Don't do what I did last night

9 replies654 views
Andy Wray avatar
I thought I'd done everything right last night and captured nothing useful.  Here's how it went:

About an hour after dusk I did the following:

* I did an autofocus with the scope pointing at the celestial pole
* I checked the polar alignment with SharpCap and it was all OK from the session from a few nights ago
* I slewed to my target (M51)
* I engaged PHD2 and got it to calibrate and start tracking
* I grabbed my previous images of M51 from a couple of days ago and platesolved them, followed by using APT's Goto++ routine to get me within 12 pixels of the previous session
* I let the tracking settle down and it ended up at about 0.7 arc secs RMS
* I kicked off my 2 hour imaging sequence

So, what went wrong as all the above was silky smooth?

Bottom line:

* I started imaging before it was really dark enough, so the sky brightness was awful and therefore I spent an hour capturing poor SNR images.  I was just too keen to start imaging before the sky was dark enough.
* I didn't notice the whispy clouds roll in;  PHD2 was still achieving 0.7 to 0.8 arc secs guiding, but I was imaging through clouds, so even when it was dark enough later on I was ending up with blurry images.
* I've ended up throwing away all of last night's session as the images will actually make my end result worse.

Unfortunately tonight looks like I have a similar problem with clouds, but at least I'm not going to kick off that routine until I'm sure there is at least half a chance of getting some good stuff.  I'll be going to bed before 1am tonight smile

On the plus side:  all the technical bits are working fine and I'm pretty relaxed about getting an imaging session set up and going now
On the downside:  I can still be a bit thick when it comes to timing and mother nature will always have her say.
Well Written Helpful Insightful Engaging Supportive
Frank Alvaro avatar
There's always a gremlin in the works that we overlook, isn't there Andy? I start and finish my imaging at the start and end of astronomical night (nice and long here in Australia at the moment 😊); I assume this is ok.

MY latest blunder occurred on the weekend when I traveled to a relative's house in a Bortle 1 zone. Hooked everything up, great PA and guiding, and target acquired …. then I see that I had forgotten to recharge the power bank after the previous night's session, leaving only enough for about 3-4 hours imaging 😡. Luckily there was a near full moon and the clouds quickly rolled in, saving me from kicking myself for missing a full night under a dark sky.

Live and learn for both of us.
Well Written Respectful Concise Engaging Supportive
Andy Wray avatar
Frank Alvaro:
Luckily there was a near full moon and the clouds quickly rolled in, saving me from kicking myself for missing a full night under a dark sky.


That made me laugh.
jjpoole741 avatar
I kicked off my imaging session and realized I input the wrong capture sequence… 30 x 1” vs 30 x 180”. Checked on it after 30 minutes thinking I’d be a few photos on and realized it was incorrect. Corrected it, captured an hour and a half then realized that I had the “warm camera after sequence” selected in NINA, so after that incorrect sequence it warmed my camera up from -15 to 20c. #FAIL
pterodattilo avatar
concerning the problem of light clouds or low atmosphere transparency, I use SubframeSelector in PixInsight to give less weight to images based
  • on the star count in the frame (first)
  • on the the new PSF Signal Weight and PSF Power Weight (now, but I have to study them better, I just noticed a correlation of these params with star count)
andrea tasselli avatar
I can't see any reason to throw away data unless its FWHM is horrible. So even in faint wispy clouds a resourceful APer has something to gain by judiciously using weighted averages in PI (that is, using SubFrameSelector process).
Andy Wray avatar
andrea tasselli:
I can't see any reason to throw away data unless its FWHM is horrible. So even in faint wispy clouds a resourceful APer has something to gain by judiciously using weighted averages in PI (that is, using SubFrameSelector process).

OK, I'll give it a try.  Up to now I have used subframeselector to analyse and select calibrated frames that I then tend to integrate in WBPP.
Andy Wray avatar
andrea tasselli:
I can't see any reason to throw away data unless its FWHM is horrible. So even in faint wispy clouds a resourceful APer has something to gain by judiciously using weighted averages in PI (that is, using SubFrameSelector process).

You were right (as you usually are) and I will not throw away images so easilly in future.  I actually just stuck with WBPP using a weighting of 20% FWHM, 20% Eccentricity and 60% PSF SNR Weight.  I also used local normalisation because of the cloud cover which was very visible in the individual subs.

The result:  quite a bit more detail in the galaxy
Helpful Respectful Concise
Pavle Culum avatar
Doesn't sound to me like you did anything wrong.

You started a session early. Most of us do that, especially if we're gonna be imaging for an extended period of time. You can easily throw out the really washed out subs, but you can't magically appear new ones when you missed them.

As for guiding, there is nothing there you could've done differently. Clouds are gonna cloud, and it will impact your guiding regardless of any precautions.

So I'd say a more appropriate title would've been "My session was foiled by weather, but I got to practice setting up my rig!".
Well Written Concise Supportive
Roger Nichol avatar
I tend to have NINA kick off my sequences before it is properly dark so that I can let PHD2 work out its PPEC corrections during the first 20 minutes or so. Those images go in the bin but the guiding on the subsequent images is marginally better (I think).  It also gives me chance to fix things on the rare occasion when something isn't working properly.
Well Written Helpful Insightful Concise