7 replies to this topic

[rlmxracer]

What do the experienced astrophotogs look for in their subs to know if they are good or if changes need to be made? 

[Notdarkenough]

For me, I look for bad things. Airplanes, trees, something crazy out of focus, whatever. The when stacking I will only keep best 70%. So, in short, I eliminate bad during that first initial cull and let the software eliminate the margin calls. 

[WadeH237]

I image fully automated, so sometimes don't look at any of the subs until the next day.

 

When I do spot check subs during the capture session, I just look at the basics:  Is the framing correct?  Is it in focus?  Are the stars round?  For most other issues, I would examine the data in full at the time I process it.  If there were any problems with the run, or the subs themselves, I examine the relevant logs.

 

I would note that I have been doing this for a very long time, and my system is very stable and very reliable.  When setting up a new system, I don't turn it over to unattended operation until I am happy with its stability.

 

So assuming a new system, I would look for things like:

 

• Are the focuser settings producing good focus runs?
• Are tilt or collimation problems evident in the corner stars?
• Is it able to plate solve reliably?
• Does it correctly handle a meridian flip?
• Is the guider working well in general?
• Does the guider recover quickly after a dither?
• Does the guider handle a meridian flip correctly?

 

I would note that I stopped guiding a few years ago and now image unguided.  I included the guider steps above because those are the steps I would use on a guided system.

 

I would also note that I don't look at the guiding RMS.  In my experience, with a well-tuned system, the guider RMS is basically a report of seeing.  What I actually care about when evaluating how my guider is working is:  Are the stars round?  Does their brightness profile look reasonable?  If those two things are correct, and the system was properly focused, the data in the subs is far more important than any single guider number.

 

If I am going to watch the guide software run, what I am looking for is the "jaggedness" of the lines in the guide graph.  That jaggedness is caused by atmospheric seeing.  What I want to see is that the RA and declination graph lines look similar.  If they look about the same, you are good.  If one of them shows either occasional spikes (not due to dithering), or a clear up-and-down oscillation that is bigger than the jaggies, then there is problem guiding.

 

I hope that helps.

[rlmxracer]

I have been spot checking the star size after I get an average from the first few frames after focusing. when the star size is creeping up I know its time to refocus. I have average to below avg seeing in my location so I know that can cause some issues with guiding. Thanks for the tips.

[joshman]

I think WadeH237 covered off on most of what I look for. Like him, my system is stable and reliable, so I don't need to babysit the runs. I don't need to but I generally like to watch the system un-park, slew, platesolve, autofocus, and the first sub come in.

When spot checking the run, i'm always looking at these items in particular:

    General exposure/median background
    FWHM/HFR (how sharp the stars are)
    Eccentricity (how round the stars are)
    Guiding performance.

[dcaponeii]

APP does the bulk of the culling for me.  I don't look at the subs much but do check the statistics graphs in APP just in case I think some additional subs should be culled.

[auroraTDunn]

Every morning I transfer my images to a project specific Quarantine folder. I then run them through Blink and look for;

 - I don't worry about satellite/plane tracks to much (sometimes a helicopter is a nasty thing though!)

- any elongation/trailing of stars

- clouds, especially high level clouds which seem so minor yet Blinking through you can tell they are there

- overly bloated or dimmed stars (another sign of clouds)

- trees are a MAJOR problem for me so I always try to note the time the incursion happened so I know the next night to make sure to not schedule that many frames and save that time for other targets.

- Dithering that continues in a particular direction to long (this is a tracking issue and if it will cause to great a cropping later I remove them.

 

I sometimes run "Subframe selector" to cull really bad subs early so I know how many replacement frames I need to take to make up for them. But for the most part I let WBPP do the heavy culling for me, but I do check the logs to find those frames, physically remove them and replace them with new frames.

[Spaceman 56]

I look at every sub visually as they arrive. if they look blurry or have trails they get discarded.

 

after that I put them into ASTAP for evaluation (in real time) if the data trends are not coinciding with the general data results I might discard them.

 

for example if an imaging run is generally getting 2850 stars according to ASTAP, and a particular sub shows 1000 stars, I can conclude that clouds have most likely obscured the target, and I would discard that sub.

 

I do the same with HFD, and overall Quality as recorded by ASTAP. discard the outliers. at the end of the night just the most consistent subs get retained.

 

then as I shoot over between 5 and 8 nights, the data sets from each night are compared. sometimes its obvious that one night does not rate as well as the other 6 or 7 nights, and I would discard the data from that particular night, and not include it in the stack.

 

all thats just where I am at in March 2025, and I hope to improve these methods. I am still working on Data acquisition.