Friday, September 10, 2010

Photography Case Study: "Fall Fest Fun" by James Green

Today I've invited my brother, a more talented photographer than I, to guest blog while I'm busy doing a number of things. His blog, "Twenty-Two For a Moment," can be found here. Take it away, Jimbo:

For today's guest blog, I've chosen to use a picture which I feel really incorporates as many I guess 'techniques' as possible which are unique to me and that I am good at: manual metering and focusing.

This is a shot from the "Fall Fest" in Lone Tree, IA on August 16, 2009, also visible here on my flickr page.

Tech specs...
Nikon D40 with Vivitar 28mm 1:2.5, ISO 200, 3 sec exposure... that's about it.
This is a lens from the 70s or 80s, non CPU meaning it doesn't talk to the D40. Implications: no metering, no auto-focus (, no Exif data).

No Metering: I have had this camera since August 2008 and it has been the basis for almost all of my learning-through-trial, error, and experience. I have always shot exclusively Raw and in full manual, just discovering a couple of months ago from my girlfriend that there's this little gauge on the screen that tells me if my settings will make a picture too light or too dark. This is how I discovered "metering". Well, it turns out 9 times out of 10, just by looking at a scene I can guess the ISO, Exp, and aperture within 1/5 EV of the camera's light meter, and when I'm off, my guess usually yields a result closer to what I desire than the meter does, so I never use the meter anyway. That being said, I don't think older lenses like this work with higher-end cameras' light meters either, so if you're using something like this you have to meter intuitively. It's really something you figure out by taking test images until you develop a feel for what's going to look right. That makes digital cameras a great tool for learning. You can take test images and review them immediately. So, I learned how to use my camera in full Manual without even being aware of the meter, and developed heuristics for honing in on the exposure I want.

For what's going on here, there is plenty of light for things I want to highlight: firecracker and rides. As such, I don't need a higher ISO (which would compromise every aspect of image quality, as the D40's high-ISO performance is, in my opinion, grossly overrated) than the base 200, nor a particularly fast aperture, like 2.5 (which would compromise sharpness and depth of view), so I went with ISO 200, and I think aperture is about F/8, though I didn't record it and this lens leaves no Exif data, so that's a guess. In a scene like this, you really won't notice a lot of difference in brightness with a 1 second exposure versus a 30 second exposure; I chose 3 seconds because that got one fire cracker, got full movement of the foreground ride transforming it from rickety swings to an immaculate top, kept moving people from being awkward semi-blurry-yet-recognizable blobs, and still retained stationary people, giving a kind of almost emo-mood to them as they seem to observe the world happening in a weird chronological isolation. So there's a lot going into the 3 second exposure, but a lot of times (not always) things work together in a way that makes it work out without leaving you wishing for different exposures for different elements.

No Auto focus: The front rim of this lens says "Auto" which is a reference to the aperture ring inside automatically closing to whatever you have it set to when you click the shutter, and there isn't a socket on the back for an AF attachment, so there is definitely no way to auto focus this lens, and furthermore, this may have been built before auto focus existed (speculation, I don't know that). However, the manual focus ring on it is very pleasant and is easy to manipulate precisely. From closest focus (.2 meters) to ∞ is about 720° of twisting, so you can't quickly shoot from focusing right in front of your face to focusing on the moon, but you can really dial it in and I can't describe how smoothly it turns. It feels as solid as a rock and as precise as brain surgery.

With F/2.5, this lens if fast and great for low light, but because it has to be manually focused, it's honestly very difficult to use for most applications. You wouldn't likely want to carry it around a party and use it to take no-flash shots of friends, you just can't accurately focus it easily enough. There are two good ways to focus this in a low-light situation that don't involve giving up and using a flash and slow aperture (so focus doesn't have to be precise), but they really limit what you can take a picture of.
1) The first way applies to taking a picture of something close. You will need a tripod to keep the distance between the camera and the subject constant. Then just take test shots and adjust the focus until your subject is sharp in them. It's dirty and it's slow, but it works. (Note: The auto focus boxes in your camera will still work so you can look for a solid green light (at least, that's how my Nikon tells me whatever is in the AF box is in focus), but at least with the D40, if your eye can't see whether or not it's sharp through the viewfinder, it's too dark for focus indicator also. A flashlight can get you around this problem, though, if you are able to shine it on your target while adjusting focus.)
2) The second way of focusing in low light is what I did here; it's very quick, easy, accurate and reliable. For this, the closest thing I really want to see is 20 feet away, and I want detail behind that also. So at 28mm, I will be fine just focusing to ∞. Because of manufacturers being wise enough to factor in thermal expansion, the lens doesn't hard-stop at ∞, but that's usually okay. The focus indicator will work if you put something bright and far away (for me, usually a street light) in the focus box. Just that dot of light and you turn until you get a solid green focus indicator light, or just until the dot of light you see is maximally sharp. This only takes a couple of seconds, and will allow you to shoot at F/2.5 in focus as long as you're shooting something far enough away.

In a way I don't think about composition at all and in a different way it consumes me. I mainly don't think of it in terms of composition, and I have no hard-and-fast rules besides avoiding "The Rule of 3rds" because someone else thought of it and I feel using it would be unoriginal. Here, I wanted to capture a ride in a way that really screamed "Carnival", and capture a fire cracker very clearly. The elements really highlight that it's a fair scene, and the setting shows you it's a small town. There's nothing to indicate what town, but it works out okay. So, besides wanting these elements, also important was getting enough of the ride to really see what it is, and I wanted there to be people in it, so no 30 second exposures here (no one stands still that long). I like important elements to be bright with the rest being dark, so I didn't want to do anything to brighten up shadows or generally less-bright foreground.
A big concern in a scene like this can be the street lights. High-pressure sodium lights are great; they are the most energy efficient bulbs available, crushing fluorescent; they don't care about weather; and they last longer than anything other than LEDs, so they're great for security, like street lights. However, the quality of light is abhorrent. They have a terrible "color rendering index" meaning the distribution of wave lengths doesn't make colors look right to our eyes, and the color temperature is very 'warm', aka yellowish orange. To fix this, I will usually either greatly 'cool' the color temperature in Ps or, if I don't want to settle, just do black and white. However, the street lights don't provide the light I'm really painting with here, the flash powder of the firecracker and the more pleasant colored ride lights do. (Note: Cooling the color temperature in this image does make it more visually appealing in a way, but at the cost of a lot of the mood.)

So that's pretty much it. The big things to overcome here are no auto-metering and no auto focus, but even with a CPU lens I would ignore the meter here, and with something like this, manually focusing is easy. Composition was pretty straight forward; the biggest obstacle there is getting the courage to stand in the middle of a crowd of cynical tweens with a camera and a tripod while they wonder 'what's this idiot doing?'.

Post Processing...
Just kidding. I literally opened the RAW in Adobe Photoshop Elements and saved as a jpg with quality of 12. Temperature is as shot, no hue, saturation, contrast, brightness, exposure etc adjustments, no cropping or noise reduction, this is 100% as shot. Everything is standard, normal, or as shot. For the record, I almost always do touching up to pictures, it just wouldn't have improved this one any.

Above average, B. It's not the most spectacular photo I've ever taken, but I like it. It incorporates a lot of techniques necessitated by the situation that I think I am good at and have developed on my own that are useful skills to have but that most people never really deal with. I don't think I could have exposed it any better; there is no noise since I got away with ISO 200 and it's sharp. I like the saturation although the colors look more interesting if cooled down a bit. I am really pleased with the composition; I could have gotten the rest of the foreground ride in the frame but I think it works better like this: you can tell the ride will be symmetrical from the light patterns being the result of rotation, but you infer the rest of the symmetry without actually seeing it, which would be distracting since there is no other symmetry in the image.

I like the fire cracker, but it would be nice if I had maybe started and stopped a second earlier, getting more of the early part of this firecracker and missing the shooting up for the next one which can be seen as a blue fishhook inside this one.

There is a tree attempting to occlude the bottom right portion of the firecracker and while it generally fails, it does interrupt the smoothness of the lines there, yet is too dark for it to be evident what is causing the disturbance there. I don't know what the ride to the left is, but the blur is kind of weird and it doesn't make a nice shape of lights the way the closer ride does. Fortunately, it kind of looks like a ghost train and with the ghosts under the near ride and the still people kind of looking like they are standing amidst ghosts, it works for me thematically.

I think the entire image is a tad crooked, but most of the light patterns are abstract enough and concrete objects so subdued that that doesn't hurt it. The flares of the street lights I like a lot here, they're beautiful. There's a kind of awkward partial house behind the near ride, but I said B, not A.

That pretty much wraps it up. I think it covers everything. Hopefully there is something to be learned from this by someone. I enjoyed writing it. Thanks for reading.

James Green

No comments: