Nex-7 vs EM5 Shutter Speed comparision at ISO 200

Started Apr 3, 2013 | Discussions thread
Detail Man
Detail Man Forum Pro • Posts: 16,707
Re: Nex-7 vs EM5 Shutter Speed comparision at ISO 200

LTZ470 wrote:

Let me post some crops of the same two to see which is brighter...

Hello, my friend. I converted both of the above crops to monochrome in order to eliminate the effect of the differences in color temperature reference that I could see in the RGB histograms:





These two images look pretty close to my eyes. The mean (average) Greyscale is 0.054 EV higher in the NEX-7 shot than in the E-M5 shot. I have located common areas in both images that have the same sRGB tone-level of 118 which is used to specify ISO Sensitivity (when using the SOS method).

Also downloaded the original crops posted on Smugmug, and the results are essentially identical:

E-M5 Crop:!i=2439533361&k=GqW9VN4&lb=1&s=O

NEX-7 Crop:!i=2439533387&k=hCxMjf3&lb=1&s=O


Regarding the different Shutter Speeds (E-M5=30, NEX-7=40) leading to approximately the same result, note that the two cameras have different sRGB JPG tone-curve transfer-functions.

Have a look at the top graph (titled Cameras Compared, Dynamic range comparison) here:

The NEX-7 plot already exists (Green, with "DRO Off", as appropriate). Select the E-M5 for one of the other tone-curve transfer-function plot (set to to "Gradation Auto", as appropriate), and select "No Camera" in order to disable the other two plots so just the E-M5 and NEX-7 plots are displayed.

The vertical scale-line labelled "0" shows you where middle grey (sRGB=118) exists. Project a horizontal line from where the NEX-7 tone-curve transfer-function intersects that vertical line. That horizontal line represents a middle grey output level (sRGB=118). Note that the E-M5 tone-curve transfer-function intersects that horizontal line at an input illumination level that is around 2/3 EV lower than is the case with the NEX-7 tone-curve transfer-function.

This would indicate that the E-M5 Shutter Speed might be around 2/3 stop slower in order to produce the same JPG sRGB output levels. In your comparison above, the cameras metered so that the E-M5 Shutter Speed was 1/3 stop slower produced approximately the same levels.

The image-file meta-data of the original crops lists the ISO specification method of the E-M5 to be SOS, and the ISO specification method of the NEX-7 to be REI (which is not constrained to the SOS procedure's use of an output of sRGB=118 for a given scene luminance level).

I looked in the DPReview NEX-7 review for a statement regarding how their own measurement of ISO Sensitivity (using a SOS-related procedure), but found no such statement in the review. The reason for that may well be because Sony uses REI (not SOS) to specify the NEX-7 ISO Sensitivity.

Due to the use of different ISO Sensitivity rating standards used for the two cameras (where REI is quite arbitrary in its requirements), it is not necessarily surprising that the Shutter Speeds that you found differed from the 2/3 stop slower that appears to be expected from the DPReview published JPG tone-curve transfer-functions for the E-M5 (and were 1/3 stop slower, as you found).


Note that the DxOMark "Saturation ISOs" that are the common source of the "cheating on ISO" theories are RAW-referenced test results that apply only to RAW recording when both camera's image-sensors are illuminated to maximum linear sensor saturation levels. Those specifications have nothing to do with JPG-referenced metering and JPG recording, and what you are measuring in your tests.

Best Regards,

DM ...

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