OM-1 HR Mode analysed for PDR at Photons to Photos

[Lester]

Bill Claff has just published a PDR graph for the 80M tripod HR Mode of the OM-1.

https://www.photonstophotos.net/Charts/PDR_HighResShotMode.htm

We can see 2 to 3 stops improvement in PDR compared to normal OM-1 20M non-HR snaps.

When we overlay the E-M1X HR and the E-M1ii HR plots, we see something interesting. The OM-1 HR is the same as the E-M1X HR from ISO 200 to 800, then it jumps a touch and is the same as the E-M1ii HR from ISO 800 to 1600. Huh.

[gazza95]

Hi Lester

Two weeks ago I would have found those results interesting, today having processed many OM-1 colour checker chart images I find it much less so. Let me try to explain.

As I understand Bills definition of Photographic dynamic range it is the difference between a peak value and the top of the low level noise at a photo site as read from a raw file. He also factors in circle of confusion and photo site size. This is a very valid scientific measurement. The issue I have with this measurement in relation to actual photo's is that we actually apply noise reduction in some form to get to jpegs. As ISO increase so does the amour of noise reduction we apply.


This difference was brought home to me while looking at 1024,000 ISO images. These should have only one stop of dynamic range according to Bill, but in reality I see 5+ stops in a jpeg once a noise reduction is applied.

So in the case of a 1024,000 iso image, Bill would calculate dynamic range on this.




Yet we as photographers would be more interested in this Noise reduced version




I appreciate this is an extreme case but it helped me clarify my thoughts on dynamic range.

With my EM-1x I would shoot to 1600 asa and not worry about noise reduction. For the OM-1 I now know Noise reduction will be a part of my processing and from the tests I have done noise reduction will show significantly more detail. Two weeks ago I was concerned about dynamic range at iso 6400, from Bills charts. Now I know dynamic range is not an issue for the type of images I capture.

When you capture images with HR, ND, or some form of bracketing you are using techniques that reduce noise in some way.


Gary

[SrMi]

gazza95 You cannot use PDR data to judge the JPG output. It is applicable only to raw files and when no noise reduction is applied.

[Ian]

Hi Lester

Two weeks ago I would have found those results interesting, today having processed many OM-1 colour checker chart images I find it much less so. Let me try to explain.

As I understand Bills definition of Photographic dynamic range it is the difference between a peak value and the top of the low level noise at a photo site as read from a raw file. He also factors in circle of confusion and photo site size. This is a very valid scientific measurement. The issue I have with this measurement in relation to actual photo's is that we actually apply noise reduction in some form to get to jpegs. As ISO increase so does the amour of noise reduction we apply.


This difference was brought home to me while looking at 1024,000 ISO images. These should have only one stop of dynamic range according to Bill, but in reality I see 5+ stops in a jpeg once a noise reduction is applied.

So in the case of a 1024,000 iso image, Bill would calculate dynamic range on this.




Yet we as photographers would be more interested in this Noise reduced version




I appreciate this is an extreme case but it helped me clarify my thoughts on dynamic range.

With my EM-1x I would shoot to 1600 asa and not worry about noise reduction. For the OM-1 I now know Noise reduction will be a part of my processing and from the tests I have done noise reduction will show significantly more detail. Two weeks ago I was concerned about dynamic range at iso 6400, from Bills charts. Now I know dynamic range is not an issue for the type of images I capture.

When you capture images with HR, ND, or some form of bracketing you are using techniques that reduce noise in some way.


Gary

This agrees with my thinking - there is a technical evaluation of the sensor output, often mathematically defined, and there is what the photographer sees. The former has been important, of course, but it's not the be all and end all.

Ian

[Lester]

Hi Gary

As you say, an ISO 102,400 image does a little better than literally 1 stop of DR, so I read Bill's graph, and pretty much all other such graphs, as mathematical abstractions. I find their practical use is for comparing an exactly defined apple with with another such apple despite never wanting or intending to actually eat one in the way defined. So I find it helpful to know that the 80M HR is, in everyday use and other things being equal, likely to be 3 ISO stops "better" than the 20M at a given ISO, or probably more usefully, that the HR at ISO 1600 is "equivalent" to the regular at ISO 200.

[SrMi]

Bill Claff has just published a PDR graph for the 80M tripod HR Mode of the OM-1.

https://www.photonstophotos.net/Charts/PDR_HighResShotMode.htm

We can see 2 to 3 stops improvement in PDR compared to normal OM-1 20M non-HR snaps.

When we overlay the E-M1X HR and the E-M1ii HR plots, we see something interesting. The OM-1 HR is the same as the E-M1X HR from ISO 200 to 800, then it jumps a touch and is the same as the E-M1ii HR from ISO 800 to 1600. Huh.

The jump in the OM-1 curve is caused by dual conversion gain, which occurs at ISO 800. On M1x, it kicks in at ISO 1600.