qhy 168c 0 wonky values

User avatar
admin
Site Admin
Posts: 4332
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: qhy 168c 0 wonky values

Post by admin » Thu May 21, 2020 6:46 pm

Hi,

I would look to see how the ranges of the digital gain control and the individual colour gain controls match up with the 0 to 4027 range shown in the Ascom driver. For the 168C the the gain ranges for the individual colour channel seem to be 0-15. If you see the 0 to 4027 range on one of those controls or the other then you probably know which one corresponds to the control in the Ascom driver. If they all have different ranges then I'm afraid I don't know!

I'm hoping that when SharpCap has the correct sensor analysis data it will realise that it only needs a low value for the black level to avoid whiteout issues and will pick a sensible value. If it doesn't then I will need to check out the sensor analysis results that you get so that I can try to understand why.

Cheers, Robin

bwells
Posts: 36
Joined: Sat Sep 08, 2018 1:01 am

Re: qhy 168c 0 wonky values

Post by bwells » Fri May 22, 2020 8:57 pm

I tried a few things last night with some results to report:
  • All three gain sliders report 0-4027 - the same as the ASCOM driver
  • When I ran sensor analysis I kept getting first one message "The exposure time required is getting quite long". I seemed to get this often. Then after an extended period of time, the graph of green "x" would show up and all seemed to be working. But then, I would get one of a couple of messages:sometimes the message would say "Could not adjust the exposure to give the required brightness level, sorry". Then other times I would get the message saying something about the minimum exposure time was reached and to lower the amount of light.
  • So in the end, the sensor analysis would not run, no matter what I did. I had a dark room with a light panel, I opened a crack in the door and let in some sunlight, I used a few other lights but no matter what I did, there was a narrow range of light needed to get the "Proceed" button to be enabled and then it always ended up aborting saying the light was not bright enough. I do not know what this means. I didnt set any values for gain or offset since the analysis automatically adjusted those anyway and I could observe the sliders moving. And it seems like the proceed button enables when the light is bright enough to put the histogram around 50% but then, later in the processing, it ends up saying there is too much light.
  • Interestingly before, I would get this message about the minimum exposure reached, and I would see the exposure value set to the very lowest setting. But now, I still get that message, but the exposure slider is not at the lowest value anymore. It is usually around 1100ms

User avatar
admin
Site Admin
Posts: 4332
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: qhy 168c 0 wonky values

Post by admin » Sat May 23, 2020 7:45 pm

Hi,

Could you send me the log file captured after you attempted to run the sensor analysis? A lot of information goes into the log recording things like gain and exposure changes and frame brightness levels and maybe from that information I can work out a bit more about what is going on.

Cheers, Robin

bwells
Posts: 36
Joined: Sat Sep 08, 2018 1:01 am

Re: qhy 168c 0 wonky values

Post by bwells » Mon May 25, 2020 2:36 am

Hey Robin, I have attached one of many log files I had. I did not see any reference to "analysis" in the more recent log files, so I am not certain but I think this log file is the one you want.

I did a lot of things as I was working my way through SharpCap and I am not certain this is the right file. I thought I saw the analysis problem a day after this date, but I have been staying up past midnight a lot lately so dates get bury!

Let me know if this is even close to what you have in mind and if not, I can go back to my site and build just what you need and send you that log file instead.
Log_2020-05-21T17_11_34-13496.log
(767.05 KiB) Downloaded 8 times

User avatar
admin
Site Admin
Posts: 4332
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: qhy 168c 0 wonky values

Post by admin » Mon May 25, 2020 9:47 pm

Hi,


The problem here looks like response the of camera to exposure changes isn't smooth. What I mean by that is that SharpCap is trying to get a particular brightness of image and it ends up with an exposure about 0.47 seconds and finds that the image is a little too dim. SharpCap gradually increases the exposure to try to brighten the image, but nothing changes in the image until the exposure reaches about 0.59 seconds at which point the image brightness jumps to a value that is now too bright. SharpCap responds to the image being too bright by lowering the exposure again gradually, but again it jumped from being too bright to too dim without ever hitting the acceptable range.

The jump in brightness is from about 61% saturation to about 80% saturation with no way to get values between. It would be interesting if you could test the camera manually between those exposure values to see if you see the same behaviour – that instead of the image brightening gradually as you increase the exposure, that it stays the same and then jumps to a new brighter level.

Cheers, Robin

bwells
Posts: 36
Joined: Sat Sep 08, 2018 1:01 am

Re: qhy 168c 0 wonky values

Post by bwells » Mon May 25, 2020 11:10 pm

I will check this soon. Is there a specific gain and offset you want me to use?

And do you want me to check the levels for a few exposure values between 0.47 to 0.59 seconds?

I want to be sure I run the specific test that will be most meaningful. And I will try these using SharpCap and using APT.

One possibility is I have the SDK DLL and the camera driver mismatched? I have been trying to get my QHY5ii guide camera working with Metaguide and PHD2 and that has been a nightmare with DLLs. Anyway, I may have created a mismatch with DLLs that might cause the behavior you describe. Once I can check to see if the levels are wrong, I will know more.

User avatar
admin
Site Admin
Posts: 4332
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: qhy 168c 0 wonky values

Post by admin » Tue May 26, 2020 8:30 pm

Hi,

Incorrect DLL versions could certainly be a factor in this. I think the problem is happening at minimum gain when SharpCap is trying to measure the linearity of the sensor response. Cutting out some of the junk, I see this in the log

Code: Select all

Info:	17:21:02.8646113  Setting Negative closest exposure to 479.640396016278ma with current value of 0.606289088726044
Info:	17:21:02.8646113  Adjusting exposure, current is 0.6063, target is 0.65+/-0.025 (pedestal = 0),  changing from 479.64ms to 496.93ms
Info:	17:21:04.7097208  Adjusting exposure, current is 0.6061, target is 0.65+/-0.025 (pedestal = 0),  changing from 496.93ms to 514.91ms
Info:	17:21:06.5593408  Adjusting exposure, current is 0.6060, target is 0.65+/-0.025 (pedestal = 0),  changing from 514.91ms to 533.59ms
Info:	17:21:08.4044082  Adjusting exposure, current is 0.6059, target is 0.65+/-0.025 (pedestal = 0),  changing from 533.59ms to 552.99ms
Info:	17:21:10.2509667  Adjusting exposure, current is 0.6059, target is 0.65+/-0.025 (pedestal = 0),  changing from 552.99ms to 573.11ms
Info:	17:21:12.0972937  Adjusting exposure, current is 0.6061, target is 0.65+/-0.025 (pedestal = 0),  changing from 573.11ms to 593.88ms
Info:	17:21:14.5302175  Setting Positive closest exposure to 593.882248675294ms with current value of 0.7981076836586
That shows that at 479 ms and a brightness of 0.606 (60.6%). As the exposure is increased up to 573 ms the brightness doesn't really change, but then suddenly at 593 ms the brightness jumps to 0.798 (79.8%)

Hope this helps, Robin

Post Reply