#1  
Old 18-04-2016, 05:52 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
guiding update rate

Hi

I measured the performance of my EQ8 system in very good seeing and with phd2 guiding exposures of 0.5, 1, 2, 5 and 10 seconds. The performance measure was the FWHM (from PI) of 2 minute subs, with the average of 3 such subs taken at each guiding exposure.

To ensure that seeing variability did not influence the results too much, the process used was to take an image for each of the 5 guide exposure times and then go through the same sequence twice more. Focus was not changed during the process, but there was no evidence of any drift (the results at each guide exposure were quite consistent).

The results contradict the conventional wisdom that short guide exposures encourage "chasing the seeing" and reduced resolution. There is fairly convincing evidence that the shortest guide exposure (0.5 seconds) resulted in quite significantly better resolution - maybe I should try 1/4 second guide exposures. The data quantify the subjective impression that the subs taken at short guide exposures were noticeably sharper than those taken with long guide exposures.

thanks for looking. regards Ray
Attached Thumbnails
Click for full-size image (seeing.jpg)
23.1 KB87 views

Last edited by Shiraz; 18-04-2016 at 06:03 PM.
Reply With Quote
  #2  
Old 18-04-2016, 06:13 PM
codemonkey's Avatar
codemonkey (Lee)
Lee "Wormsy" Borsboom

codemonkey is offline
 
Join Date: Jul 2013
Location: Kilcoy, QLD
Posts: 2,058
That's consistent with my experiences. I think it basically depends how quickly your mount bounces around.

Long guide exposures are for people with extremely smooth mounts.

The trick to working with short guide exposures it to get your min motion honed in. This will change on a day to day basis, basically the better the seeing the smaller the min motion.
Reply With Quote
  #3  
Old 18-04-2016, 06:14 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Good on you for trying it out, Ray! I still tend to guide as quickly as my guide star SNR will allow but I've never spent the time to measure experimentally whether it helped. It does make my little ACP guiding squiggle look tighter...

Cheers,
Rick.
Reply With Quote
  #4  
Old 18-04-2016, 06:23 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by codemonkey View Post
The trick to working with short guide exposures it to get your min motion honed in. This will change on a day to day basis, basically the better the seeing the smaller the min motion.
I set my min and max move based on the image scale (min move no less than 1/4 image scale and max move at least 2x image scale) and then tweak the Aggressiveness to account for nightly variations in seeing. That's derived from the gospel of Roland Christensen (Mr. Astro-Physics) and has always worked well for me...

Cheers,
Rick.
Reply With Quote
  #5  
Old 18-04-2016, 06:36 PM
codemonkey's Avatar
codemonkey (Lee)
Lee "Wormsy" Borsboom

codemonkey is offline
 
Join Date: Jul 2013
Location: Kilcoy, QLD
Posts: 2,058
Quote:
Originally Posted by RickS View Post
I set my min and max move based on the image scale (min move no less than 1/4 image scale and max move at least 2x image scale) and then tweak the Aggressiveness to account for nightly variations in seeing. That's derived from the gospel of Roland Christensen (Mr. Astro-Physics) and has always worked well for me...

Cheers,
Rick.
The basis for my thinking is simply that if the centroid is bouncing around +/- 0.2px due to seeing fluctuations, then you can't trust movements within those bounds, only once it goes outside of that can you be sure there's something other than atmospheric turbulence at play, which is why I adjust the min motion to account for seeing.

I only tweak the aggression when I see under correction (large swings of corrections in the same direction) or over correction (zig-graph, with frequent alternating corrections).

Damping the aggression to compensate for seeing may also work well, clearly it does for you, just doesn't make as much sense to me so hasn't been the way that I've tackled it.
Reply With Quote
  #6  
Old 18-04-2016, 09:33 PM
rustigsmed's Avatar
rustigsmed (Russell)
Registered User

rustigsmed is offline
 
Join Date: Mar 2012
Location: Mornington Peninsula, Australia
Posts: 3,950
That is interesting, thanks for sharing Ray. Of course a guide star isn't always so bright to allow short ones but good to know short as possible is the go.
Reply With Quote
  #7  
Old 18-04-2016, 09:57 PM
Atmos's Avatar
Atmos (Colin)
Ultimate Noob

Atmos is offline
 
Join Date: Aug 2011
Location: Melbourne
Posts: 6,980
With my guide scope I know I can guide on 0.1s exposures so maybe I should try super short subs
Reply With Quote
  #8  
Old 18-04-2016, 10:56 PM
astronobob's Avatar
astronobob (Bob)
Casual Cosmos Capturer

astronobob is offline
 
Join Date: Jun 2010
Location: Gold Coast SE QLD
Posts: 4,190
I know not of the detailed guiding you guys do but one thing I question is why, when you say Ray quote "EQ8 system in very good seeing and with phd2" I ask, would it not defeat the purpose if your testing in as said, Very Good seeing ? Why would you not do such a test when ' seeing' is mediocre or worse ?
If you get my ' Drift" :-)
Respect & Rgrds !
Reply With Quote
  #9  
Old 19-04-2016, 06:26 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by RickS View Post
I set my min and max move based on the image scale (min move no less than 1/4 image scale and max move at least 2x image scale) and then tweak the Aggressiveness to account for nightly variations in seeing. That's derived from the gospel of Roland Christensen (Mr. Astro-Physics) and has always worked well for me...

Cheers,
Rick.
Quote:
Originally Posted by codemonkey View Post
The basis for my thinking is simply that if the centroid is bouncing around +/- 0.2px due to seeing fluctuations, then you can't trust movements within those bounds, only once it goes outside of that can you be sure there's something other than atmospheric turbulence at play, which is why I adjust the min motion to account for seeing.

I only tweak the aggression when I see under correction (large swings of corrections in the same direction) or over correction (zig-graph, with frequent alternating corrections).

Damping the aggression to compensate for seeing may also work well, clearly it does for you, just doesn't make as much sense to me so hasn't been the way that I've tackled it.
Thanks Rick and Lee. with aggression, hysteresis, min motion, guide rate, bearing stiction, drive stiffness, load imbalance, processing latency, scintillation etc. there seems to be too many (possibly dependent) variables to come up with a single optimum approach, so it is great to know what works.

Quote:
Originally Posted by rustigsmed View Post
That is interesting, thanks for sharing Ray. Of course a guide star isn't always so bright to allow short ones but good to know short as possible is the go.
I use a fast guide scope and can almost always find a 0.5 sec guide star - oag would be different.

Quote:
Originally Posted by Atmos View Post
With my guide scope I know I can guide on 0.1s exposures so maybe I should try super short subs
would be worth trying Colin. I also did some tests with varying sub lengths, but stuffed it up. will redo and post again.

Quote:
Originally Posted by astronobob View Post
I know not of the detailed guiding you guys do but one thing I question is why, when you say Ray quote "EQ8 system in very good seeing and with phd2" I ask, would it not defeat the purpose if your testing in as said, Very Good seeing ? Why would you not do such a test when ' seeing' is mediocre or worse ?
If you get my ' Drift" :-)
Respect & Rgrds !
good point Bob. I will redo the test when the seeing is crappy. I suspect that the results may also be specific to the mount type - the EQ8 has 8 inch wheels and probably generates less mechanical noise than smaller ones, so it may possibly do better with short guide exposures - but that is entirely conjecture.

regards Ray

Last edited by Shiraz; 19-04-2016 at 07:49 AM.
Reply With Quote
  #10  
Old 19-04-2016, 10:57 AM
alocky's Avatar
alocky (Andrew lockwood)
PI popular people's front

alocky is offline
 
Join Date: Aug 2010
Location: perth australia
Posts: 1,291
Hi Ray, it's good to challenge convention and especially in such a scientific manner. Do you get the same trend if you plot the ellipticity of the stars rather than fwhm? I'd be curious if the poorer results coincided with an elongation in RA.
Cheers
Andrew.
Reply With Quote
  #11  
Old 19-04-2016, 01:04 PM
rustigsmed's Avatar
rustigsmed (Russell)
Registered User

rustigsmed is offline
 
Join Date: Mar 2012
Location: Mornington Peninsula, Australia
Posts: 3,950
Quote:
Originally Posted by Shiraz View Post
I use a fast guide scope and can almost always find a 0.5 sec guide star - oag would be different.


regards Ray
that makes sense ray for some reason I thought you had an OAG;
still with an OAG it is rare for me to go more than 1 second (at f4 and a lodestar) occasionally it's required but I could definitely go to 0.5seconds atleast half of the time I reckon no problem. I can't recall going shorter than that.
Reply With Quote
  #12  
Old 20-04-2016, 06:04 AM
DavidTrap's Avatar
DavidTrap (David)
Really just a beginner

DavidTrap is offline
 
Join Date: Jun 2009
Location: Brisbane
Posts: 3,032
Quote:
Originally Posted by RickS View Post
I set my min and max move based on the image scale (min move no less than 1/4 image scale and max move at least 2x image scale) and then tweak the Aggressiveness to account for nightly variations in seeing. That's derived from the gospel of Roland Christensen (Mr. Astro-Physics) and has always worked well for me...

Cheers,
Rick.
Rick,

Is the image scale you refer to the arcsec/pixel of the imaging camera or the guide camera??

Ta
DT
Reply With Quote
  #13  
Old 20-04-2016, 06:42 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Interesting results Ray and as usual you have done a thought provoking post.

Did you know PDH2 has a guide assistant that actually measures the seeing and PE and recommends the settings from that? Its quite good.

I have used it on both PME and AP1600. Generally it recommends the same exposure length each night of between 2 and 4 seconds.

I had a Tak NJP mount and it typically gave better results with short guide exposures like 1 second if I could get it.

PME seemed to get best results around 4-6 seconds and AP1600 around 4 seconds.

I think its highly mount dependent with the smoother better mounts with larger gears etc doing better with longer guide exposures and lower aggression and those with more PE shorter exposures as there is more moving around from the PE. I have often got best results with aggressiveness at 3 on mounts like PMX, PME but usually I use about 5.

Roland also recommends to increase aggressiveness on nights of good seeing so the number of secs before the mount is corrected back is lower.

I tend to tighten up max move as sometimes you see a bit of seeing induced larger corrections that seem to be outliers and worsen the guiding for a few secs. So now the mount is correcting the corrections.

A night of good seeing also may be skewing your results. Also mounts like PME/AP use a different system to EQ8 and its direct guide or pulse guide (AP version) rather than relays and this further improves guiding.
Short guide exposures on a night of worse seeing may in fact chase the seeing much more but in good seeing short exposures would work better.
FWHM may also not be the best measure as seeing fluctuations could account for some of the results although there is no arguing there is a definite trend there in your results. PI has a measure of eccentricity of stars you could use if you have PI. If not you can download a trial of PI.
At the end of the day round stars is what you are after so it would be the better measure. Still affected by seeing though.

There is a fabulous video at the PHD site where one guy talks about guiding and all its aspects. I found it very educational and the best resource around.

Another possibility is that using a guide scope differential flexure may also favour shorter exposures as that occurs usually over longer time periods. It may not and its just a thought.

Autoguiding is not often talked about yet it is critical to good results and a continual source of problems. So it would be good for a long thread on it and see what others use and what results they get.

It also may come down to software too. I have found CCDsoft my main autoguiding program over the years mainly because it was the first software I used from my 1st camera which was an SBIG.

The Sky X also seems to be the guts of CCDsoft transferred over to Sky X with a few extra features. I use that now sometimes as my AP mount needs pulse guide and CCDsoft does not offer that as an option but Sky X does.

PHD2 is very sophisticated. I am not sure if it gives better results than CCDsoft though. It will do Pulse Guiding or Direct Drive via an ASCOM plug in which is free.

A lot use Maxim. It has more bells and whistles and I am not sure if they ever mastered multiple star guiding or not but it sounded interesting. I have never used it.

Using a guide scope usually means flexure to some degree although I have managed to get 10 minute subs with round stars on a PME with 4 second guide exposures and 2 metres focal length. But it usually means elongated stars beyond 3-5 minute exposure times plus a guide scope tends to unbalance your scope at different angles due to raising the centre of gravity.

Getting round stars at 10minute exposures has always been the starting point for any astrophotography. Its quite a hard target no matter how good your gear is!

Not really on topic but probably interesting to a lot is I measured the PE of both my AP1600 and PME recently using the same setup. The AP1600 was considerably better at around 1.5 arc secs error often less. The best I saw was .6 arc secs and .8 is common. The PME was more like 3.5. Both give round stars when properly setup though. The PME has been a reliable performer for me so I was surprised the AP guide graph was so much smoother and less peaky than the PME's (about a 10 year old PME which may have better PE than modern versions which state higher PE in their specs). This is with PEC turned on. Home position on the PME is a hell of a useful feature though.

Another point I would make is the guide star selected. I find it can change guide errors a lot. If I get increased errors the first thing I do is select another star perhaps one less saturated and nice and round. Also tighten up guide star focus which improves results a bit. Rounder stars should mean more accurate centroid calculations than an odd shaped star. Sometimes when using OAG depending on the scope, guide stars near the outer edge of the FOV are distorted and aren't the best choice for a guide star. This can be worse with some scopes compared to others. Refractors don't usually have a problem but some scopes do give distorted guide stars in the corners and sides.

Greg.

Last edited by gregbradley; 20-04-2016 at 06:53 AM.
Reply With Quote
  #14  
Old 20-04-2016, 10:13 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by alocky View Post
Hi Ray, it's good to challenge convention and especially in such a scientific manner. Do you get the same trend if you plot the ellipticity of the stars rather than fwhm? I'd be curious if the poorer results coincided with an elongation in RA.
Cheers
Andrew.
Hi Andrew. yes, there is elongation in RA - easiest to see in the attached image comparison where RA is aligned roughly diagonal from bottom left to top right.


Quote:
Originally Posted by gregbradley View Post
Interesting results Ray and as usual you have done a thought provoking post.

Did you know PDH2 has a guide assistant that actually measures the seeing and PE and recommends the settings from that? Its quite good.

I had some difficulty understanding how the assistant works, so have not used it.

I think its highly mount dependent

FWHM may also not be the best measure as seeing fluctuations could account for some of the results although there is no arguing there is a definite trend there in your results. PI has a measure of eccentricity of stars you could use if you have PI. If not you can download a trial of PI.
At the end of the day round stars is what you are after so it would be the better measure. Still affected by seeing though. Getting round stars at 10minute exposures has always been the starting point for any astrophotography. Its quite a hard target no matter how good your gear is!

my criteria for good guiding is for round and small stars - I can easily get just round stars by defocusing a little or in post processing, but the only way to get finest detail is with good guiding. FWHM is directly affected by seeing, so I ran the experiment to make sure that seeing was very close to the same for each set of comparison images.

Another possibility is that using a guide scope differential flexure may also favour shorter exposures as that occurs usually over longer time periods. It may not and its just a thought.

The only exposures that were changed were the guide exposures. The light exposures were all the same at 2 minutes - my system has no measurable flexure up to about 10 minutes.

Not really on topic but probably interesting to a lot is I measured the PE of both my AP1600 and PME recently using the same setup. The AP1600 was considerably better at around 1.5 arc secs error often less. The best I saw was .6 arc secs and .8 is common. The PME was more like 3.5. Both give round stars when properly setup though. The PME has been a reliable performer for me so I was surprised the AP guide graph was so much smoother and less peaky than the PME's (about a 10 year old PME which may have better PE than modern versions which state higher PE in their specs). This is with PEC turned on. Home position on the PME is a hell of a useful feature though.

interesting results. My criterion for mount quality is that it must be just good enough to do the job. Clearly your two mounts are at least that good.

Another point I would make is the guide star selected.

The guide stars for this test were chosen to be in the centre of the frame and the SNR was about 10 for all tests

Greg.
responses in italics Greg.

regards Ray
Attached Thumbnails
Click for full-size image (stars.jpg)
74.8 KB41 views
Reply With Quote
  #15  
Old 20-04-2016, 10:29 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Thanks Ray.

The assistant is easy to use. You simply click start and wait say 2 minutes for the results to settle down and then click on finish and it gives recommendations for min/max move and exposure lengths. You can click on accept and it implements them.

Thanks for the test. Always illuminating. There's always trial and error with guide settings. I did some of this very type of check just the other night. In my case on an AP1600 4 seconds seemed optimum with 3 metre focal length and a MMOAG. Faster exposures got into correcting the corrections. Seeing wasn't that great at the time. I used to use 6 seconds on the PME and it seemed to be the best for it using a MMOAG at 3 metres focal length.

Greg.
Reply With Quote
  #16  
Old 20-04-2016, 10:57 AM
jenchris's Avatar
jenchris (Jennifer)
Registered User

jenchris is offline
 
Join Date: Nov 2010
Location: Ormeau Gold Coast
Posts: 2,067
I'm not sure how to compare the results above with my phd2 graphs.
My min mov is set to .14 and .09. My exposure to 2 seconds on a qhy5 on a 102swachro.
Graph attached with target and fwhm graph.
Attached Thumbnails
Click for full-size image (20160406_222258.jpg)
124.6 KB48 views
Reply With Quote
  #17  
Old 20-04-2016, 09:05 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by gregbradley View Post
Thanks Ray.

The assistant is easy to use. You simply click start and wait say 2 minutes for the results to settle down and then click on finish and it gives recommendations for min/max move and exposure lengths. You can click on accept and it implements them.

Greg.
thanks Greg. the only problem I have with the advisor is that I have no idea what is under the bonnet and how to get the best from it - for example, if my best imaging exposure is 0.5 arc sec, but 2 sec is used for the calibration run, how can it possibly sample fast enough to end up suggesting 0.5 seconds. I don't see any way that it can, but the documentation does not explain it well enough that I can be sure. Rather than accept something so "black box", I decided to test the effects of guide exposure in another way. That is not a criticism of the software - it's just that I would prefer to understand how it actually works.

Quote:
Originally Posted by jenchris View Post
I'm not sure how to compare the results above with my phd2 graphs.
My min mov is set to .14 and .09. My exposure to 2 seconds on a qhy5 on a 102swachro.
Graph attached with target and fwhm graph.
the graph given earlier is a plot of the star sizes (in arcseconds at full width half maxiumum) taken from a series of 2 minute images (taken with an astro camera) and for different guide exposures. It shows how the star size gets smaller as the guide exposure is reduced.
Your screen shot shows the error plot for phd, which is a measure of how much error phd was correcting at any time over a past period. This is not the same as the resolution plot, since the data in your graph comes from the guide camera, not an imaging camera. There is a tenuous tie up between the two types of data (reduced guide errors may possibly result in smaller stars), but there is no direct comparison.

having said that though, your plot looks good to me, with quite low and consistent guiding error. You might try using shorter exposures and see if you get smaller stars in DSLR images, but it already looks like it is guiding OK. FWIW, my EQ6 guides well at 1.5sec guide exposure in most conditions - that is close to your chosen value, so either that is a good choice or we are both wrong.

Last edited by Shiraz; 20-04-2016 at 09:25 PM.
Reply With Quote
  #18  
Old 20-04-2016, 09:23 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
One of the software writers went over roughly how it works. Its in that astro channel you tube video presentation on the PHD website. He mentions the best thing you can do for improving your autoguiding is to turn off the autoguider and observe how your mount performs. So I think its measuring drift in RA and Dec, PE, flex etc. It may also be measuring FWHM to get an idea of the seeing.

I have found it to work and be a good tool as its measuring the conditions on the night rather than a canned approach which may not match the current conditions.

Greg.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 06:26 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Bintel
Advertisement
Testar
Advertisement