Well, I think I have the technique down now. This is at least as good technically as any 4×5 I ever shot.
It’s an HDR composite of 63 exposures on the 5D, three rows of seven frames at -2, 0 and +2EV. Stitched in PTGui, tonemapped in CS3. The full resolution image is about 150 MP or so. I uploaded a fairly big version so that anyone who’s interested can check it out. View large or huge (>3000 pixels).
A few notes: making these images is fairly labour intensive. It’s certainly not an ‘easy alternative’ to 4×5. In fact, in terms of difficulty, it’s a wash. There are a billion things that can go wrong when you’re shooting 4×5 and the same is true here. You have to learn a ton of technical stuff to shoot and process 4×5 properly, and the same here.
If you want to shoot RAW, you’re using about a gigabyte of storage for every composite. That’s eight shots if you carry an 8 gig card into the field. Practically speaking, you don’t really lose any quality by shooting big jpegs because of the huge dynamic range BUT you lose the ability to forego HDR (if movement ruins the shot, say) and pull the RAWs from the 0EV exposure. You just have to pray your JPEGs are good enough.
A few more tech details. Camera shake is a potential problem. I am using a Nodal Ninja which is borderline for the 5D at low shutter speeds. Anything under 1/60 I stead it with my hand. Under 1/30 I use mirror lockup. I always use a cable release.
Another wrinkle is the speed of the camera when writing to the card. With jpegs, you’re usually okay. The camera dumps the images about as fast as you shoot them. But with RAWs there’s no way the camera can keep up. So you end up waiting with your finger on the cable release as the camera churns through the -2, 0, 2 EV brackets. It’s boring.
Batteries are something of an issue. I left the kit in my car one night and the next day went out to shoot in -10C weather. The camera wouldn’t even turn on. I guess I could have warmed up the batteries next to my body but instead I went home.
Finally, it’s not obvious, but I’m not framing these in any conventional sense. I’m just picking a spot, trying to figure out where the center of the picture is, and framing in my head. My standard 7 horizontal x 3 vertical shot matrix (15 degree increments horizontally, 20 vertically, with the camera vertical) is designed to capture as close as I can to a 90mm on 4×5’s field of view with a bit of wiggle room for slight reframing. I’ve thought about taking a 4×5 finder into the field. I will probably do that from now on.
Having said all that, there are some real advantages over 4×5. The kit is massively easier to carry around. You can postpone a lot of editorial judgements (BW or color? Filtration? Exposure?) until you are sitting at your computer. I don’t have to fight SCSI to get my 4×5 scanner going. I don’t have to do dust spotting. I don’t have to buy or process film.
And, maybe most important, I can shoot for different formats in the field. If it looks like a 4×5, I can shoot a 4×5. But if it looks square I can shoot it square. If it looks like a Noblex shot, I can shoot that. If it’s a 360, I can shoot that. For a camera whore like me that’s nice.
Posted by John Brownlow on 2008-04-07 13:40:45
Tagged: , indian , brook , thornbury , ontario , canada , tree , trees , forest , wood , woods , wild , tangle , bramble , blackandwhite , bw , monochrome , 400 , stitch , stitched , panorama , composite , ptgui , HDR , tonemapped , cs3 , pinkheadedbug , bestviewedlarge , viewbigplease