Something different for me: street photography by night

[jlw]

I'm in Kansas City at the moment, and this evening I went to "First Fridays," a monthly event at which art galleries in the hip Crossroads Arts District collaborate to have openings, parties, and other events during the evening.

I had never attended before, and it really draws a large, widely assorted crowd: everything from the desperately trendy to very dressed-up young women to families with kids to people walking their dogs. It was interesting enough that I was motivated to do something I almost never do: just wander around snapping, without worrying too much what the pictures would look like or even if I'd get any results at all.

It was night, so I set my R-D 1 to EI 1600 and used the fastest lens I had with me, the 35/1.2 Nokton. This produced an interesting side effect that I've noticed before when taking pictures in very dark environments: the camera actually picks up more than my eye can see, so when looking at the pictures afterward I'd notice details I hadn't seen originally, or in some cases see social interactions that left me completely baffled as to what was going on.

I suppose that ultimately most of the pictures are trivial and I'll probably lose interest in most of them fairly quickly, but I've learned that when trying something different it's best for me to let it "settle" for a while rather than making snap judgments about it.

Meanwhile, if anyone else wants to see them, I've put them in a gallery on my new website; click here for the link.

I'm not fishing for compliments on these, but there's one topic on which I'd particluarly appreciate the group's thoughts: When taking pictures in the dark, how important is it for the image to look dark? As I said, in most of these, my impression of the actual scene was much darker than the picture suggests. But if I darkened the images to the point of "looking dark," they'd be almost illegible!

The same goes for white balance: In some cases I adjusted the white balance to try to get natural color, but the scene under the street lights hadn't looked natural, so the ones with "correct" color balance actually don't communicate my impression of what I experienced.

How does everyone else handle this...?

 

[rxmd]

I find this is one of the areas where film still excels, because of the high contrast situations at nicht. Your shot is nice, but one can still see how detail is lost in the highlights. The high latitude of negative film allows a very selective approach, having a general atmosphere of darkness and still keeping detail in the highlights.

At present I try to shoot XP2 wherever possible - I've just ordered a bulk roll of it. The film is not very pushable beyond about 600 ASA, though; I just lost a couple of good portraits from Tashkent due to that. Where XP2 is impossible I shoot either Neopan 1600 or Delta 3200, both at 1000 to 1200 ASA.

 

[ChrisN]

JLW, I was about to ask how you felt the RD-1 performed in this situation, compared with film. I like the series - it's a great way to show us the character and feel of the evening - looks like fun! What prompts my question is the apparent digital noise in the darker shots (my dSLR gives similar noise, which I do not enjoy).

 

[tomasis]

wonderful photo there! I like it.

For me, night photography can be easily stimulating. I'd snap much more good and interesting shoots at nights than at daytime (maybe it depends that I'm living in small boring town) Though using colors at night is tricky part. But B&W becomes so lovely at iso 1600 which I enjoy of.

 

[jlw]

JLW, I was about to ask how you felt the RD-1 performed in this situation, compared with film. I like the series - it's a great way to show us the character and feel of the evening - looks like fun! What prompts my question is the apparent digital noise in the darker shots (my dSLR gives similar noise, which I do not enjoy).

Well, one difference compared with film is that if I had shot these on film, you'd be seeing them in about a month (after a lot of scanning and cleanup) rather than on the same night I shot them!

Another thing to keep in mind is that the ones that look especially noisy are ones that I inadvertently underexposed drastically (hey, it happens!) In these cases, I had to apply big boosts to the "exposure" and "fill light" controls in Adobe Lightroom, just to get a legible image. I knew this also would exaggerate the shadow noise -- but since the whole point of this exercise was just to relax and take what the process would give me, I went ahead and did it anyway. The shots that got more normal levels of exposure look decently smooth to me, which is typical of my experience with the R-D 1 in low light.

If I had done this project on film, I suspect I would have used b&w film. I've just never liked the results of the ultra-speed CN films under low light, especially when the end product is going to be a scan. The way they grain up (and acquire flecks of color) in shadow areas actually looks worse to me than digital sensor noise.

If I had shot on conventional b&w film (as I did in my late-night drag racing series a few years ago; click here for link) I could have shot at nearly the same EI rating I used with the R-D 1, and have gotten a good tonal range. But then I'd be fighting grain aliasing and dust -- I'm sure most people know by now that all the film-scanner trick features such as "digital ICE" do not work when scanning non-chromogenic (conventional silver-based) b&w films.

Shooting in b&w also would have spared me all the decision-making about color balance, but it also wouldn't have given me the unexpectedly vivid colors you sometimes find under low light at night. In low-light photography there are always tradeoffs!

 

[jlw]

I find this is one of the areas where film still excels, because of the high contrast situations at nicht. Your shot is nice, but one can still see how detail is lost in the highlights.

I shot these in raw mode, so if I had the patience to open them in Photoshop I probably could have recovered a lot of the highlight detail. But that wasn't the point of this exercise -- I've been doing a lot of very controlled shooting lately, so I did this project to get back in the rhythm of working fluidly and without worrying too much!

 

[furcafe]

Some interesting shots there. We have a similar 1st Friday here in Washington (though it's mainly for the Dupont Circle area & other neighborhoods have their arts nights on other days of the week).

I prefer to keep the image approximately the same level of darkness as the actual scene (or @ least how I remembered it being), my reasoning being that the most interesting thing about night scenes is usually the light itself (which tends to be highly directional & therefore inherently more interesting than diffuse light) &/or the lack thereof (I also find shadows inherently interesting, perhaps why I like film noir & Neopan 1600). While it is true that the camera often captures more shadow detail than I remember seeing, I only rarely find that what was hidden in the shadows to be of great interest in the sense of adding anything to the visual appeal of a particular shot.

As far as color balance, I also generally prefer to hew to what my eyes saw, w/a minimal level of correction to keep a scene from looking too green, etc. Essentially, I do the same as if I was shooting a show that incorporates colored stage lighting.

I'm not fishing for compliments on these, but there's one topic on which I'd particluarly appreciate the group's thoughts: When taking pictures in the dark, how important is it for the image to look dark? As I said, in most of these, my impression of the actual scene was much darker than the picture suggests. But if I darkened the images to the point of "looking dark," they'd be almost illegible!

The same goes for white balance: In some cases I adjusted the white balance to try to get natural color, but the scene under the street lights hadn't looked natural, so the ones with "correct" color balance actually don't communicate my impression of what I experienced.

How does everyone else handle this...?

 

[jlw]

As far as color balance, I also generally prefer to hew to what my eyes saw, w/a minimal level of correction to keep a scene from looking too green, etc. Essentially, I do the same as if I was shooting a show that incorporates colored stage lighting.

That's an interesting thought, but I'm not very confident that what the camera records is similar to what my eyes saw! Where I was shooting was lit mostly by vapor lamps, which have a somewhat amber color... but my perception of that color was that it wasn't as strong as it looked in the uncorrected photos.

I wound up correcting most of them subjectively, which is what it sounds like you do as well.

It's different with stage lighting, in which there can be all kinds of colored gels but the basic illuminant tends to be a well-controlled source (3200K theatrical lamps) so if you set the camera white balance to that of the basic illuminant, you'll capture the colors of the gels fairly accurately.

If I remember the theory correctly, vapor lamps (and fluorescents) are different from naturally-occurring light sources in that they emit a "discontinuous spectrum" (parts of the spectrum are actually missing) so it's impossible in principle to set a truly accurate white balance for them.


I'm still not sure what to think of the darkness-level issue. In general I agree with your idea that the goal is to capture the impression of what you saw. But because the eyes are connected directly to the brain, whereas a camera isn't, I think there are perceptual differences in what you can actually SEE in a dark environment vs. what would show up in a photo corrected to look similarly dark.

If I kick up the photo to the point where someone else can see what I knew was there, then the photo no longer looks realistically "dark." But if I leave it uncorrected enough to retain the impression of darkness, a viewer might not notice what motivated me to take the picture in the first place.

It reminds me a bit of another thing I did that weekend, which was go to the Nelson-Atkins Art Museum. I spent some time looking at a painting by Ad Reinhardt, who created abstract paintings consisting entirely of geometric patterns of very slightly different shades of black.

The wall label noted that Reinhardt deliberately painted these works so that the pattern would NOT be visible at first glance -- he was interested in the idea that you'd have to invest some time standing in front of the painting and give your eyes a chance to adjust until you could perceive the different shades.

That works fine in the environment of a museum, where presumably viewers assume that every object on the walls is important and is there for a reason, and might be willing to invest the time and mental effort to figure it out.

But I'm concerned that if I tone my low-light photos so that the subject is only barely visible, people viewing them in casual environments (such as an online gallery) might not even bother to wonder if there's anything more to see. After all, there are a lot of really crappy photos online; it would be pretty easy to conclude, "Oh, well, another technically incompetent crappy photo," and move on without trying to look at what I was trying to get into it.

I don't like art that's too obvious, and I've always liked the line in the Wallace Stevens poem that goes:

The poem should resist the intelligence
almost successfully.​

But sometimes it's a fine dividing line between "non-obvious" and "too obscure to be worth the bother." Kind of interesting to think that how you set the exposure slider can determine which side of that line your photo winds up on, but there you are... that's why photography isn't easy!



(PS -- Looking back at the original post, I just noticed that the photo in it looks a lot darker and more atmospheric on my desktop computer than it did on my laptop -- so that's another perceptual variable I can't control! In other words, if I set the photo so the perceived darkness looked "just right" on one particular computer, it might easily look too light or too dark on the next computer!)

 

[furcafe]

You obviously shoot in better-lit stage environments than I do! OK, I suppose I use the term "stage" rather loosely, as I often shoot in performance spaces & clubs, etc. where the light can be made up of everything from household incandescents to Xmas string lights, to compact fluorescents, to halogen spot lights, to theatrical lamps, or some unholy combination of the above. I wish I could simply set my color temperature manually (possible on the M8, not on the R-D1) & get the white balance correct in-camera (yes, it's more of a convenience, not a necessity, when shooting raw as I always do, but it would be nice to occasionally shoot JPGs when a fast turnaround is needed). 😛

That's an interesting thought, but I'm not very confident that what the camera records is similar to what my eyes saw! Where I was shooting was lit mostly by vapor lamps, which have a somewhat amber color... but my perception of that color was that it wasn't as strong as it looked in the uncorrected photos.

I wound up correcting most of them subjectively, which is what it sounds like you do as well.

It's different with stage lighting, in which there can be all kinds of colored gels but the basic illuminant tends to be a well-controlled source (3200K theatrical lamps) so if you set the camera white balance to that of the basic illuminant, you'll capture the colors of the gels fairly accurately.

If I remember the theory correctly, vapor lamps (and fluorescents) are different from naturally-occurring light sources in that they emit a "discontinuous spectrum" (parts of the spectrum are actually missing) so it's impossible in principle to set a truly accurate white balance for them.

If I have a shot that has a small but significant detail or area that would otherwise fall completely into shadow, I will occasionally use some limited dodging to bring it up or burn in the other parts of the picture, but that's about it. The shadow/highlight tool in Photoshop helps too, but is obviously rather blunt. At some point, once you're shooting enough w/a particular film or sensor, I think previsualization kicks in & you start adjusting your shooting & composition w/the end result & the limitations of the medium in mind.

I am in complete agreement w/you about the online viewing conundrum. And the effects of monitor calibration & correct color profiles, or more likely the lack thereof, are very significant; so much so that I've basically given up on the idea of conveying subtlety through any of my online uploads, as there are just too many variables involved. Judging from sites like flickr & photo.net, "non-obvious" just doesn't get any traffic online, anyway (though real life isn't always much better).

I'm still not sure what to think of the darkness-level issue. In general I agree with your idea that the goal is to capture the impression of what you saw. But because the eyes are connected directly to the brain, whereas a camera isn't, I think there are perceptual differences in what you can actually SEE in a dark environment vs. what would show up in a photo corrected to look similarly dark.

If I kick up the photo to the point where someone else can see what I knew was there, then the photo no longer looks realistically "dark." But if I leave it uncorrected enough to retain the impression of darkness, a viewer might not notice what motivated me to take the picture in the first place.

It reminds me a bit of another thing I did that weekend, which was go to the Nelson-Atkins Art Museum. I spent some time looking at a painting by Ad Reinhardt, who created abstract paintings consisting entirely of geometric patterns of very slightly different shades of black.

The wall label noted that Reinhardt deliberately painted these works so that the pattern would NOT be visible at first glance -- he was interested in the idea that you'd have to invest some time standing in front of the painting and give your eyes a chance to adjust until you could perceive the different shades.

That works fine in the environment of a museum, where presumably viewers assume that every object on the walls is important and is there for a reason, and might be willing to invest the time and mental effort to figure it out.

But I'm concerned that if I tone my low-light photos so that the subject is only barely visible, people viewing them in casual environments (such as an online gallery) might not even bother to wonder if there's anything more to see. After all, there are a lot of really crappy photos online; it would be pretty easy to conclude, "Oh, well, another technically incompetent crappy photo," and move on without trying to look at what I was trying to get into it.

I don't like art that's too obvious, and I've always liked the line in the Wallace Stevens poem that goes:

The poem should resist the intelligence
almost successfully.​

But sometimes it's a fine dividing line between "non-obvious" and "too obscure to be worth the bother." Kind of interesting to think that how you set the exposure slider can determine which side of that line your photo winds up on, but there you are... that's why photography isn't easy!



(PS -- Looking back at the original post, I just noticed that the photo in it looks a lot darker and more atmospheric on my desktop computer than it did on my laptop -- so that's another perceptual variable I can't control! In other words, if I set the photo so the perceived darkness looked "just right" on one particular computer, it might easily look too light or too dark on the next computer!)