Please recheck your Low Light ISO Result for A99

[lightdreamer]

This a comment for this page on the website

I have compared the A99 against the A900 and the A99 is for sure 1 EV better in this regard. I made my comparisons in M-Mode and used dcraw without any sharpening or noise reduction. The A99 should reach the 2000+ level easily.

Regards

[a odd guy]

I can't understand what you mean.

All of DxOMark score is based on RAW data comparison.

It doesn't have relevance to sharpening and noise reduction.

[Revenant]

DXOMarks definition of the Low-light score: "Thus low-light ISO is the highest ISO setting for a camera that allows it to achieve an SNR of 30dB while keeping a good dynamic range of 9 EVs and a color depth of 18bits."

In other words, it's not just about the visible noise in the image, but also about DR and color depth. When the SNR goes below 30dB, or DR goes below 9 EV, or color depth goes below 18 bits, whichever happens first, then the image quality is no longer excellent according to DXO's definition. Obviously, at least one of these three things happens at a much lower ISO on the A99 than on its full-frame rivals, which explains the low-light score.

[Bobo_SAN]

Yes it will, and I see it as well and may very well be easily compared because the database studio samples (http://www.dpreview.com) with other cameras that not are SLT.
DXOmark SLT disadvantage precisely because it leads to a slight loss of light and measurement settings that the DXO are performed against a very tight although the differences are micro small but measurable laboratory.

Low-Light-ISO is actually somewhere in the 1950's až 2050. It can be deduced from the databases studio samples (http://www.dpreview.com) in comparison to other cameras that Low-Light ISO-2000 were around.

For example:
Nikon D3X has a low-light-iso 1992 in stiodio but the results are compared to the 3200 and 6400 very much alike. The ISO 3200 seems to A99 x places better and D3X one in a watch or I would even say that the A99 has ISO a tad better.

Second example:
Nikon wrote in 7100 that the ISO is very similar to the A99 but it is not true, A99 is hell somewhere else but in 7100 measured in 1256 Low-light-A99 and ISO 1550 but realistically it is somewhere around 2000, and why it's so very much to know in comparative study of still pictures at ISO 3200/6400 in the raw Nikon 7100 vs Sony A99 (large differences in favor of A99).

DXOmark SLT can not properly assess the reality is somewhere else then.

[kevinli]

There's an obvious misunderstanding about DXO's low-light ISO measurement, which is based on the measured ISO, not the manufacturer's ISO number.

If you check the ISO sensitivity data of A99, you can find that A99's ISO 1600 is actually just measured ISO 913, and ISO 3200 is measured ISO 1825. I guess Sony did this on purpose to get a better looking ISO number.

That is to say, the DXO low-light ISO score 1555 corresponds almost ISO 3200 on the camera. This last number (ISO 3200) might match most people's experience of A99's acceptable ISO setting.

[Emilie_DxOMark]

Hello!

Thanks for your interest in DxOMark.

The DxOMark sensor score only evaluates RAW image quality on the sensor noise point of view.
The comparison is a JPEG comparison, with a raw convertor that could have slight different behavior between the different cameras.
The ISO lowlight score of the SLT A99 is pretty good and is a third stop better than the last Nikon D7100.

The DxOMark team