One thing I've wanted to try out for a long time is the zone system. It's a system for establishing which exposure you should use that was invented primarily by Ansel Adams. It's been around for a long time, and I suppose it's lost a lot of it's relevance in this age of fancy matrix metering and, with the ability to check the histogram after taking a photo, you hardly need it at all. One of the things that appeals to me about using it though, is that it slows the picture taking process down, forcing you to consider what parts of the image are important, and how you'd like them to appear in the final print. Print? What's that? Do people still print photos?
I've kind of vaguely understood some of the concepts, but never really understood just how you apply the system to your photogaphy out in the field. There's a section in this book though that covers it quite well. It included a couple of key points that motivated me to try it out.
The system needs to be adapted a little to apply it to digital photography. Apparently digital only has five stops of exposure latitude, but I'm not sure how relevant that still is, and I don't know just how good my camera is in this regard. Also, I know that images taken in RAW mode have more usable exposure latitude than those taken in jpg. One of the points of trying the zone system out is that it involves recording the range of exposure in your scene, and in doing so you can then find out just how much exposure latitude your camera has. My landscape photography sure could do with a bit of work, so I see this as a good thing.
Anyway, my main motivation for going out for a walk last Sunday was to try out the zone system. I selected my scene, drew a little sketch of it in my notebook and decided on a good aperture to use to capture it (f/14). I set my camera to spot metering, aperture priority on f/14 and metered away on the main zones in the scene. These are the results:
I opted for an exposure time of 1/15s @f/14. My focal length was 17mm so I could've easily taken the shot hand-held with image stabilisation on, but I still used a tripod.
The histogram looked pretty good, with only a few blinking pixels in the sky, and no visible blip at the right hand end. With a bright overcast sky, a few blown highlights in the sky was what I expected with a correct exposure, so I was happy with the result. Having said that, it was a pretty easy scene, with not a lot of contrast on a afternoon like that. Also you can't really be 100% sure you've nailed it until you see the print - because the zone descriptions refer to the appearance of the zones in the print.
Here's the photo:
Not exactly a stunning landscape photo, but that wasn't really the point. The point was all about learning the technique.
Interesting technique. Unfortunately my ageing Canon 350D body doesn't have spot metering. :( Just thought I'd talk a bit about digital sensors. There are several limitations with them that are quite different to film.
ReplyDeleteThe first is the dynamic range, as you say, which is essentially the range of values that can be represented between the brightest and darkest. Film has a dynamic range of approximately 14 stops. Digital has a dynamic range of about 5.5 stops, as you say. So the overall range is much less than in film. There is a technique that can extend this range by something in the order of four stops, depending on the auto exposure bracketing feature on your camera body, called High Dynamic Range (HDR) photography. See here: http://tutorialblog.org/hdr-tutorials-roundup/
The second limitation is that while film degrades gracefully as it approaches saturation (the response curve of film trails off non-linearly, meaning it actually takes longer to fully saturate the film or "clip" the highlights), digital sensors are linear devices, meaning that they respond linearly right up to the point when they clip, which can result in some nasty things like colour casts in the sky.
The fact that they're linear devices has another implication as well: our eyes are non-linear, logarithmic devices, while the sensor is linear. So this is converted in the camera, leaving bigger gaps between dark values than light values. This means that as brightness increases, as we head towards the lighter end of the exposure range, we have more and more granularity in the information (another way to look at it is that there are more possible colour values in the lighter end of the dynamic range). This means that if you are trying not to clip the sky, using exposure compensation or underexposing the image, pushing everything left on your histogram, you are probably causing your lowlights to have very little detail (especially as you only really have 5 stops to play with). You're damned if you do, damned if you don't! :)
http://www.dpreview.com/learn/?/key=sensor+linearity
The third limitation, as you say, is the actual granularity of the digital format you use to store the image. While JPG provides 8 bits per colour channel, allowing 256 absolute values, RAW stores the full range recorded by the camera, often 12 bits per channel or higher. This means you have 4,096 discrete colour values per colour channel. Now, if you shoot in RAW, you have to convert back to JPG anyway to view the image on your computer, but the massive advantage you get is that if you do any kind of manipulation of the image like colour balancing, exposure compensation, white balance, or any of the other things, you do this in the 12 bit range, meaning that when you downsample it to JPG, you still have the full 256 colours per channel at your disposal. If you were to do the same thing to a JPG, you end up with what's called "posterisation", often seen as colour bands in skies and other large areas with subtle colour variation. You can see posterisation as a "comb like" histogram, which shows that some of the intervening colour values are not being used.
The advantages of using RAW really can't be overstated; I think it's one of the things that made a huge difference to my photography.
Here's an article on film vs digital: http://www.kenrockwell.com/tech/filmdig.htm
This is the RAW processing tool I use, it has a 15 day trial: http://www.bibblelabs.com/
Ok I'll stop now! ;)
Ha ha! Rob, your comment was longer than my post! My 400D has the same limitation as your 350D unfortunately, re: the spot metering. I actually used partial metering rather than spot, and just used the wrong term in my post. I wish my camera did have spot metering, I'm sure Canon just leave it out to give you another reason to upgrade to a 50D or similar.
ReplyDelete