Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

M101 and Vicinity - PHD2 Multistar Guiding Test During an Otherwise Stark, Starless Winter

Astrophotography CCD DSO Equipment Imaging Refractor
  • Please log in to reply
9 replies to this topic

#1 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,205
  • Joined: 13 Mar 2013

Posted 02 March 2021 - 06:45 PM

Greetings!

 

It has been a long, stark and starless winter in the Pacific Northwest. I have only had one night of imaging thus far in 2021, and this is the result of it. This is M101 and its vicinity, a luminance only image made into a mosaic showcasing M101 and its satellite neighbor NGC 5477. At the time I knew I was only going to get one night under a clear sky, and so I devoted the whole night to the L filter with the hope of coming back later with RGB. (I'm still waiting for that opportunity.)

 

It was, however, a nice opportunity to try out PHD2's new multistar imaging capability. I upgraded to the latest version of PHD2 but did not turn it on at first. After about 30 - 45 minutes of imaging I then switched it on for the rest of the frames, and I was pleased to see an immediate guiding improvement. As I recall, the guiding curves' RMS values decreased by about 25%. I later calculated FWHM values from the calibrated frames, and I saw about a 20% reduction in star size after switching it on. This capability is a welcome addition to guiding.

 

M101 (Ursa Major) is my favorite galaxy to image. Not as close to us as the galaxies of our Local Group, it is still a relatively close neighbor at 20.9 Mly (6.4 Mpc) with diameter 170 kly and apparent size 28.8' x 26.9'. This apparent size is large enough that it defies usual longer focal length images. Here I have used my TS 130/910 refractor with my FLI ML16200 CCD resulting in a resolution of 1.36 "/px.

 

I found this resolution to be small enough for the image to benefit from a modest touch of deconvolution. With less than 5 hours of integration time, I needed to apply noise reduction, however I did not wish to overdo it. This can be seen in the background. Hopefully, more imaging time in the near future will allow me to add more signal and reduction of noise.

 

The mosaic highlights M101 and I also wished to zoom in on NGC 5474. In this image, N is to the left, S right, E down, W up. Below M101 is NGC 5477, one possible candidate satellite galaxy responsible for distorting that outer arm as if unwinding it.

 

Hopefully, I'll get more opportunity to image again soon. I do have quite a bit of data left from last summer and fall that I haven't had time to process or upload, although much of it suffered from mediocre seeing conditions.

 

Best Regards,

Ben

 

Higher Resolution Link on AstroBin

 

M101_L2_133-01A.v12D_Mosaic-1.jpg

 


  • dghent, Scott Mitchell, PeteM and 9 others like this

#2 Linwood

Linwood

    Vanguard

  • -----
  • Posts: 2,206
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 02 March 2021 - 07:13 PM

I love the shot.  I've had similar experience with multi-star, but someone asked me a question that I did not know the answer to, and maybe someone else does.... 

 

Does the measurement of total RMS error change when you enable multi-star?   I.e. if one sees (as I do as well as the OP) a definite decrease in error -- is the measurement made the same way, or does the measurement also change and use multiple stars? 

 

I've imaged M101, but not quite so wide -- I need to think about doing it again and getting the whole set.  Thanks for sharing that, very nice.



#3 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,205
  • Joined: 13 Mar 2013

Posted 02 March 2021 - 07:50 PM

I love the shot.  I've had similar experience with multi-star, but someone asked me a question that I did not know the answer to, and maybe someone else does.... 

 

Does the measurement of total RMS error change when you enable multi-star?   I.e. if one sees (as I do as well as the OP) a definite decrease in error -- is the measurement made the same way, or does the measurement also change and use multiple stars? 

 

I've imaged M101, but not quite so wide -- I need to think about doing it again and getting the whole set.  Thanks for sharing that, very nice.

Thank you, Linwood!

 

It is my understanding that the RMS error reported by PHD comes directly from the fluctuations of the centroid measurement.  For single star guiding, this is, of course, the centroid of that single star.  For multistar guiding, I can only assume that this is the centroid averaged over the multiple stars, and therefore it will tend to fluctuate less than a single star since, in principle, the fluctations due to seeing conditions should be quite a bit smoother as you are averaging over the random seeing.  This explains the immediate reduction in RMS that I observed in the guiding curves as well as measured reduction in star sizes.

 

A goal in guiding would be to adjust the parameters (exposure time, guide hysteresis and/or other algorithm settings) such that seeing fluctuations are minimized or, if possible, eliminated.  In other words, we're advised to not "chase the seeing".  In my situation, I have never been able to adequately remove the seeing conditions from my guiding, but I welcome a method like multistar guiding that helps me to at least minimize and smooth it out.

 

Best Regards,

Ben



#4 Linwood

Linwood

    Vanguard

  • -----
  • Posts: 2,206
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 02 March 2021 - 07:55 PM

If that's true - that the measure changes -- there's no way from the reported errors to know if your guiding improved.  Well, other than if your stars are more round.



#5 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,205
  • Joined: 13 Mar 2013

Posted 02 March 2021 - 08:04 PM

I'm not following what you mean, Linwood.

 

The RMS fluctuations smoothed out immediately when I turned on multistar guiding.  This settled the fluctuations, which in turn reduced the number of and amount of guiding corrections.  And, as I said above, the real telling evidence of better guiding was that my star sizes (FWHM) were reduced measurably with multistar guiding turned on.  It's the star sizes, not the roundness, that is the measure of better guiding when making comparisons under the same seeing conditions.

 

But, do you have another interpretation that I may be missing?


  • rockstarbill likes this

#6 Linwood

Linwood

    Vanguard

  • -----
  • Posts: 2,206
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 02 March 2021 - 08:09 PM

Absolutely, FWHM is a good measure.  And I do think PHD2 guides better multistar.

 

I just wonder about the actual RMS measurement for error. 

If, for example, it measured error the same way on the primary start, then the RMS error before and after multi-star would be a valid comparison.

 

If on the other hand it measures error during multi-star based on all stars, not the target star, then the measurement is measuring something different, and you could not compare total error before and after.

 

I do not know the answer as to which it does, so that's the first question.



#7 Birddogoby

Birddogoby

    Vostok 1

  • -----
  • Posts: 103
  • Joined: 14 Jan 2020
  • Loc: Poulsbo, WA

Posted 02 March 2021 - 08:11 PM

Great mosaic!  M101 is also one of my favorites. I don't use PHD2 and use Lacerta's MGEN-3 multi-star guider, which I'm assuming uses the same basic principle.  It takes up to 100 stars and creates a "multi-star" that is used for guiding.  It's been really good at tracking with my EQ6R-Pro mount and Takahashi TSA120.  I typically get .4-.7" with the MGEN-3 even with our notorious seeing here and would like to get a guiding camera to compare the two with PHD2.  It's only money, right?

 

As an aside, I can't help but chime in as a fellow Pacific Northeasterner (Poulsbo) about the lousy AP opportunities we've had over the last two months.  I just received my QHY600M-PH and it soooo wants to see the stars!


  • psandelle likes this

#8 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,205
  • Joined: 13 Mar 2013

Posted 02 March 2021 - 08:27 PM

Absolutely, FWHM is a good measure.  And I do think PHD2 guides better multistar.

 

I just wonder about the actual RMS measurement for error. 

If, for example, it measured error the same way on the primary start, then the RMS error before and after multi-star would be a valid comparison.

 

If on the other hand it measures error during multi-star based on all stars, not the target star, then the measurement is measuring something different, and you could not compare total error before and after.

 

I do not know the answer as to which it does, so that's the first question.

OK, thanks, I think I follow your point.  My understanding is that multistar RMS values represent the average motions over all the target stars and I don't know whether or not you could access the value of a single star from that.  Maybe it's recorded in the debug or guide logs.

 

 

Great mosaic!  M101 is also one of my favorites. I don't use PHD2 and use Lacerta's MGEN-3 multi-star guider, which I'm assuming uses the same basic principle.  It takes up to 100 stars and creates a "multi-star" that is used for guiding.  It's been really good at tracking with my EQ6R-Pro mount and Takahashi TSA120.  I typically get .4-.7" with the MGEN-3 even with our notorious seeing here and would like to get a guiding camera to compare the two with PHD2.  It's only money, right?

 

As an aside, I can't help but chime in as a fellow Pacific Northeasterner (Poulsbo) about the lousy AP opportunities we've had over the last two months.  I just received my QHY600M-PH and it soooo wants to see the stars!

Thank you, Birddogoby!  Hang in there, we'll get a clear night or two eventually ...



#9 rockstarbill

rockstarbill

    Cosmos

  • *****
  • Posts: 8,445
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 02 March 2021 - 09:46 PM

If that's true - that the measure changes -- there's no way from the reported errors to know if your guiding improved.  Well, other than if your stars are more round.

The roundness of stars is not the real measure of guiding effectiveness. FWHM is. I can have perfectly round and fat stars. In fact, the eccentricity of a star will be lower on stars with higher FWHM values. However if you have very low FWHM and very low eccentricity, you can bet those are pretty darn good frames.

 

At the end of the day, astroimaging is not about guide graphs.


  • psandelle likes this

#10 spokeshave

spokeshave

    Mercury-Atlas

  • *****
  • Posts: 2,567
  • Joined: 08 Apr 2015

Posted 02 March 2021 - 10:31 PM

If that's true - that the measure changes -- there's no way from the reported errors to know if your guiding improved.  Well, other than if your stars are more round.

The primary value of multi-star guiding, as I understand it, is to act as a low-pass filter for seeing effects. If the high-frequency effects of seeing are reduced, the number of false guiding corrections in response to seeing is also reduced. The RMS error should drop but that will not translate directly to a reduction of star roundness or FWHM of the same proportion. Guiding will be improved because a larger fraction of the guiding corrections will be in response to real tracking error rather than random movements from seeing. Since guide star movement from seeing is random, the resulting guiding corrections will only make guiding worse. This is the aspect of guiding that MSG improves. For me, on the Tak, my RMS error may be improved by as much as 50%, but I only see about a 10% reduction in image FWHM.

 

Tim


  • psandelle likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: Astrophotography, CCD, DSO, Equipment, Imaging, Refractor



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics