QHY174M-GPS calibration StartPos and EndPos

[skogpingvin]

Hi brains trust, Bill here

My excellent bosses (who run a telescope factory in Australia) are into citizen science, and we're getting into occultation timing (amongst other things). There's an occultation in about three days nearby and I'm preparing to go and catch it. I've been testing with a 107mm refractor and an ASI1600MM-P.

But yesterday another surprise - a QHY174M-GPS turned up! I've got it physically onto the refractor and I've got the drivers all sorted in SharpCap, but now I'm having trouble with the calibration. This is the mysterious bit where you set StartPos and EndPos with the calibration LED flashing. I've read the nearly non-existent documentation, as well as seeing a couple of YouTube presentations and it seems fairly plain.

But then when I fired it up, my StartPos defaulted to way higher than the EndPos (see the screenshot). This is weird. Should I do the normal thing and set StartPos to something low, like 20000, and push EndPos up to something high, and try to continue - edging StartPos up to when I see the LED, then backing it back until I can't, then pushing EndPos from something a bit higher than StartPos (so I can't see the LED) up until the LED reappears fully?

It's just that with the strange settings that appeared in the screenshot make me think there's something automatic happening.

Does anyone know of an up-to-date tutorial about this?

Many thanks!

[admin]

Hi Bill,

there is a long discussion worth reading here - viewtopic.php?t=2241 - a number of the denizens of the SharpCap forums put in a considerable amount of effort in order to work out the way that the calibration parameters varied as you changed various camera settings. The result of this was being able to build an 'autocalibration' feature into SharpCap that updates the calibration values as the relevant camera settings (exposure, USB speed, ROI size) change. The autocalibration is usually good to within a few hundred.

The end value being less than the start value seems odd, but is expected. Both are measured from an internal pulse in the cameras circuitry, and the 'end' value is the time of the end of the current frame after that pulse (in counts of a 75Mhz clock). The 'start' value is the time of the start of the next frame from that internal pulse, so has to be higher than the end. One way to miscalibrate is to go looking for an end value that is higher than start - you will find one, but it will be the end of the frame after, and will cause problems.

To be honest, SharpCap's autocalibration should be good for timing down to the sub millisecond level in almost all cases - manual calibration will get you down to microseconds, but you probably don't need that. In fact, no calibration at all would probably be good for 0.01->0.1s accuracy.

cheers,

Robin

[Jean-Francois]

Hello Bill,

I can confirm that the actual automatic LED pre-setting works very well.
In the past I did a lot of work for finding a general equation of the StartPos and EndPos values.

If you want the best LED calibration, you can tune the value manually.

Click "up" or "down" of the StartPos ("10" position) until the LED illumination disappear.
Click again until it reappears completely. Now guess where the illumination is at the half intensity.
After the StartPos, click the "up" and "down" of the EndPos until you find the half intensity. ... so finished

Note that if you change the exposure time, the USB traffic or the ROI height, you need to redo the manual calibration.
You will see that the SharpCap automatic calibration work at the end very well.

Alternative is that you use one of my automatic Python script for doing the same.

Regards,
Jean-Francois

[skogpingvin]

Thanks for your help Robin - I'll leave the calibration that automatically populates when I set the exposure (and other parameters).

Weather is looking - meh, possibly OK for Sunday morning (Saturday 15:24 UT). I sure hope it comes off, I'm missing a big star party in central Victoria for this!

Cheers, Bill

Hi Bill,

there is a long discussion worth reading here - viewtopic.php?t=2241 - a number of the denizens of the SharpCap forums put in a considerable amount of effort in order to work out the way that the calibration parameters varied as you changed various camera settings. The result of this was being able to build an 'autocalibration' feature into SharpCap that updates the calibration values as the relevant camera settings (exposure, USB speed, ROI size) change. The autocalibration is usually good to within a few hundred.

The end value being less than the start value seems odd, but is expected. Both are measured from an internal pulse in the cameras circuitry, and the 'end' value is the time of the end of the current frame after that pulse (in counts of a 75Mhz clock). The 'start' value is the time of the start of the next frame from that internal pulse, so has to be higher than the end. One way to miscalibrate is to go looking for an end value that is higher than start - you will find one, but it will be the end of the frame after, and will cause problems.

To be honest, SharpCap's autocalibration should be good for timing down to the sub millisecond level in almost all cases - manual calibration will get you down to microseconds, but you probably don't need that. In fact, no calibration at all would probably be good for 0.01->0.1s accuracy.

cheers,

Robin

[skogpingvin]

Thank you Jean Francois, it's good to know that the automatic settings are pretty accurate. If I need to get better accuracy (this camera is likely to be located in a permanent remote observatory somewhere) it's great that they can be made even more accurate.

Cheers, Bill

Hello Bill,

I can confirm that the actual automatic LED pre-setting works very well.
In the past I did a lot of work for finding a general equation of the StartPos and EndPos values.

If you want the best LED calibration, you can tune the value manually.

Click "up" or "down" of the StartPos ("10" position) until the LED illumination disappear.
Click again until it reappears completely. Now guess where the illumination is at the half intensity.
After the StartPos, click the "up" and "down" of the EndPos until you find the half intensity. ... so finished

Note that if you change the exposure time, the USB traffic or the ROI height, you need to redo the manual calibration.
You will see that the SharpCap automatic calibration work at the end very well.

Alternative is that you use one of my automatic Python script for doing the same.

Regards,
Jean-Francois

[procyon12]

Hi Bill, all,

I fully agree with the statements about auto-calibration. On your picture I see that the "USB traffic" is set to 50. Please try to reduce this value (usually around 3 ... 0) to avoid dropped frames. Reducing the capture area also helps against dropped frames. And, start the GPS at least 30 minutes before the occultation to ensure stable UTC timing (check the GPS data window regularly, also enable GPS logging in the settings).

Good luck with your observation!

Christian

[skogpingvin]

Thanks for the advice Christian, I'll reduce that number - do you mean "by around 3" (ie. to about 47) or "to around 3"?

I'll also enable the GPS logging.

Thanks - I suspect I'm going to need a little luck. The weather looks iffy, but the thing I'm worried about is that due to clouds I haven't had a chance to have the rig set up under stars, and I won't get one until game time. I need to check that SharpCap can adequately polar align (my refractor's 560mm FL and the QHY174 produce an FOV that might be marginal for plate solving). I also need to check that ASTAP can plate solve with the new camera - I'm going to find my way to the star using SGP. I'm pretty confident there, it absolutely nailed it in the dead centre of the frame with my ASI1600.

With luck I'll have a few hours before the occultation to troubleshoot things.

Cheers, Bill

Hi Bill, all,

I fully agree with the statements about auto-calibration. On your picture I see that the "USB traffic" is set to 50. Please try to reduce this value (usually around 3 ... 0) to avoid dropped frames. Reducing the capture area also helps against dropped frames. And, start the GPS at least 30 minutes before the occultation to ensure stable UTC timing (check the GPS data window regularly, also enable GPS logging in the settings).

Good luck with your observation!

Christian

[procyon12]

Hi Bill,

I meant "to" about 3 (e.g. 3 on my W7 machine, 0 on my W10 machine). You can test it with SC's FPS display at the bottom of the GUI. For plate solving 560 mm FL should be sufficient, you could also try a reducer.

Cheers

Christian

[Jean-Francois]

Hello Bill, Robin and other user of the QHY-174-GPS,

I have a look at my script for the LED calibration.
I do some minor changes and I update a little bit the "documentation".

Here two versions ... one script has to be started from the IronPython script console, the other script can be insert in the SharpCap menu.

One point, maybe can Robin explain, ideally find a way to solve the small difference.
The script running from the script console shows the live images ... so you can follow the intensity changes of the LED illumination.
The script running from the menu does not show the live images ... so, you see nothing, you need to wait the end of the calibration.

Note that the forum does not allow to upload directly a *.py file. I add ".txt" at the end of the file name.
So, delete the ".txt" after the download of the files.

Calibration_LED_menu_2022-10-02.py.txt

Menu version

(11.87 KiB) Downloaded 45 times

Calibration_LED_2022-10-02.py.txt

Console version

(11.49 KiB) Downloaded 45 times

You can place the files everywhere.
The menu version has to be load in SharpCap settings before use.

Robin, for the selection rectangle ... in the past SharpCap memorized the dimension and position of the rectangle.
I remark that the last version 4.0.9382 does not remember. The rectangle is on the top of the image after each start of SharpCap.

Best regards,
Jean-Francois

[admin]

Hi,

just to note that I will have a look at both the non-remembering of the selection area and the failure to update.

cheers,

Robin