Introduction
In this page you'll find a technical test. Please refer here for an introduction, details and testing method.
Available sensitivities:

• standard: from 100 to 25600 ISO with 1/3 EV steps
• extended: 50 ISO, 51200 ISO and 102400 ISO
  

Main results

• Gain. The measured gains follow quite accurately the power law. Exceptions are 50 ISO, as expected, that is exactly the same as 100 ISO (only JPG files are affected, RAW are absolutely identical). Another exception is 102400 ISO that is a little bit less sensitive than declared.
  
• Unity gain at 575 ISO
  
• Read noise. A constant decrease is found up to 6400 ISO. Using 12800 ISO or more produces no improvements on the read noise (and even worser, read noise appear to increase a little bit for 25600 ISO and above). So I advise not to use anything more than 6400 ISO, the only effect of using more is to reduce the dynamic range. Obviously this advice is valid only for RAW files, in JPG it is worth using higher sensitivities. For best results in low light circumstances mainly use 6400 ISO.
  
• Dynamic range. Considering a full well capacity of 13235 ADU, the approximate dynamic range is computed and reported both in the table and graph below. The best dynamic ranges are both at 100 ISO and 200 ISO; but also 400 ISO has a very similar dynamic range. Using 50 ISO does not improve in any way the dynamic range (in the RAW files). Above 400 ISO the dynamic range start to decrease noticeably. For best results on bright subjects, use 400 ISO or less.
• 1/3 stop sensitivities. Main result: absolutely don't use them! In example 1000 ISO has the same read noise of 800 ISO, but lower dynamic range, so it is better to use 800 ISO. And 1250 ISO has the same read noise of 1600 ISO and also the same dynamic range (because saturation is at a lower level, 11177 ADU), so there is no advantage to use 1250 respect 1600 ISO. As a general rule, simply avoid 1/3 stop ISOs!
  

True sensitivities
To distinguish between true and software interpolated sensitivities, the best method is to look at the histogram of a nearly uniform image (like a flat field). The result is shown in the figure below, and the conclusion is:

• 51200 ISO and 102400 ISO are not true sensitivities, as expected.
  
• 1000 and 1250 ISO (intermediate 1/3 stops sensitivities): as explained above, they do not produce anything better that full stop ISOs. So I advise to avoid them. Also looking at the histogram of the 1250 ISO image, something strange appear.
  
• 100, 200, 400, 800, 1600, 3200, 6400, 12800, 25600 ISO are true.

Gain and read noise