Hi Ben back from my break. Just pulled the latest code. I see [Error] Starting exposure (which appeared in log before). Next line is [Info] Exposure has begun. Then it hangs with the EKOS camera interface showing the countdown whirlygig but the countdown doesn't progress.
... After 3 minutes it times out and tries unsuccessfully to capture again failing to set binning and other exposure parameters on the second attempt.
Hey Dave. I just pushed some changes that undid some changes that I made to "clean up" the driver a bit. They were unrelated to fixing the download issue, which I hope should now work. Let me know how it goes.
Images of neighbours chimney pots attached!
The raw image captured in EKOS can be opened in OCS and other packages like GIMP or Affinity.
I notice if doing a file save as in the FITS viewer to a jpeg or png format it doesn't do it but saves it as a fits file.
If I do a save as in the FITS viewer to a fits format in a different folder the x-axis starts at a different point. Very roughly a vertical band about 1/6th the image on the left end of the image gets moved to the right hand end of the image?
The captured image looks very similar in OCS and EKOS which is good.
I think there is something unusual about the RAW format of the starshoot image which software packages struggle to handle correctly. I struggle to get capture a 'true' daytime colour representation of say the neighbours chimney pots and that's using OCS. Is there anything the Indi driver needs to consider to handle the G3 data correctly? For example in Forums I have seen an offset value mentioned
More testing needed ....
The color issues are due to the fact that the sensor used on the G3 color (ICX419AKL) has a CMYG pattern rather than more common RGBG patterns. The driver is currently advertising that it's monochrome, so I can fix that but I think it will still be demosaiced incorrectly. I think that's a separate issue to be sorted out.
Do you see any other issues with the image capture? Does it work reliably?
Here's a list of things that still need to be done, in priority order:
1. Hook up the gain and offset controls. This is 99% done and i just need a minor code change.
2. Get cooling working. I know the firmware commands to get this working. A capture where you plug in 12v, wait 10s, then unplug 12v might be helpful.
3. Get binning working. I have everything I need to make this work.
4. ST-4 control. I don't have any wireshark dumps that show this. If you can capture one, that would be great. Does OCS allow you to manually send guide commands? If so, a N,S,E, and W guide of different guide times would be helpful.
Do you see anything missing or anything that you'd like to prioritize to be able to actually use your camera?
Also, if there's anyone out there with a mono camera could you send the detailed lsusb output (lsusb -vv)?
I will do more testing like binning and whatever the EKOS camera interface permits. I did notice sequential imaging and image looping works.
1) Over to you Ben
2) Not had cooling power plugged in yet. Let me know when you want this tested?
3) Let me know when you want this tested as well?
4) I can swap my ZWO mini for the starshoot and test guiding using PHD2 or EKOS internal guider to drive my Meade Telescope. Will capture wireshark log. OCS has a PHD Link. Could try this as well.
For the ST-4 support, I think you can use PHD via ASCOM and use the Manual Guide function to send the guide commands without actually using a guide star. No mount or stars needed.
I'll re-review some previous dumps that Bob S provided with regards to cooling. I remember that I hadn't quite worked out how the setpoint and actual temperatures were encoded, but know enough to turn the TEC on and off.
I'll work on removing debug code and cleaning things up for now.
I just pushed some changes (and verified that they were pushed up to github). It includes some pretty significant cleanup, but it's mostly internal stuff. I did quite a bit of testing on my end but all of the hardware stuff was, of course, faked out. It would be helpful if you could test doing some basic captures to make sure nothing blows up.
I did also plumb in the gain and offset settings. It would be helpful if you could take a bias frame at two different offsets and verify that the statistics track. For gain, you could try a light frame with same exposure times at two different gains and verify that the pixel values change. I did verify that changing those settings resulted in the usb call that I expect, so I think that's probably scientific enough.
If nothing has been broken, then I suppose we could try and get binning working...but I wonder how valuable that would really be. The pixels are already big and it seems that most people who have ever posted about this camera have the color version. I think that cooling or ST-4 might be more interesting, but we would need additional wireshark dumps for those. Do you have an opinion?
For cooling, it would be helpful to get a capture where you turn on the cooler (with 12V connected) and change the setpoint a few times and then let it sit until the cooler reaches the setpoint.
Hi Ben. From Monday I am away for two weeks so if I don't respond on the forum for a while please don't take that as a lack of interest.
Things to do:-
I need to test the driver and the features you mention such as cooling, offset and gain. Will do this before Monday.
Capture USB traffic when sending ST-4 commands to camera.
'Save As' in the Fits Viewer does not work. PNG and JPG conversions saved as FITS. Save as another FITS in a different location shifts the left hand edge of pixels to the right hand end of the image.