, or what kinds of instruments/electronics/lenses/accessories are acceptable to use, how much something should be "made" to look like it "actually" looks (to whom? when? how?), etc. It often seems like the purpose is less to conform to some Platonic ideal or some objective reality than to distinguish a "high photography" from "low photography" or to conform to what, at some point, emerged from practitioners as a universal code of what kinds of environmental, electronic, optic, etc. distortions or compensations are permissible and when that makes the image lousy rather than better. I'm not exempt from this either, of course. There are certain kinds of radical edits that put me off, too. Just venting a bit here is all. :) — Rhododendrites talk |  06:42, 12 November 2018 (UTC)

ANU thread[edit]

As Jcb closed the ANU thread I will ask you here. Can you apologize for the accusations of vandalism and wikilawyering? - Alexis Jazz ping plz 01:13, 12 November 2018 (UTC)

In Commons:Deletion requests/Files on User:Faebot/SandboxT Fae listed 258 images that provided portraits of notable figures. I know Jan isn't flavour of the month, being blocked and all that, but think about what was involved in creating those. I'm not sure how many YouTube videos they came from? A handful, a dozen, several dozen? Each of them watched until you find the best frame for the person. Screen grab, crop, adjust. Then Jan labels the JPG with the title of the person in the frame, the creator of the video, the event the image was taken from, the source video url (though sometimes imprecise) and the licence url (which may or may not have a 4 rather than a 3 it should have had). He then uploads each one with Upload Wizard, supplying again those details for the file description page, along with the appropriate category for the person (which can be time-consuming to locate).
He created those images over a period of several months earlier this year. Fae wanted them all deleted based on a single digit being possibly incorrect in the EXIF. That can only be described as vindictive gravedancing, and there's no way they would have been deleted. There are simply too many examples, in the upload history of anyone who uploads third-party JPGs, where the EXIF of a JPG is inconsistent with the claims made on Commons or with reality. And despite Jan's weird obsession with colour alterations, at least he is using a professional image application that correctly saves the colour profile in a JPG, unlike the broken applications like MS Paint or Paint.Net.
So, we're talking hours and hours of work. And the extra EXIF information was added by Jan in good faith because he knows such information survives when the JPG is downloaded from Commons and used elsewhere. It is useful stuff. It wasn't perfect, but there are plenty cases where people link to a source URL that isn't as precise as it should be or has moved, or where they get the licence wrong.
If Fae had got is way, all those hours of work would have been wasted. So on the one had it is good that you volunteered to fix the issue. But you were way too quick, and didn't wait to see if anyone agreed with Fae that there was a problem, that it needed fixed or what it should be fixed to. Simply volunteering was all that was needed to neutralise Fae's claim that volunteers wouldn't/shouldn't fix it. As you see, most people thought the DR ridiculous and it closed with a keep. Your edits, which totally destroyed all the EXIF in 30 files (at that point, with the intention of doing the rest) was taking a bulldozer to all that metadata. Data which someone had taken time to type and insert file-by-file. There really isn't another word for that than vandalism, which my dictionary describes as "action involving deliberate destruction of or damage". Whether your overall intentions were good or bad, you were thoroughly reckless and destructive. And when asked to stop, responding with a "fuck you, I'm actually going to speed up" kind of response, is a sure way to end up at AN/U. -- Colin (talk) 10:17, 12 November 2018 (UTC)