Building Sekonic Profile . . . Weird Data?

I'm building an advanced Sekonic (478D) profile for my 5D3 with 21mm Zeiss Distagon. Things are looking off to me.

Here is ISO 200:

I would expect the dynamic range to decrease as ISO increases. How did you run the test to get your exposures (fixed aperture or shutter speed, target, etc.)? The following graph is with the EF 24-105 f/4.0 @80mm. Aperture was fixed at f/8.0 and I used 5 bracketed exposures. The target was a color checker passport. The default dynamic range and clipping levels are fairly conservative. I usually bump those up, but it still leaves plenty of wiggle room. I didn't run a high ISO profile for comparison. If I have time, I'll run one this weekend.

I spoke to someone locally who had a lot to say about this. But, he confirmed that as ISO increases dynamic range increases. There are, of course, things that deteriorate at higher ISO.

I did my profiles we three bracketed shots, in two different ambient lighting conditions (daylight and incandescent in my kitchen), at three different ISOs (200, 6400, and 12,800). I used the ColorChecker Passport. While doing each set I kept aperture and ISO the same (outdoors was f8, indoors was f2.8), and changed the shutter speed to get the brackets. I did use a tripod. In total I did six bracketed sets.

I'll be adding 5 shot sets to the profile. I'll see if DR increases with those for each ISO (which I suspect it might).

I did watch a Sekonic YouTube video where the demonstrator indicated that he felt the clipping points were too conservative and he bumped them up. But, he didn't indicate what he adjusted them by.

You might be interested in looking at the data at the photonstophotons site or at DxoMark. Both show that dynamic range decreases as ISO increases.

http://www.photonstoph​otos.net …n%20EOS%205D%20​Mark%20III

http://www.dxomark.com …easurements#mea​suretabs-3

Keep in mind that the main purpose of the data transfer software is to calibrate your light meter to your camera. I think either of the above sites will give you a better indication of the sensor's dynamic range capability.

When creating a meter profile, I usually set the highlight dynamic range level to 250 and the clipping point to 245 and set the shadow dynamic range level to 15 and the clipping point to 20. Canon sensors perform well when exposing to the right (highlights) and less so with shadows, so the shadow settings are a bit more conservative.

I had a look at your links.

There was something that I had read somewhere (I can't find it again, I'll post it if I do), that indicated while the DR reading would increase, there was something that was increasing at the same time that should be factored out of the final value, and thus it was really a decrease in effective DR. I can not remember what the thing that should be factored out was.

I typically ETTR, but I've been playing around with not pushing exposure as far lately when dealing with skies and looking at the other end of the histogram more closely.