Sony A7S : Tests and Reviews

Hi adeadetayo,
Thanks for you feedback and for the kind words. Dynamic Range measurement depends on threshold definition. You'll find all the information you need on our protocol detailed here: http://www.dxomark.com/About/In-depth-measurements/Measurements/Noise
Sorry but we cannot make any comment on other measurements as we do not have protocols and definition.

Sincerely

Hi,

Thank you for your interest, and our apologies for the late reply. When available, we systematically use mechanical shutter to perform our analysis.

Best regards

Not a fan of these sharp edged little Sony boxes... but in looking at the measurement numbers (please look there instead of just the categorized numbers) this sensor achieves significantly higher numbers across its entire operating range in every tested area.. What were your expectations that were not realized??

Good Spec i like it.

@SiliconVoid

Well, given the v. low pixel count, I was waiting for a much higher ISO result. For instance, with only half the resolution, I was waiting for a full stop better performance in low light. Or at least half a stop. For instance, against the 24MP and ~3000 ISO score of the D600, I was hoping to be between 4500-5000 with the 12MP 7s.

But maybe that's just my ignorance of sensors' physics.

Also the samples I had seen had an amazing dynamic range. But it seems it was only commercials.

In conclusion, given that the pixel count has no direct impact on the DXO score (if I’m not mistaken), I was hoping that this camera woud break the 100 result after the Red Epic Dragon.

Well I strongly suggest basing your evaluation at the pixel level (the [Screen] option when viewing the measurements tab) as it is becoming obvious that DxO does not actually take measurements from a 'normalized' image, but merely calculates the 'projected' downsample from the pixel level measurements. Which is understandable as other variables would be introduced affecting consistency and accuracy dependent on what software or printer was used.

The reason I make that assertion is because there are some aspects of analog to digital conversion, and the eventual oversampling and averaging that occur during downsampling/printing, that will not result in a physical improvement of the data. Noise levels are the only aspect of DxOs measurements that actually improve when an image is downsampled - not only less perceivable, but due to oversampling and averaging of surrounding pixels the noise levels improve because data is effectively replaced. To have your calculations return a result that is higher than the baseline value in other areas should raise a flag and make one realize a flaw in the applied calculations. The output in the other areas of measurement like color, tone, DR, and SNR do not physically improve when downsampled.

If a sensor is only able to capture say 22bit color at the sensor level (pixel level) it would be impossible for it to produce 24bit color at any point afterwards regardless of print/downsample size as the missing bits of color were not captured to begin with. Same thing with tonal accuracy, DR, and SNR - as they are all physical characteristics and limitations of the sensor affecting the baseline data. If that data was not captured to begin with it is not possible to produce it later unless being manipulated by the manufacturer during ASIC processing - which is no longer a measure of what the sensor is capable of, but what the manufacturer is able to 'simulate' - and no longer holds any empirical value in comparing the 'sensors' performance.

If you review the results provided for the sensor level performance, and take into consideration the arbitrarily selected 'subjective' criteria DxO considers 'acceptable' (30db, 9DR, 18bit color) you will see that the A7S far exceeds the performance of the D600, and anything else with higher mp in the higher ISO ranges at the sensor level. Which means that when you do 'physically' downsample or print (as opposed to speculative calculations) you will see better output than shown in DxOs evaluation.

Honestly, at this point I cannot even say what value DxO provides in evaluating camera sensor performance unless your targeted output is exactly the arbitrary parameters DxO 'measures' (speculates by calculation), and even then it would be a scenario where one is more interested in speculative performance than physical performance.

...and while DxO does not award any points for pixel count, their 'calculated' downsample is significantly effected by pixel count. So whether any physical benefits are provided by higher and higher pixel counts, speculative benefits are.

What DxO should do, if they are going to evaluate the camera sensors at anything other than pixel level, is figure out what percentage of downsample no longer yields any measurable benefit. For example if the D800 no longer shows any improvements beyond say 8mp, and that downsample represent a ~78% reduction/compression of the image data, then all cameras tested should be measured at a ~78% reduction of their image output, regardless of what physical print size is produced.

Hi François, hi all,
The lowlight ISO score is primarily relative to the sensor size, and to a lesser extent on the pixel pitch. Pixel pitch could be important if the percentage of effective surface were smaller on a smaller pixel pitch, but it is not exactly the case here. To sum it up, the Sony A7s results are very good indeed.

Best regards

I'm curious about it as well

Really confused too.

I confirmed with Sony that the new firmware version supporting 15.3 stops in raw is the launch version, v1.0.  Sony said that some pre-release cameras may have had older firmware versions.

The DxOMark review didn't say which firmware version was tested.

I also learned something about the way DxO measures dynamic range.  DxO's dynamic range is, for one pixel, the ratio of the saturation illuminance at the sensor to the darkest iluminance at the sensor with an output SNR greater than 1 dB.  But then they multiply that ratio by the square root of the camera's pixel count divided by 8 megapixels.  It's a way of converting a per-pixel measurement to a per-area measurement, normalizing all of the measurements to the same effective resolution and removing the noise penalty of a high megapixel sensor.  The extra factor is the theoretical increase in dynamic range that you'd get from the noise reduction that comes from downsampling the picture to 8 megapixels, *if* the low end of the dynamic range were noise limited, and not limited by the luminance encoding scale.  It also assumes that the image is uniform and the noise is Gaussian.  So high megapixel cameras get an artificial boost in their dynamic range measurement that may or may not correspond to the the actual dynamic range of a picture downsampled to 8 megapixels. 

Let's not forget that dynamic range is a real thing that can be measured per-pixel, and it has a real maximum and a real minimum.  Just as there are high sensor illuminance levels above the saturation point (clipping), there are small illuminance levels that cannot be distinguished from total darkness in the camera's raw output, even if you ignore the noise. Those small illuminances are outside of the camera's dynamic range. 

A 14-bit linear raw scale, for example, can never express more than 14 stops of dynamic range, even if you remove all of the noise.  But there are 14-bit linear raw cameras with DxO measurements of dynamic range greater than 14 stops.

thanks,but if "A 14-bit linear raw scale, for example, can never express more than 14 stops of dynamic range, even if you remove all of the noise." how could Sony claims that A7S has 15.3 stops dynamic range?

I also really want to know what firmware the A7s was running?

It really strange that Sony would market the camera with "unprecedented 15,3 stops of dynamic range"  only to have it measured 1 stop less than the other Sony A7 cameras!

Sony must have tested the sensor against their other sensor, before they decided to market it as a unprecedented dynamic range monster.

Something is really weird here. Please let us know what firmware you were using on your test camera.

Hi balazer, hi all,

Thank you for your feedback, and our apologies for the late reply. For now the firmware updates do not seem to affect RAW image quality, at least on the DxOMark sensor point of view (dynamic, noise, color sentivity).

Best regards

Thank you for your interest in DxOMark and our apologies for the late reply.
Eventhough we cannot put Sony's measurement methodology forward, there may be some differences in our respective dynamic range definitions. 
However, here is the information you are looking for concerning the firmware used for the Sony A7s measurments: - Key: Image Software
- Value: ILCE-7S v1.00

Best regards

Thats because A7 has phase detect AF pixels on sensor.

~1000 higher ISO is superior...

After 10 days, no answer ? The camera is in stores now, why still no test ?

I'm a great admirer of your work here, and would be unwelcome to complain about something of quality that I'm not even paying for. But I really want a new camera for these holydays, and I’m waiting for your tests to decide between D610 and Alpha 7S (I konow, they’re quite different…)

Hi,

Thank you for your support to DxOMark, and sorry for the late reply. The Sony Alpha 7s test results will be published today!

Sincerely