Sensor Analysis (Sony a6400) brightness clamped at 50%?

[jprincen]

I just connected my mirrorless sony a6400 via its ascom driver, and when trying to do sensor analysis the brightness seems to clamp at 50%, and no matter what I do it will never go above this. I have attached a screenshot.

Screenshot from 2022-12-02 11-38-03.png (202.63 KiB) Viewed 5664 times

I am new to Sharpcap. Am I missing some setting on the camera or sharpcap itself?

John.

[admin]

Hi John,

I suspect that this is something coming from the ASCOM driver that you are using for the Sony camera. An ASCOM driver can report to SharpCap the maximum pixel value that it will produce (via a property called MaxADU). SharpCap uses this when building the image from the ASCOM driver data. For instance if the driver reports a MaxADU of 20000, but none of the pixels can ever go above 10000 in the actual image then you will see exactly the sort of thing you have noted with the brightness never going above 50%.

There may be options to the ASCOM driver that let you configure this sort of thing - try pressing the 'Show' button for the ASCOM driver options on the right to see what you can adjust. Also the gain looks like a weird value - what gain values are available to you and does anything change if you use other gains?

cheers,

Robin

[jprincen]

Thanks for the feedback. The ASCOM driver is just reporting back a fixed value (not camera dependent) of 20000 for MaxADU. Unfortunately I can't change it from the UI, since the developer hard coded it. Not sure if it is correct or not, but I will take a more detailed look at the actual data returned and see. If it is wrong, I can probably tell the developer, and also recompile the driver myself.

I get all the typical ISO values for gain, except the weird one shown is the first one in the list (not sure what it means). Changing the gain doesn't change the max value, but just makes all the values get closer to the max, as expected.

John.

[admin]

Hi John,

I wonder if the maximum pixel value is something that varies from camera model to model with no way of finding out without testing each model -that would explain choosing a fixed value as 'good enough' from the developer's point of view. Anyway, if you use the pixel value readout tool in SharpCap you can see what the exact maximum value is when over-exposed and pass that on or change the code yourself.

The weird gain value is the largest number you can store in 3 bytes (2^24 - 1). Seems likely to be a bug in the ASCOM driver that is putting that into the gain list (unless it is a really weird SharpCap bug!)

cheers,

Robin

[jprincen]

Thanks Robin. Actually I did a check myself, by connecting to the driver through alpaca and python, and it looks like the max value for the RAW output is 16383, which means it is 14 bits. This is more than 50% of 20000, which I confirmed is the MaxADU returned by the driver. Would the debayering be doing anything to limit the range?

I will also confirm through SharpCap the way you suggested.

John.

[jprincen]

Also, the weird value for gain doesn't look like a sharpcap bug, I see the same weird value as the first gain through a diagnostic tool provided by the sony ascom driver writer.

John.

[jprincen]

Hi Robin,
What I see with SharpCap's pixel value tool is max values of 32744 for both RAW and RGB values. This is strange.

John.

[admin]

Hi John,

SharpCap will try to scale up the pixel values to fill the whole range, but it only does it in powers of 2. I wonder if the maximum pixel value is 16383, but the ASCOM driver perhaps reports 16384 (2^14). SharpCap might think it can only double the pixel values, since if it multiplied them by 4 then a pixel with maximum value would go out of range.

Try running the ASCOM conformance check ('Conform.exe' which will be installed with the ASCOM platform) on the camera driver. Fairly early on in the check it reads the Maximum ADU property and prints out the value in the output, like this

Code: Select all

14:59:13.474 MaxADU                            OK       4095

(Once it has printed this you can stop the test).

It would be interesting to know what value the driver is reporting.

cheers,

Robin

[jprincen]

Hi Robin,
I think this explains it. I checked the MaxADU value myself, and it is reporting 20000. When I checked the actual code, the 20000 is hard coded. I will see if I can change the driver code, and then confirm.

John.

[jprincen]

Hi Robin,
Fixing the drivers MaxADU return value allows the sensor analysis to run. Only one remaining issue for a complete set of information is that strange gain value. Turns out this strange gain (16777215) means set the camera ISO to Auto. I'd actually like to disable that, since it messes up the graph of sensor data. If there is a way to do that in SharpCap, let me know. Otherwise I will try to do it through the driver.

Thanks for the help so far!
John.