[ratpack] Lens flare

From: Ray Buck <rbuck@xxxxxxxxxxxx>
To: ratpack@xxxxxxxxxxxxx
Date: Mon, 22 Mar 2010 15:23:19 -0600

I've been shooting indoor car shows for the last 3 weeks and I've found some interesting stuff. Well, it's interesting to me cuz it messes with my photos.

I'd swear that I'd seen something about this on Ken Rockwell's site but can't find it now.

The issue is not just a single or even a double flare in a backlit subject, but a ton of flares that show up when using small apertures, especially on long exposures. Last Friday I shot the "Chrome in the Dome" car show in Pocatello, Id. Almost every shot was done on a tripod and all were done with my Canon 7D and the 18-200mm lens. When on the tripod, I used a cable shutter release to prevent camera shake from pressing the shutter button.

I did this stuff because I wanted to blur any moving spectators, car owners, untamed rug rats, etc. from the photos. The exposures were anywhere from 3 to 15 seconds, generally with an ISO of 100 and an aperture of f22. That allowed me to keep the "shutter open" (I dunno how much of this is actual mechanical timing and how much is electronically controlled) for a sufficiently long time that a person walking through the frame might only be seen as a slight blur. This photo:

was a 15 second exposure at f22, ISO 100. It's a bit overexposed. But look at the flares in it, particularly behind the B pillar (above the back door handle) which I'd seen before and thought it was a smudge on the lens.

When I applied a little more editing, this was the result:

The exposure's much closer but the flares are still there. A friend suggested that this is probably the result of scattered light refraction in the lens. Digging back into the caverns of memory, I thought I remembered Rockwell writing about this, but stating that it only happened with extremely small apertures like f32 and smaller.

Btw, there was a guy kinda watching over my shoulder as I shot this and he saw 2 people walk right between me (and camera) and the car. I said "watch what happens." Then when the 15 seconds was up, I showed him the result in the LCD. He said, "what are you? Some kind of magician?" I explained that I had the lens stopped down and the sensor sensitivity turned down with the ISO setting so that for the small percentage of the total exposure time that the intruder was in the frame didn't register in the image.

But it was when I got home that I saw these flares and was really puzzled by 'em. They weren't in any of the shots I'd made with the flash, and they seemed to decrease as the lens opened up. The worst of the flaring at f22 and the minimum at f9. Ok. I could understand that this might be what I'd read about and it was just showing up on the long exposures.

Then I went back to the first shoot of the year (the Auto Expo) that was done while the 18-200 lens was at the Canon repair facility. I'd used the Sigma 17-70 lens...and there was virtually zero flare in it, no matter what aperture was used.

Here's one at f22, 10 seconds, ISO 100 from the 7D with the Sigma lens:

No editing other than cropping and resizing...and the only flare is in the windshield and I knew that one was going to be there.

Now I'm real confused cuz it seems that the problem might be lens-dependent as well as aperture-dependent. And the 18-200 is about twice the price of the 17-70 plus it has IS and I'd hate to find out that in addition to the lens lock issue (I haven't used it at all since I got it back out of protective custody) it's prone to these flares at apertures smaller than f9. I guess I need more testing, next with the Canon 28-135.

Back to editing.

RtR

-- 
Sent from my Dreadnought using that barely tolerable Thunderbird email program

[ratpack] Lens flare

Other related posts: