What I learned from Gollywop -- and what I wonder

Started Mar 26, 2013 | Discussions thread
Detail Man
Detail Man Forum Pro • Posts: 16,975
Re: GH2 Information

Anders W wrote:

Detail Man wrote:

Anders W wrote:

Detail Man wrote:

Anders W wrote:

You are largely (though not exactly) right about the GH2 whose only "real" ISOs are 160, 320, 640, and 800.

I don't think that it is clear that ISO=800 on the GH2 is unmanipulated. See bg2b's OP here:

Could be that you are right about that. Logically, it seems that the series of "real" ISOs should end at 640 rather than 800.  But if so there appear to be some unresolved mysteries here. I remember asking Ken W about this in a prior thread and he found the ISO 800 distribution to be perfectly continuous, without the gaps we would expect as a result of digital scaling. On the other hand, he didn't find that going from 640 to 800 had any tangible benefits in terms of reduced read noise. See here:


Less than 0.1 EV (net) difference. Here are some RawDigger histograms of GH2 dark-shots that I recorded. The vertical Y-axis display scaling was adjusted to be in direct proportion to the (rated) ISO value for viewing. Though they appear a bit "rough" to the eye, it seems possible to get something of a feel from observing them about how much digital processing manipulation(s) may be going on. ISO=800 does perhaps appear as being the closest to the ISO=640, ISO=320, and ISO=160 series:

I don't think those (partial) distributions do much to resolve the issue of what goes on at ISO 800 relative to ISO 160, 320, and 640. When the expected number of hits in each bin is very low, as it is here towards the right edge of the distribution, you should expect gaps as a result of mere chance. You need to look at a distribution such that you'd normally expect plenty of hits in each bin. If, under such circumstances, you nevertheless find some bins empty, that's a sign of digital scaling.

I am sure that's what Ken did at ISO 800 and didn't find any empty bins where they would have been expected if digital scaling were involved.

Really ? Do you have a link to that information from Ken ? I recall asking him at one point whether he had done that, mentioning bg2b's statement about ISO=800 (quoted below), and my recollection is that Ken communicated to me that he had not done that. After that, he switched to an E-M5.

Real ISOs are probably just 160, 320, and 640. 200, 400, and 800 show odd artifacts (gaps or near gaps) in the RAW histograms. 250, 500, and 1000 show them even more. Everything beyond ISO 1000 is completely artificial; they're obviously just digital scalings of lower ISOs.


This is kenw's RMS Read/Dark Noise data:

1/4000 NR-2:

160 0.73554
200 1.117
250 1.5487
320 1.3875
400 1.8514
500 2.4756
640 2.6903
800 3.513
1000 5.4109
1250 5.7917
1600 7.9696
2000 13.557
2500 13.168
3200 16.576
4000 25.919
5000 28.305
6400 35.017
8000 40.398
10000 54.246
12800 68.36


I calculated the number of EV that certain ISOs (from rated ISO=160 through ISO=800) deviate from a straight line projection (in EV) from that which would be derived from multiplying the Read/Dark Noise data by the ratio of the Saturation ISO divded by Saturation ISO=167.

Here are results using DxOMark Saturation ISO (extrapolated SatISO values for ISO=320 and 640):

Saturation ISO -------- Read/Dark Noise deviation in EV

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow