Commons:Photography critiques/November 2017

Columbus past the CSX Upper Scioto River Bridge

I've been trying to improve on this but I don't feel I quite have it. I would really apreciate any advice or comments. I can't seem to finally be happy with this one. Thank you everyone in advance, -- Sixflashphoto (talk) 03:37, 9 November 2017 (UTC)

@Sixflashphoto: Mmmh, this is difficult. You've got two subjects at once in the same place (bridge + skyline), and they are really fighting a battle for attention. I don't really have an idea for solving this, but it seems to be a bit on the darker side over-all. Have you tried a color version? The skyscrapers look great in b&w, but I'm not so sure about the rest. --El Grafo (talk) 10:39, 27 November 2017 (UTC)

Beautiful scene is drab -- what went wrong?

A few weeks ago I went to a park and took these two pictures of a tree by a pond. It was about 1:00 PM, with the sun roughly behind me on a mostly clear day. Camera is an Olympus E-PL6 (mirrorless).

Taken with an Olympus 40-150mm f/4-5.6 lens (at 40mm, f/8, 1/500, iso 200)

Taken with a Panasonic 20mm prime lens with f/6.3, 1/640, iso 200

I had a wide angle lens with me, but I couldn't see at the time that the important parts weren't all in focus so I didn't bother uploading that version.

This was a very nice scene there in person, and I was surprised to see how drab the images came out, relatively speaking. What could I have done either in taking the picture or in post-processing to improve? — Rhododendrites ^talk |  07:02, 28 November 2017 (UTC)

Well, I wasn't there so I can only speculate …

• One reason might be having the sun right in your back: Everything is illuminated evenly with hardly any shadows visible, which makes the scene look pretty flat.
• I don't know how high the sun is standing in New York this time of the year at noon. But if it was rather low, it might have produced this special kind warm light that makes scenes like this look so beautiful. In that case, white balance becomes important: The EXIF data says you camera was set to auto white balance, which will lead it to produce rather neutral colors. That's pretty much the opposite of what you want in a situation like this: If you want to retain the color cast caused by a low-standing sun, you'll have to manually set your white balance. The "daylight" setting would probably work rather well, or at least provide a good starting point (if you shot RAW, this can still be done in post, and to some degree also on a JPG).
• A polarizing filter could have helped to cut down on the reflections on the lake and bring out some more texture in the sky (but be careful not to over-do it and note that polarizers also work better when the sun comes from an angle).

I general, I like the wider shot better composition-wise, but maybe try to keep the tree off-center to the left or right in a rule-of-thirds kind of way? Maybe go even wider? BTW: At the left side of this shot there's an interesting-looking birch tree with what appears to be some kind of birds net at the top – looks like that might be an interesting subject as well!

In summary, if you plan to re-shoot, my suggestions would be: 1) come a few hours earlier or later for better sun position (not straight in the back and lower on the horizon for even warmer light). 2) Set your white balance and/or shoot RAW. 3) If you've got a polarizer, give it a try. Hope that helps, --El Grafo (talk) 10:46, 28 November 2017 (UTC)

We get a lot of bright sun where I'm at and I always have a polarizing filter on my lenses (protects the glass too). You can turn them to adjust how the scene looks. Try the golden hour too. PumpkinSky talk 02:59, 30 November 2017 (UTC)

@El Grafo and PumpkinSky: Thanks. Admittedly, I almost never change it from auto white balance. Worth playing with. I keep meaning to take a white piece of paper with me, since it seems to want me to have one at all times to customize it (but there is a separate daylight setting -- or an equivalent). I just got a polarizing filter for one of the lenses (neither of those that produced these images, though -- unfortunately they're all different sizes) and tested it out yesterday. Still figuring it out and haven't gone through the images yet. Regarding the bird's nest in the background, I tried to find a way to get a bit closer to that. It's a nature reserve so many areas are fenced off, including that area, and if you get closer it gets blocked by the trees between the path and that tree. Ah well. :) — Rhododendrites ^talk |  15:54, 3 December 2017 (UTC)

@Rhododendrites: I got myself one of these white/black/grey cards but I can't remember when I last used it. They are helpful for precise exposure metering sometimes, but not so much for setting white balance outdoors. You see, calibrating your white balance using a neutral surface (such as those cards or a sheet of paper) basically does the same thing as the camera's auto white balance in a more precise manner. It will make your paper appear perfectly white, but in a situation like this that's not really what you want. To keep the golden glow of the landscape, you want your sheet of paper to be a bit on the yellow side. In the film days, you'd just choose a daylight balanced film for that. Digital sensors give you more control, but you have to tell the camera what you want it to do. You can play around with different color temperatures, but as colors can be difficult to judge on the camera's screen in the field, for critical work I usually just set it to "daylight"/"sunny" (or very rarely "cloudy"), shoot RAW, and fine-tune WB to taste in post. If you've never used RAW before, I'd recommend shooting RAW+JPG in the beginning, as editing RAW files takes some practice. They will likely look totally flat and boring at first and need much more editing than a camera-processed JPG, but once you get the hang of it it opens up a whole new world. Getting a piece of software that is more focused on raw development rather than pixel-level retouching makes things much more convenient (there are lots of options there from excellent free software like darktable, rawtherapee, lightzone etc. to expensive professional stuff like lightroom or captureOne). --El Grafo (talk) 09:35, 5 December 2017 (UTC)

Rhododendrites: Good suggestions above. I'd also recommend first reading https://luminous-landscape.com/understanding-histograms/ and then having a look at the histograms of these photos, if you haven't already. You should never trust a histogram more than you trust your eyes, but it can be helpful in understanding why your eyes tell you something's off with the exposure.

I think the main reason these photos look washed out is that, especially in the first one, you've got almost no action in the top 10% of the histogram, meaning there's very little in the way of highlights. A good start in remedying this is to play with Levels adjustment. Again, looking at the first photo, try dragging the slider for your white point input level from the 255 mark down to 235 or 230 where the action starts. You might even find that despite that big spike between the 205 and 235 mark, there's not actually much detail lost by dragging it down to around 205 (but maybe I'm just a sucker for contrasty photos). You can also bring up the black point from 0 to about 15. (In the second image, I'd bring the black point up to 30 and the white point down to 230.)

With that done, you've got something much more lively to play with. If you want more control, you can adjust the levels for individual channels or make Curves adjustments instead of levels. Hope that helps, —LX (talk, contribs) 20:25, 30 November 2017 (UTC)

@LX: This is very helpful. I've been meaning to get a better handle on the histogram for a while now. I never really use it. I should also admit that I dragged the highlights down in order to see the sky a bit better. I did try playing with the tone curve to try to restrict those adjustments to the whitest whites, but may have shot myself in the foot. I'll read that article and play with this some... — Rhododendrites ^talk |  15:54, 3 December 2017 (UTC)

Unfortunately, there is a lot of vague, overly simplified/idealized, misleading and sometimes just plain wrong information about the histogram out there on the web. Because of that, it took me quite a while to get the hang of it. This is indeed one of the best explanations I've seen so far; you might also want to have a look at this video by Matt Granger where he clears up some of the common misconceptions. --El Grafo (talk) 09:35, 5 December 2017 (UTC)