QHY5III200M with much more noise after v4.0.9545 upgrade

[admin]

Hi,

thanks Jean-Francois. Those figures are effectively the same to within the expected margin of error of the analysis, so it looks like nothing significant has changed in terms of the camera response to gain in 16 bit mode. Was the original problem happening in 16 bit mode or 8 bit mode?

cheers,

Robin

[A.Cidadão]

Thank you both!

Completely different results from mine (see below; made them in triplicate)

So, different camera behavior, before and after last SC update (now both 4 and 4.1 with lastest updates).

Also, different camera behavior, at the same date, with SC4 (after updated; strange behavior) and 4.1 (at the time not yet updated; normal behavior; unfortunately cannot use 64bit versions due to ASCOM driver for focuser). Same day, before updating SC4 normal behavior of the camera. Same day, after strange effect due to SC4 update, normal behavior with SC4.1 while not updated.

So problem should not be the camera.

To complicate things even more, the QHY download site downloads a March10 all in all file (https://www.qhyccd.com/file/repository/ ... .10.14.exe) on the 230228 link. Besides, the SDK update log has two 230228 entries, one for (https://www.qhyccd.com/file/repository/ ... .10.14.exe), and the other for (https://www.qhyccd.com/file/repository/ ... .28.15.exe).

For the SC updates I made, installing one or the other all in one files does not solve the issue.

Can I revert the last SC4 update (without loosing settings) to try to solve this or need to wait for a new version?

Thanks!

António

Gain Value e/ADU Read Noise (e) Full Well (e) Relative Gain Rel. Gain (db) Dynamic Range (Stops)
0 2,165759107 3,15004749 1108,868663 1 0 8,459499205
50 1,201699298 2,599044233 615,2700406 1,802247126 5,116286835 7,887094748
100 0,692172048 2,099570877 354,3920887 3,128931762 9,907921838 7,399108097
150 0,381101679 0,942297365 195,1240597 5,682890486 15,09138574 7,693993556
200 0,216704211 0,769064345 110,9525562 9,994079456 19,99485596 7,172622882
250 0,131434291 0,690299247 67,29435721 16,47788475 24,33802922 6,607119816
300 0,079125768 0,628111971 40,51239327 27,37109742 28,74584421 6,011197739
350 0,046786272 0,509609258 23,95457144 46,2904822 33,30983409 5,554765705
400 0,034866419 0,459413637 17,85160646 62,11590341 35,86405612 5,280116416
450 0,028478467 0,478882385 14,58097495 76,04900678 37,62187093 4,928272009
500 0,024192202 0,511033095 12,38640763 89,52302364 39,03869484 4,599197294
550 0,02134921 0,612586673 10,93079577 101,4444589 40,1245666 4,157340644
600 0,021576212 0,783429689 11,04702059 100,3771699 40,03269894 3,817709711
650 0,035567952 1,711255676 18,21079159 60,89074475 35,69102572 3,411666403
700 -0,019163594 -1,203026996 -9,811760257 -113,0142435 NaN 3,027842966
750 -0,013245892 -1,030152183 -6,781896832 -163,5042069 NaN 2,718831357
800 -0,015787708 -1,6335545 -8,083306336 -137,1800865 NaN 2,306930935

[A.Cidadão]

Hi Robin, Jean-Francois, went back to SC4 v 4.0.9538 and problems disappeared.

Also, sensor analysis gave a proper result (attached).

Something went wrong with my update, but for the moment I'm leaving things as they are.

Thank you both! Best wishes

António

[admin]

Hi,

the values you get with the latest SDK are really odd - they only make sense if the data coming back from the camera has only 9 bits of information in it (rather than the 12 you would expect). For instance at gain 0, the ratio between the FWD and the e/ADU is 512, but we would expect 4096 in 12 bit mode. Even with the older SDK, you are getting 11 bit data, which still seems wrong. (I looked at the analysis data I have in my database and that is also 11 bit!). The FWD is correct with the older SDK, so I suspect that the SDK is somehow throwing away the least significant bit of data from the senosr.

The most likely cause is that the SDK is adding some processing to the data that we are not expecting - it could be a rogue setting that is different or has a different default value, but it's really hard to be sure. It would be worth comparing the 'CaptureSettings' files created with the old and new SDK just to see if there are any camera settings that SharpCap knows about that have different default values.

As to going back to the older version, that will be fine and won't mess up your settings. I think there is something not quite right with the handling of this camera in the QHY SDK (even in older SDKs) and possibly worse in newer ones, which will probably need to be fixed by QHY in a future SDK update.

cheers,

Robin

[Jean-Francois]

Hello,

OK ... it means that my camera "lost" a bit too

Here the begin of the camera analysis report:
Gain Value e/ADU Read Noise (e) Full Well (e) Relative Gain Rel. Gain (db) Dynamic Range (Stops)
0 4.06 3.00 8318.69 1 0 11
50 2.27 2.54 4645.07 1.79 5.06 10.84
100 1.31 2.03 2677.54 3.11 9.85 10.36
150 0.74 0.91 1492.062 5.58 14.93 10.68

I use the last QHY beta driver version.

Regards,
Jean-Francois