There’s a debate that occasionally raises its head in kite aerial photography circles: to shoot RAW or not to shoot RAW. Like so many technical debates, this tends to split people into two distinct camps: those who have real reasons to shoot RAW, and those who have real reasons not to. Often this depends more on the kind of KAP the person does than on what is considered “right” in other photographic circles.
The most apparent deciding factor is whether the person is doing remote controlled KAP or autoKAP. With RC KAP, the KAPer makes a conscious decision to trip the shutter on the camera. With autoKAP, the rig or the camera typically runs an intervalometer that trips the shutter every X seconds. RC KAP sessions may result in anything from a handful of photos to a few hundred. AutoKAP can often result in thousands of photos from a single flight, depending on how fast the intervalometer is running and how long the camera is airborne.
All of which has very little to do with RAW and JPG unless you think about the implication on file storage. For my T2i, a typical JPG is 6MB. A typical CR2 RAW file is 24MB. A hundred frames is only 600MB using JPG. That same session is 2.4GB using RAW. I typically do RAW+JPG, so every hundred frames, I’m looking at 3GB of storage. A lot, but not unmanageable. Not when doing RC KAP, in which a hundred frames is a pretty big haul.
However, if I put that same camera in my panoramic rig running at one frame every one and a half seconds (about how fast it goes) I run out of space on a 32GB card in just under half an hour shooting RAW+JPG. You can start to see the problem here. It gets compounded when you actually start processing the photos. A one hour autoKAP session may run upwards of 60+GB of disk space, and unless you do RAW+JPG there’s no way to proof the photos before committing to processing them. Most RAW workflows get unmanageable once you have that kind of volume going through them.
All of which makes the RAW format look pretty unattractive when doing autoKAP. But what about RC KAP, in which the photographer is intentionally tripping the shutter for each frame? Even then, not everyone sees the need to use RAW. In this case it has more to do with subject than anything else.
Years ago when I was doing large format film photography, I read Ansel Adams’s Technical Series and fell in love with his Zone System. I was attempting to do interior architectural photography at the time. Attempting, but failing most of the time. There were simply too many stops of dynamic range in the subject for me to be able to photograph it with any film I had available to me. Color transparency film only has about five stops of usable range. Color negative has about seven. Black and white negative film typically has ten stops of range, and was the film I used most of the time. But sometimes my subjects – rooms lit by sunlight streaming in through an open window, coupled with dark shadows – presented eleven, twelve, or more stops of dynamic range, all of which I wanted to represent on film.
The Zone System is a method of tuning your developing to expand or contract the tonal range of the film. With some care you can get twelve stops of tonal range out of a black and white negative, and capture detail in highlights and shadows that would’ve been lost with normal processing. Likewise you can compress the tonal range of the film so it would just cover a less contrasty scene, offering more choices when printing the picture.
A JPG file has, by definition, eight stops of range available to it. Each stop represents a doubling of the light being sensed by the film. With digital files, each bit also represents a doubling of the light. Eight bit file = eight stops of range. If your scene contains more than eight stops of tonal range, you’re up a creek. You can push or pull-process film, but you can’t pull or push-process a digital file. Or can you?
It turns out you can. This is what RAW is good for. My T2i makes 14-bit RAW files. 14 bits of data means 14 stops of range. During processing, I can choose what part of those 14 stops are represented by the 8 stops of range in the output JPG. If the scene isn’t very contrasty, I can stretch five or six stops of data to fill the 8 stops I have available. On a more contrasty scene I can do the opposite, and compress the data to fit into the 8 stops of the JPG format.
This is similar to what a photographer does in a positive darkroom. When the negative is made, they use pull or push processing so that the scene they photographed will fill all ten stops of range the film can handle. \When that negative is then printed, they select a paper that will give them precisely the range they need to represent all of the tonal information from the negative on the final print. In this analogy a RAW is like a negative, and a JPG is like a final print.
Not every subject requires this level of processing. If they did, the Polaroid camera never would’ve taken off the way it did. Likewise, for some subjects the in-camera JPG is fine. The benefits in terms of storage, transfer time, ease of processing, etc. means that for some KAPers the JPG is the file format of choice.
But some subjects really do benefit from the additional processing. Here in Hawaii, I routinely photograph a subject that pushes the JPG format past its limits: rocky shorelines. The basalt rock that makes up much of the coast on the Big Island is extremely dark. Foaming surf, by contrast, is extremely bright. In direct sunlight the two present a scene that is several stops wider in tonality than a JPG can handle. The rocks become a muddy blur of black, and the brightest parts of the surf become a washed out white. I’ve lost entire KAP sessions to this effect. For years I didn’t even bother flying at Laupahoehoe Point because the resulting JPG images looked so bad.
It was primarily for this reason that I started using a DSLR on my KAP rig. And it’s for this reason that I use RAW+JPG whenever that camera is flying. Below is a pair of photos from a recent KAP session. The upper one is the in-camera JPG. The lower one was processed from the RAW file of the same exposure.
In the processed RAW image the rocks are slightly darker by my own choice. I could’ve “printed” them at the same tonality as the in-camera JPG, but working from a RAW file gave me the freedom to represent them the way I saw them with my eye .
But the real difference lies in the surf and the driftwood. In the in-camera JPG, all detail is lost. The RAW file showed texture in those parts of the photo, so I tweaked the tonal range during processing so the highlights were better represented in the final JPG.
One criticism I’ve heard of RAW workflow is that it’s difficult to mass-process an entire KAP session worth of photos. But I haven’t seen this as a problem any more than I saw making individual black and white art prints as a problem back in my days of film. After choosing which negatives to work with, I loaded each one into the enlarger and made a test print. Then I went back and decided which paper I wanted to use for each, whether I needed contrast filters or not, whether I wanted to dodge any areas of the print or burn other areas in a little more. I had as much opportunity for artistic expression when printing my negatives as I did when I exposed them in the first place. A picture worth printing was a picture worth printing individually.
To me, working with RAW files is a similar process to printing a negative. I make RAW and JPG files when I’m out doing photography, and use the JPG files as a proof sheet to decide which photographs get further treatment. I’ll load the corresponding RAW files into my software, and then the fun begins. In the end I get a set of JPG files that are individually white balanced, individually adjusted to make the most of the JPG’s eight stops of tonal range, and individually processed to best represent what I saw when I was standing there.
It’s not for everyone. And it’s not for every form of KAP. But I appreciate what RAW workflow has given me when it comes to kite aerial photography. Despite the weight of the camera, despite the added storage requirements and complexity of processing, it’s worth it.