×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Is binning accounted for when guiding?

  • Posts: 28
  • Thank you received: 1
I was troubleshooting my guiding when I tried changing the binning of my guide camera to 2x2 and my guiding improved significantly. Went to bed last night quite happy but someone pointed out it might be that the guiding software doesn't take binning into consideration and the error could've just been halved without real improvement.

Could someone clarify how binning is handled by the internal guider? Thanks
2 years 7 months ago #74160

Please Log in or Create an account to join the conversation.

  • Posts: 28
  • Thank you received: 1
From what I can see it appears the improvement I saw was not true. The internal guider doesn't take the binning into consideration as I was seeing total RMS between 0.8" - 1.5" which is well below my ~2.2 arcsec/px resolution but this still resulted in elongated stars which makes sense if you double the values reported.
2 years 7 months ago #74174

Please Log in or Create an account to join the conversation.

  • Posts: 1208
  • Thank you received: 559
I quickly checked the code, and it seems to me that binning is correctly accounted for when computing the internal guider's rms error.

Read on if you're interested to see the details in the code:

Here's where the "drift" (difference is RA or DEC in arc-seconds) for a given time sample is computed:
invent.kde.org/education/kstars/-/blob/m...uide/gmath.cpp#L1130

Those "arc" variables (meaning they are in units of arc-seconds as opposed to pixels) are computed slightly above:
invent.kde.org/education/kstars/-/blob/m...uide/gmath.cpp#L1122

The "convertToArcseconds() method can be found here:
invent.kde.org/education/kstars/-/blob/m.../calibration.cpp#L65

wherexArcsecondsPerPixel() and yArcsecondsPerPixel() seem to take binning into account:
invent.kde.org/education/kstars/-/blob/m...calibration.cpp#L107

If you wanted to double check that these calculations were being done properly, 
you could look at the guiding log being stored at a sister directory to the logs directory
(e.g. on linux that's ~/.local/share/kstars/guidelogs) and look at the header of your most recent guide log, 
e.g. mine from last night looks like this:

KStars version 3.5.5. PHD2 log version 2.5. Log enabled at 2021-08-04 22:42:34
Guiding Begins at 2021-08-04 22:42:34
Pixel scale = 3.87 arc-sec/px, Binning = 2, Focal length = 400 mm
RA = 22.64 hr, Dec = 34.5 deg, Hour angle = N/A hr, Pier side = West, Rotator pos = N/A, Alt = 40.6 deg, Az = 73.5 deg
Mount = mount, xAngle = 128.5, xRate = 1.599, yAngle = 219.4, yRate = 1.788
Frame,Time,mount,dx,dy,RARawDistance,DECRawDistance,RAGuideDistance,DECGuideDistance,RADuration,RADirection,DECDuration,DECDirection,XStep,YStep,StarMass,SNR,ErrorCode
1,3.488,"Mount",0.000,0.000,0.021,-0.045,0.000,0.000,0,W,0,,,,3045,39.02,0
2,6.798,"Mount",0.058,-0.047,0.027,-0.036,0.000,0.000,0,W,0,,,,3039,38.98,0
3,10.246,"Mount",-0.042,0.018,-0.063,-0.059,-0.037,-0.054,23,E,30,N,,,3108,39.42,0


Yours may be a bit different, as I store the calibration, and thus there is no calibration at the start.
Search for the line that looks like

Pixel scale = 3.87 arc-sec/px, Binning = 2, Focal length = 400 mm

and you can see whether the parameters look right.

E.g. that line is printed by 
invent.kde.org/education/kstars/-/blob/m...de/guidelog.cpp#L152
where info.pixelScale is filled in here:
invent.kde.org/education/kstars/-/blob/m...rnalguider.cpp#L1608

 
Last edit: 2 years 7 months ago by Hy Murveit.
2 years 7 months ago #74178

Please Log in or Create an account to join the conversation.

  • Posts: 28
  • Thank you received: 1
I really appreciate you having a look. I won't spend too much time looking through the code. If you feel the binning is taken into account I'll trust that. I tried to check if the presentation of total RMS used the same values but I couldn't quite follow the code just using the browser..
Can you give any reason why my guiding would improve drastically from just changing the binning of the guide camera? I'll include my guide log, which doesn't seem to show any great guiding

I also took some (poor) pictures of the screen.I really appreciate you having a look. I won't spend too much time looking through the code. If you feel the binning is taken into account I'll trust that. I tried to check if the presentation of total RMS used the same values but I couldn't quite follow the code just using the browser..
Can you give any reason why my guiding would improve drastically from just changing the binning of the guide camera? I'll include my guide log, if I open it in the PHD2 log viewer then it doesn't look very great to me at least.

I also took some (poor) pictures of the screen which had me very happy seeing this results.

File Attachment:

File Name: guide_log-...5-32.txt
File Size:69 KB


2 years 7 months ago #74181
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 1208
  • Thank you received: 559
Re the binning, is it repeatable that your error goes down when you bin 2x2? E.g. can you switch back a forth a few times and see? Could the sky conditions have changed when you made the adjustment? However, IMHO, looking at your guide log, you have other issues you should try to address first.

Looking at your log in phdlogview, I see roughly the same guiding RMS error in the 4-minute and 6-minute 1x1 binning runs, than I see in the 24-minute 2x2 run at the end. That is, I don't see any improvement when you went to 2x2. Do you think it displayed differently in Ekos?

More importantly, looking that the 24-minute-long 2x2 binned run at the end (and truthfully in all the runs) there seem to be guiding spikes +/- 5 arc-seconds every 100 seconds or so. phdlogview seems to say you get about 2.3" RMS during that run. This all seems a bit too much error. Is this periodic error in your mount? Does the mount need a gear adjustment? Is the mount/scope unbalanced? How's the polar alignment? My suspicion is that there is something mechanical that you need to fix/adjust.
2 years 7 months ago #74183

Please Log in or Create an account to join the conversation.

  • Posts: 28
  • Thank you received: 1
Just to give a bit of background, I've always had trouble guiding with this second hand mount and getting pretty bad performance and lately I've been putting more effort into troubleshooting it.
So during the night I was initially always guiding bin 1x1 like I always have and got the same crappy guiding. Then I did two things, I reset the mount model and I switched to bin 2x2. Like turning on a switch I was suddenly getting sub-arcsecond results. I thought maybe this was some fluke and slewed to another target and saw the same result. Now, as you say, the logs aren't really showing anything impressive though, but the plot you see in my picture is pretty much what I always saw with bin 2x2, at the most it went to 1.5 arcseconds.

Since this is a quite old second hand mount, once again like you're suggesting, I'm thinking it has some mechanical issues that I've also tried getting verified from the logs but nobody seems proficient at reading the logs and telling me if they indicate any issue with the mount. Because I've also been worried about if it's my rather cheap Omegon guide scope that could be the issue.

What information is possible to get from the calibration, I haven't understood how to read that exactly?
2 years 7 months ago #74184

Please Log in or Create an account to join the conversation.

  • Posts: 1208
  • Thank you received: 559
If I were you, my first shot would be to make sure that the worm gear is adjusted and lubricated well. That is you can only feel a small amount of backlash in RA and DEC. You may be able to find YouTube videos showing how to adjust those for your mount. I can't be sure that that's your issue though. You also have to make sure that your mount can carry the amount of weight you're putting on it.
2 years 7 months ago #74185

Please Log in or Create an account to join the conversation.

  • Posts: 984
  • Thank you received: 160
Simon, until a year or so ago guiding did not take binning into consideration. Do you use a current version of Kstars/Ekos?
2 years 7 months ago #74187

Please Log in or Create an account to join the conversation.

  • Posts: 28
  • Thank you received: 1
Hy, yes I've started taking the mount apart and I'll go through the mechanical details to see that it's cleaned and lubricated as well as in general good condition. I'll make sure it's tightened correctly too.
The mount is quite an old NEQ-6 and I'm using a William Optics Z73 with a mono setup and a 50 mm guide scope. I'm really not very worried about too much weight as I haven't had to use more than one counterweight at just over halfway up the bar.

Alfred, I'm running a remote indi server which is running quite new software from just a month ago and I checked on the laptop it's running kstars 3.5.3 Stable and it says the build is from 2021-05-06 so it shouldn't be too old I think.
2 years 7 months ago #74198

Please Log in or Create an account to join the conversation.

  • Posts: 984
  • Thank you received: 160
3.5.3. certainly should do. Good luck with your mount upgrade efforts!
2 years 7 months ago #74202

Please Log in or Create an account to join the conversation.

  • Posts: 28
  • Thank you received: 1
Alright, I got the mount into pieces and found that the RA worm shaft was likely over-tightened and couldn't move smoothly. I've loosened it, cleaned and regreased a bunch of stuff. I've tried to adjust everything and I'm pretty much waiting for a clear night to test it now.
2 years 7 months ago #74214

Please Log in or Create an account to join the conversation.

Time to create page: 0.790 seconds