Go Back   IceInSpace > Equipment > Astrophotography and Imaging Equipment and Discussions

Reply
 
Thread Tools Rate Thread
  #1  
Old 17-10-2020, 11:05 AM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Plate solving & sync question in EQMOD

So, as far as I understand it, a limitation of syncing platesolved coordinates to EQMOD is that it won't sync if you are too close to the pole. For example, when setting up each night I will typically Polar Align in SharpCap (Which obviously uses Platesolving to work) and then open APT and use the PointCraft function to PS and sync. Now, if you just Platesolve whilst still pointing at the Pole, EQMOD will refuse to sync - but I always have a 100% success rate platesolving at the pole, so this is annoying. What I typically find I have to do then is slew around the sky, until I land on a target far enough away from the pole that can be successfully platesolved.

Sometimes, for reasons I am not clear on, it can take a few attempts on a few targets before I can get a successful PS once away from the pole. There does not appear to be any technical limitation in successfully platesolving as Pointcraft (using both ASPS and PlateSolve2) will work fine when pointed at the pole, as does SharpCap. But once I start slewing away, it can take a while sometimes before I can successfully Platesolve, and then sync.

Am I doing something wrong here? Obviously the point of PS+Sync is so that the mount knows exactly where it's pointing, allowing me to slew to targets accurately, but the limitation on syncing coordinates close to the pole is very frustrating.

Typically I will run the PA platesolve routine, as well as initial APT PS+Sync through my guide scope, as focusing my imaging scope without a bright enough star can be tricky, and finding a bright enough star can be tricky without first syncing the mount to correct coordinates

Any tips for how to speed up the initial routine of getting the mount properly synced to where it is pointing? Some nights I get lucky and it will PS on the first section of the sky I slew to, others it can take many attempts across different areas of the sky before it decides to cooperate.

Cheers
Reply With Quote
  #2  
Old 17-10-2020, 01:52 PM
Startrek (Martin)
Registered User

Startrek is offline
 
Join Date: Dec 2017
Location: Sydney and South Coast NSW
Posts: 5,990
I don’t plate solve but I’ve been using EQMOD with Stellarium for 3 years now
Just a thought, maybe your meridian limits are set in EQMOD ( say 5 or 10 degrees after the meridian ) and this may be in conflict with your PS software when close to the meridian or SCP ??
Just a hunch ?
Reply With Quote
  #3  
Old 17-10-2020, 03:27 PM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Craig,
Not sure I can help, but I do use EQMod (and Astroart for plate solving and syncing)
If you're using SharpCap to refine your PA I assume you're tweaking the Alt/Az to give you a final correction?
Ok, at this stage you have the mount aligned and the scope pointing to the SGP.....this is the home position for EQMod.
When you open EQMod it automatically assumes (unless you've set a different Custom park position) it's pointing at the SGP, so there's no need to have a sync at that point.
The next step is to obviously slew to your intended target. If it's not 100% centred, then do a Plate Solve and sync which will be recognised by EQMod.
As you change target during the night, you can continue to apply PS and add sync points to improve your pointing model.
Does that make any sense????
Reply With Quote
  #4  
Old 18-10-2020, 03:01 PM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Quote:
Originally Posted by Startrek View Post
I don’t plate solve but I’ve been using EQMOD with Stellarium for 3 years now
Just a thought, maybe your meridian limits are set in EQMOD ( say 5 or 10 degrees after the meridian ) and this may be in conflict with your PS software when close to the meridian or SCP ??
Just a hunch ?
Thanks for the reply. As I understand it, and I'm happy to be corrected on this, it is a software limitation of EQMOD that you cannot sync coordinates when close to the pole. That said, sorry, I'm not sure I quite understand what you are saying here? If my meridian limits are set in EQMOD, this would allow the mount to track only as far as those limits before initiating a flip, correct?

Quote:
Originally Posted by Merlin66 View Post
Craig,
Not sure I can help, but I do use EQMod (and Astroart for plate solving and syncing)
If you're using SharpCap to refine your PA I assume you're tweaking the Alt/Az to give you a final correction?
Ok, at this stage you have the mount aligned and the scope pointing to the SGP.....this is the home position for EQMod.
When you open EQMod it automatically assumes (unless you've set a different Custom park position) it's pointing at the SGP, so there's no need to have a sync at that point.
The next step is to obviously slew to your intended target. If it's not 100% centred, then do a Plate Solve and sync which will be recognised by EQMod.
As you change target during the night, you can continue to apply PS and add sync points to improve your pointing model.
Does that make any sense????
Yes, I do use SharpCap for my PA, so I am tweaking the Alt/Az yes.
To clarify, when I say I want a sync initially, it is because after PA takes place the "home position" isn't pinpoint perfect on the SCP. (That is to say, the mount is aligned by alt/az, but the RA/DEC are not being aligned in SharpCap as I understand it.) So the reason I need to PS and sync after PA is that if I complete my PA routine and attempt to slew to a target, invariably it won't accurately slew to that target as where the mount thinks it is pointing is not exactly where it is pointing. After successfully PS+syncing, I can slew with perfect accuracy to any target.

That then leads me to the crux of the issue, which you touch on - "If it's not 100% centred, then do a Plate Solve and sync which will be recognised by EQMod." This step has mixed success when PS away from the pole. I can start APT, run the Pointcraft module and have 100% success night after night PSing my coordinates near the pole - it just won't sync them. Let's say I want to image M8 - I then slew to where M8 "should be", invariably find it isn't there because I haven't yet synced the mount to PS'ed coordinates, and then the fun begins. Sometimes I can PS successfully whatever area of the sky I end up in, sync, and then off I go. Other times it just won't PS successfully, and I end up having to slew all over the sky until I find an area it can successfully platesolve.

So my problem is basically two-fold:

1) With a 100% success rate in Platesolving near the Pole, this is useless as EQMOD refuses to sync coordinates near the pole (if there is a way around this I'd love to hear it)

2) Despite no apparent technical limitation to Platesolving (given 100% success rate near the pole) once I start slewing away from the pole my success rate drops dramatically. Sometimes it can take 10-15 minutes before I land on an area of the sky the software can successfully platesolve - immensely frustrating.

Thanks for the assistance so far.
Reply With Quote
  #5  
Old 18-10-2020, 03:06 PM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Quote:
Originally Posted by Startrek View Post
I don’t plate solve but I’ve been using EQMOD with Stellarium for 3 years now
Just a thought, maybe your meridian limits are set in EQMOD ( say 5 or 10 degrees after the meridian ) and this may be in conflict with your PS software when close to the meridian or SCP ??
Just a hunch ?
Thanks for the reply. As I understand it, and I'm happy to be corrected on this, it is a software limitation of EQMOD that you cannot sync coordinates when close to the pole. That said, sorry, I'm not sure I quite understand what you are saying here? If my meridian limits are set in EQMOD, this would allow the mount to track only as far as those limits before initiating a flip, correct?

Quote:
Originally Posted by Merlin66 View Post
Craig,
Not sure I can help, but I do use EQMod (and Astroart for plate solving and syncing)
If you're using SharpCap to refine your PA I assume you're tweaking the Alt/Az to give you a final correction?
Ok, at this stage you have the mount aligned and the scope pointing to the SGP.....this is the home position for EQMod.
When you open EQMod it automatically assumes (unless you've set a different Custom park position) it's pointing at the SGP, so there's no need to have a sync at that point.
The next step is to obviously slew to your intended target. If it's not 100% centred, then do a Plate Solve and sync which will be recognised by EQMod.
As you change target during the night, you can continue to apply PS and add sync points to improve your pointing model.
Does that make any sense????
Yes, I do use SharpCap for my PA, so I am tweaking the Alt/Az yes.
To clarify, when I say I want a sync initially, it is because after PA takes place the "home position" isn't pinpoint perfect on the SCP. (That is to say, the mount is aligned by alt/az, but the RA/DEC are not being aligned in SharpCap as I understand it.) So the reason I need to PS and sync after PA is that if I complete my PA routine and attempt to slew to a target, invariably it won't accurately slew to that target as where the mount thinks it is pointing is not exactly where it is pointing. After successfully PS+syncing, I can slew with perfect accuracy to any target.

That then leads me to the crux of the issue, which you touch on - "If it's not 100% centred, then do a Plate Solve and sync which will be recognised by EQMod." This step has mixed success when PS away from the pole. I can start APT, run the Pointcraft module and have 100% success night after night PSing my coordinates near the pole - it just won't sync them. Let's say I want to image M8 - I then slew to where M8 "should be", invariably find it isn't there because I haven't yet synced the mount to PS'ed coordinates, and then the fun begins. Sometimes I can PS successfully whatever area of the sky I end up in, sync, and then off I go. Other times it just won't PS successfully, and I end up having to slew all over the sky until I find an area it can successfully platesolve.

So my problem is basically two-fold:

1) With a 100% success rate in Platesolving near the Pole, this is useless as EQMOD refuses to sync coordinates near the pole (if there is a way around this I'd love to hear it)

2) Despite no apparent technical limitation to Platesolving (given 100% success rate near the pole) once I start slewing away from the pole my success rate drops dramatically. Sometimes it can take 10-15 minutes before I land on an area of the sky the software can successfully platesolve - immensely frustrating.

Maybe this entire thread can be better summed as me asking: how can I improve my platesolving success rate when all the software is correctly set up already? (demonstrated by the fact I can PS successfully on the pole, 100% of the time.)
Reply With Quote
  #6  
Old 18-10-2020, 03:35 PM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Craig,
""I complete my PA routine and attempt to slew to a target, invariably it won't accurately slew to that target as where the mount thinks it is pointing is not exactly where it is pointing."
OK
EQMOD (I don't know about the other programs you mention) doesn't require a plate solve at SCP.


It is reliant on the scope being set to the "home position" (ie pointing towards the SCP. This doesn't have to be 110% accurate.


The "encoders" used by EQMOD then use the "home position" as a starting point for the subsequent slew. When you reach the first target position ( according to EQMOD) there may be a error ( the home position wasn't 100%, flexure/ movement of the scope etc etc). By then moving the scope to the actual target (handcontroller/ plate solving) and syncing on that point you are providing corrections to the EQMOD alignment model which is then applied to subsequent slews.

For accurate slews you really should have a series of syncs which surround your target area(s).

Once you have a good set of sync points they can be re-used on subsequent nights....

All this is well covered in the EQMOD manual (in the EQMOD download folder)
Reply With Quote
  #7  
Old 18-10-2020, 03:39 PM
jahnpahwa (JP)
Registered User

jahnpahwa is offline
 
Join Date: Apr 2020
Location: Canberra, AUS
Posts: 593
Craig, my routine is exactly the same as yours except I use my main scope images for PS.



One thing that might be mucking you around here is the focal length setting in APT in "tools" tab, down the bottom. Do you have this set to the focal length of your imaging scope or guider? When i swap between OTAs for imaging, I need to change the FL here, or else my PS will not work (455 to 1000). I assume you are getting the feed from your imaging camera into APT? I'd definitely be using my imaging OTA for APT feed, and therefore platesolving. I use mine for sharpcap too.



So, my advice would be to find a way to approximate your focus point before setting up (maybe take a pic of the focuser draw tube during a session once you've focussed it?). It only needs to be close enough so that you can live view and dial it in with APT or sharpcap once you've turned it on. Once you're focused this will all be no problem at all. Maybe changing exposure times in either software will help with the solving/focussing intel too.



The other thing I'd mention is that, like you I slew back to home in APT after PA in sharpcap, then solve and sync there, then go straight to a nearby bright star (Rigel Kent), solve and sync and GOTO again, and it should be bang on. I focus properly there. If I go straight from a sync at the SCP to M8 or similar, I am sometimes a way off... I think you need to give the map a bit of space from the pole so it can chew on where it is, but if you go half way across the sky straight away, it can be miles off.


I hope that helps?
If not, I'd be happy to PM or chat!
Reply With Quote
  #8  
Old 18-10-2020, 04:48 PM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Quote:
Originally Posted by Merlin66 View Post
Craig,
""I complete my PA routine and attempt to slew to a target, invariably it won't accurately slew to that target as where the mount thinks it is pointing is not exactly where it is pointing."
OK
EQMOD (I don't know about the other programs you mention) doesn't require a plate solve at SCP.


It is reliant on the scope being set to the "home position" (ie pointing towards the SCP. This doesn't have to be 110% accurate.


The "encoders" used by EQMOD then use the "home position" as a starting point for the subsequent slew. When you reach the first target position ( according to EQMOD) there may be a error ( the home position wasn't 100%, flexure/ movement of the scope etc etc). By then moving the scope to the actual target (handcontroller/ plate solving) and syncing on that point you are providing corrections to the EQMOD alignment model which is then applied to subsequent slews.

For accurate slews you really should have a series of syncs which surround your target area(s).

Once you have a good set of sync points they can be re-used on subsequent nights....

All this is well covered in the EQMOD manual (in the EQMOD download folder)
Thanks for the additional info. To address the points you've raised -


The "encoders" used by EQMOD then use the "home position" as a starting point for the subsequent slew. When you reach the first target position ( according to EQMOD) there may be a error ( the home position wasn't 100%, flexure/ movement of the scope etc etc). By then moving the scope to the actual target (handcontroller/ plate solving) and syncing on that point you are providing corrections to the EQMOD alignment model which is then applied to subsequent slews.


Yep, this bit I do understand. I think my problem is that sometimes the home position is clearly out by enough that where I end up slewing to is far enough away from my desired target that getting it into my frame, and then syncing to it, is impossible. For example, let's say I am at the home position and wish to slew to Jupiter to focus on (I find it makes a good focus target.) I execute the slew and unless Jupiter happens to be at least partially in my frame (which it often would not be), I am "lost" - I know I am somewhere near by but I end up blindly slewing around trying to find it, wasting time. If I can accurately sync my coordinates before getting started, it ultimately means that when I want to land on a specific target I can with accuracy.

For accurate slews you really should have a series of syncs which surround your target area(s).

Indeed, but it comes back to my problem is that if Platesolving fails, I am blind, so to speak, unable to sync on a target I cannot find.

All this is well covered in the EQMOD manual (in the EQMOD download folder)

I did read through this when starting out but I should probably dive back in and read it again, thanks.

Quote:
Originally Posted by jahnpahwa View Post
Craig, my routine is exactly the same as yours except I use my main scope images for PS.



One thing that might be mucking you around here is the focal length setting in APT in "tools" tab, down the bottom. Do you have this set to the focal length of your imaging scope or guider? When i swap between OTAs for imaging, I need to change the FL here, or else my PS will not work (455 to 1000). I assume you are getting the feed from your imaging camera into APT? I'd definitely be using my imaging OTA for APT feed, and therefore platesolving. I use mine for sharpcap too.



So, my advice would be to find a way to approximate your focus point before setting up (maybe take a pic of the focuser draw tube during a session once you've focussed it?). It only needs to be close enough so that you can live view and dial it in with APT or sharpcap once you've turned it on. Once you're focused this will all be no problem at all. Maybe changing exposure times in either software will help with the solving/focussing intel too.


The other thing I'd mention is that, like you I slew back to home in APT after PA in sharpcap, then solve and sync there, then go straight to a nearby bright star (Rigel Kent), solve and sync and GOTO again, and it should be bang on. I focus properly there. If I go straight from a sync at the SCP to M8 or similar, I am sometimes a way off... I think you need to give the map a bit of space from the pole so it can chew on where it is, but if you go half way across the sky straight away, it can be miles off.


I hope that helps?
If not, I'd be happy to PM or chat!
Thanks for the reply!

To answer your points:

One thing that might be mucking you around here is the focal length setting in APT in "tools" tab, down the bottom. Do you have this set to the focal length of your imaging scope or guider?

An excellent question, and this has mucked me up before. But, on recent evenings I've been sure to check this and it hasn't been the source of my woes!

I assume you are getting the feed from your imaging camera into APT? I'd definitely be using my imaging OTA for APT feed, and therefore platesolving. I use mine for sharpcap too.


I run PA in Sharpcap off my guide scope, and my initial Platesolving in APT is also off the guide scope. I find the imaging OTA near impossible to focus without a bright enough star, and I find it near impossible to accurately slew to a bright enough star without platesolve and sync. Classic chicken and egg My guidescope on the other hand is a cinch to focus on any star.

So, my advice would be to find a way to approximate your focus point before setting up (maybe take a pic of the focuser draw tube during a session once you've focussed it?). It only needs to be close enough so that you can live view and dial it in with APT or sharpcap once you've turned it on.

Yeah, I do do this, I will always set the focus on the imaging OTA to "roughly" where it should be (helps with balancing too) but I find the focuser on the Esprit 80 super touchy, and even being fractionally out from the proper focus spot can leave you wildly out of focus. Even matching up the focuser position exactly to a photo from a previous session usually leaves me a long way off

The other thing I'd mention is that, like you I slew back to home in APT after PA in sharpcap, then solve and sync there,

I'd like to drill in on this actually. You go back to home position in APT after PA, platesolve and sync on the home position? How are you able to sync the PS results in the home position? I consistently, without fail, get an error when I try and sync coordinates platesolved in the home position, which is supposedly an EQMOD limitation (can't remember where I read this though.) As soon as I slew away from home/SCP, assuming a successful platesolve of course I have no issues syncing. So my issue comes down to: 100% success rate platesolving at the pole, 0% success rate syncing at the pole. 30% (maybe?) success rate platesolving away from the pole; 100% success rate syncing away from the pole. What's your trick to sync at home/SCP??

I hope that helps?
If not, I'd be happy to PM or chat!
Very helpful thanks, and I may well take you up on that offer! Much appreciated.
Reply With Quote
  #9  
Old 18-10-2020, 05:13 PM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Craig,
It sounds like your platesolving capabilities are part of the problem.
Have you tried Astrometry.net?
I find Astroart solves pretty well.
If the plate solving software is any good it should be able to solve any good image.....
Reply With Quote
  #10  
Old 18-10-2020, 06:09 PM
SteveInNZ
Registered User

SteveInNZ is offline
 
Join Date: May 2008
Location: Auckland, NZ
Posts: 239
You are asking EQMOD to do something that is illogical.


When the scope is pointing at the pole, the Dec is obviously 90 degrees, but the RA is undefined. Consider what would happen if you asked the scope to move east by 10 degrees when it's pointing at the pole.


If it can't determine an RA value, it can't say where it is and therefore, can't sync.





Steve
Reply With Quote
  #11  
Old 18-10-2020, 06:43 PM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Quote:
Originally Posted by Merlin66 View Post
Craig,
It sounds like your platesolving capabilities are part of the problem.
Have you tried Astrometry.net?
I find Astroart solves pretty well.
If the plate solving software is any good it should be able to solve any good image.....
Yep, something is awry with my PS but I can't figure out what. Astronomy.net has a 100% success rate for me, but the problem is that I have no way of taking it's successful result and syncing it into EQMOD (as far as I know?) so it's kind of useless.

APT doesn't do the PS itself, it lets you plug in other programs that do it. I have used ASPS and PS2 within APT. Both have had their share of success and failure for me.

Quote:
Originally Posted by SteveInNZ View Post
You are asking EQMOD to do something that is illogical.


When the scope is pointing at the pole, the Dec is obviously 90 degrees, but the RA is undefined. Consider what would happen if you asked the scope to move east by 10 degrees when it's pointing at the pole.


If it can't determine an RA value, it can't say where it is and therefore, can't sync.





Steve
Thanks for this. So the illogical action you refer to is me expecting to sync RA/DEC coordinates into the mount when it's pointing at the pole?
Reply With Quote
  #12  
Old 18-10-2020, 07:30 PM
The_bluester's Avatar
The_bluester (Paul)
Registered User

The_bluester is offline
 
Join Date: Feb 2011
Location: Kilmore, Australia
Posts: 3,342
You can sync near the pole (As it is unlikely you will be pointed exactly at the pole) but it is not a good idea as it is error prone. As Steve pointed out, if you are pointed directly at the pole RA is undefined. If you plate solve at the pole and the plate solve result is on the opposite side of the meridian from where EQMOD assumed the scope was it will reject the sync. That one tripped me up getting meridian flips automated at one point due to cone error, after the flip the telescope was pointed just a few minutes east of the meridian and it would fail there. Delaying the flip an extra ten minutes solved it.

Make sure your polar alignment is good, then the park position is reasonably accurate and do your first goto. Then polatesolve and sync. Or even just randomly point off to the east somewhere and plate solve/sync will do, just get away from the pole and meridian.

It was mentioned above about syncing multiple points which builds a pointing model. I Know APT recommends working that way and when I was using APT I had a routine down which captured maybe ten sync points each side of the meridian (Append on sync) before I would start imaging after which gotos were quite accurate, but it still usually took more than one pointing iteration to get right on target.

SGP and Voyager (Which I use now) use a different routine, the sync there in EQMOD speak is "Dialog based" What it does is realign the whole pointing model on the one sync point. I can't recall how SGP does it precisely so I will only mention Voyager here. When you request a precise pointing operation, it issues a goto and when that finishes, it plate solves and syncs the mount, if it has missed the target by more than your defined value, it issues another goto (Much smaller now) and then plate solves again, iterating in typically smaller and smaller steps until it hits the threshold for accuracy. Even with my AZEQ6, it generally took no more than three iterations to be within the limits.

I prefer the dialog based sync, while it is cool to see accurate gotos first time after building the pointing model it had to be done every night.
Reply With Quote
  #13  
Old 18-10-2020, 08:16 PM
jahnpahwa (JP)
Registered User

jahnpahwa is offline
 
Join Date: Apr 2020
Location: Canberra, AUS
Posts: 593
Quote:
Originally Posted by Craig_ View Post

The other thing I'd mention is that, like you I slew back to home in APT after PA in sharpcap, then solve and sync there,

I'd like to drill in on this actually. You go back to home position in APT after PA, platesolve and sync on the home position? How are you able to sync the PS results in the home position? I consistently, without fail, get an error when I try and sync coordinates platesolved in the home position, which is supposedly an EQMOD limitation (can't remember where I read this though.) As soon as I slew away from home/SCP, assuming a successful platesolve of course I have no issues syncing. So my issue comes down to: 100% success rate platesolving at the pole, 0% success rate syncing at the pole. 30% (maybe?) success rate platesolving away from the pole; 100% success rate syncing away from the pole. What's your trick to sync at home/SCP??

Hmmm.... so once I've done the 90deg RA turn for sharpcap, I slew roughly back to the SCP and hit another PS (I always do auto PS, using the planewave set of point data). I think it actually auto syncs everytime you PS. I hit sync manually each time too. It may be a rubbish synch, but its not far to rigel kent, and my understanding is the auto platesolve solves in a spiral out from where it thinks it should be, so that first solve away from the pole usually takes from 10 to 70 seconds... but usually under 25. I wish I knew of the trick I'm doing! If its not working for you, I'd give it a slight kink in Dec when slewing back after PA, just tohrow it off the smallest amount, and that will give it something to chew on and develop a model around.


The other back to basics thing is, how precise are you with levelling your mount? I was using an app for spirit level, but am now using a little clear spririt cube thing, and get things super level. In my mind, being out of level could make it harder for that first solve, and again, going a long way from home will open up that gap wider and wider if your mount is half a degree skewy.



Overall, I think a little move from the pole first up might help, and see that this is also being mentioned by others
Reply With Quote
  #14  
Old 19-10-2020, 06:53 PM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Quote:
Originally Posted by The_bluester View Post
You can sync near the pole (As it is unlikely you will be pointed exactly at the pole) but it is not a good idea as it is error prone. As Steve pointed out, if you are pointed directly at the pole RA is undefined. If you plate solve at the pole and the plate solve result is on the opposite side of the meridian from where EQMOD assumed the scope was it will reject the sync. That one tripped me up getting meridian flips automated at one point due to cone error, after the flip the telescope was pointed just a few minutes east of the meridian and it would fail there. Delaying the flip an extra ten minutes solved it.

Make sure your polar alignment is good, then the park position is reasonably accurate and do your first goto. Then polatesolve and sync. Or even just randomly point off to the east somewhere and plate solve/sync will do, just get away from the pole and meridian.

It was mentioned above about syncing multiple points which builds a pointing model. I Know APT recommends working that way and when I was using APT I had a routine down which captured maybe ten sync points each side of the meridian (Append on sync) before I would start imaging after which gotos were quite accurate, but it still usually took more than one pointing iteration to get right on target.

SGP and Voyager (Which I use now) use a different routine, the sync there in EQMOD speak is "Dialog based" What it does is realign the whole pointing model on the one sync point. I can't recall how SGP does it precisely so I will only mention Voyager here. When you request a precise pointing operation, it issues a goto and when that finishes, it plate solves and syncs the mount, if it has missed the target by more than your defined value, it issues another goto (Much smaller now) and then plate solves again, iterating in typically smaller and smaller steps until it hits the threshold for accuracy. Even with my AZEQ6, it generally took no more than three iterations to be within the limits.

I prefer the dialog based sync, while it is cool to see accurate gotos first time after building the pointing model it had to be done every night.
Thanks heaps for the info. I think your first paragraph probably answers the challenge of why EQMOD rejects my sync near the pole. With respect to your second paragraph:

Make sure your polar alignment is good, then the park position is reasonably accurate and do your first goto. Then polatesolve and sync. Or even just randomly point off to the east somewhere and plate solve/sync will do, just get away from the pole and meridian.

This is, more or less, what I do now. My challenge is that very often after I slew away from the pole, the platesolve fails for reasons unknown - all settings are configured correctly (demonstrated by the successful platesolve at the pole!)

The building of the sync points to improve accuracy makes sense, although for now GOTO accuracy is not so much my primary concern as being able to sync to begin with



Quote:
Originally Posted by jahnpahwa View Post
Hmmm.... so once I've done the 90deg RA turn for sharpcap, I slew roughly back to the SCP and hit another PS (I always do auto PS, using the planewave set of point data). I think it actually auto syncs everytime you PS. I hit sync manually each time too. It may be a rubbish synch, but its not far to rigel kent, and my understanding is the auto platesolve solves in a spiral out from where it thinks it should be, so that first solve away from the pole usually takes from 10 to 70 seconds... but usually under 25. I wish I knew of the trick I'm doing! If its not working for you, I'd give it a slight kink in Dec when slewing back after PA, just tohrow it off the smallest amount, and that will give it something to chew on and develop a model around.


The other back to basics thing is, how precise are you with levelling your mount? I was using an app for spirit level, but am now using a little clear spririt cube thing, and get things super level. In my mind, being out of level could make it harder for that first solve, and again, going a long way from home will open up that gap wider and wider if your mount is half a degree skewy.



Overall, I think a little move from the pole first up might help, and see that this is also being mentioned by others
Thanks for the info! In this sentence "Hmmm.... so once I've done the 90deg RA turn for sharpcap, I slew roughly back to the SCP and hit another PS (I always do auto PS, using the planewave set of point data)." What do you mean by planewave data sorry?

"so that first solve away from the pole usually takes from 10 to 70 seconds... but usually under 25."

Yeah this one has been a mixed bag for me. PS at the pole is usually 3-10s depending on whether I use ASPS or PS2 (ASPS is faster IME) but away from the pole can take 30 seconds or so. I've tried blind solving and can be waiting minutes for no result sometimes.

"If its not working for you, I'd give it a slight kink in Dec when slewing back after PA, just tohrow it off the smallest amount, and that will give it something to chew on and develop a model around."

Interesting, I will try that, thanks!

"The other back to basics thing is, how precise are you with levelling your mount? I was using an app for spirit level, but am now using a little clear spririt cube thing, and get things super level. In my mind, being out of level could make it harder for that first solve, and again, going a long way from home will open up that gap wider and wider if your mount is half a degree skewy. "

I'd like to think I am fairly accurate, I do put the time in to do it right and use a physical spirit level like you mention with a dual-axis bubble.

Would I be correct in saying that from the pole, where I can always successfully PS, the further away from the pole I slew for my initial first PS+Sync, the less likely the chance is for it to work? I should try and only move a little away?
Reply With Quote
  #15  
Old 19-10-2020, 08:26 PM
jahnpahwa (JP)
Registered User

jahnpahwa is offline
 
Join Date: Apr 2020
Location: Canberra, AUS
Posts: 593
Yeah, by planewave data I mean they own the dataset that eqmod uses to solve. I think there are three datasets or methods to chose from? I just use the auto and it seems to chose that one. Its probably not something to think about. If yours is solving at all, it has a dataset to refer to.

Insofar as moving it a bit off the pole, I think a small movement would be a good start, perhaps 3 or so degrees in both dec and RA, in line with Paul's suggestion about direction, too

Sounds like you're very close to having it dialled it!
Reply With Quote
  #16  
Old 20-10-2020, 09:23 PM
The_bluester's Avatar
The_bluester (Paul)
Registered User

The_bluester is offline
 
Join Date: Feb 2011
Location: Kilmore, Australia
Posts: 3,342
If you use PS2 for near solving the time taken to solve will depend entirely on how close to the target you land. It uses a spiral search so every time it moves out a layer the number of index files it uses to solve increases significantly. I found with mine if it was within a degree or two it solved fairly quickly, and then a little further out it was slow and a bit more again was basically unusable.

Using ASPS your pixel size needs to be right and the focal length needs to be very close to spot on. It is worth using the settings assistant to check that you have your focal length bang on as it can make a big difference to time taken for a blind solve using ASPS or even solving at all.
Reply With Quote
  #17  
Old 21-10-2020, 10:09 AM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Quote:
Originally Posted by jahnpahwa View Post
Yeah, by planewave data I mean they own the dataset that eqmod uses to solve. I think there are three datasets or methods to chose from? I just use the auto and it seems to chose that one. Its probably not something to think about. If yours is solving at all, it has a dataset to refer to.

Insofar as moving it a bit off the pole, I think a small movement would be a good start, perhaps 3 or so degrees in both dec and RA, in line with Paul's suggestion about direction, too

Sounds like you're very close to having it dialled it!
Thanks, that helps! I will try this out next session.

Quote:
Originally Posted by The_bluester View Post
If you use PS2 for near solving the time taken to solve will depend entirely on how close to the target you land. It uses a spiral search so every time it moves out a layer the number of index files it uses to solve increases significantly. I found with mine if it was within a degree or two it solved fairly quickly, and then a little further out it was slow and a bit more again was basically unusable.

Using ASPS your pixel size needs to be right and the focal length needs to be very close to spot on. It is worth using the settings assistant to check that you have your focal length bang on as it can make a big difference to time taken for a blind solve using ASPS or even solving at all.
Thanks, that info makes sense. When you say the 'settings assistant' is that within APT or ASPS?

Also, in somewhat of a revelation for me, I've just figured out how to sync to the mount any coordinates I manually input. The astronomy.net Platesolver has frustrated me previously as it has always, 100% of the time, platesolved a result but I've not been able to work out how to take the RA/DEC coordinates the website gives you and feed them into the mount. Today I discovered that under the "GOTO RA/DEC" coordinates box in APT, there is a little Sync button, that will sync to the coordinates entered into that box instead of telling the mount to GoTo them. Hallelujah! That discovery increases my platesolve arsenal of tools and importantly adds one that has, to date, always worked for me. The downside is I need to make sure any imaging site has 4G data to access it
Reply With Quote
  #18  
Old 21-10-2020, 12:56 PM
The_bluester's Avatar
The_bluester (Paul)
Registered User

The_bluester is offline
 
Join Date: Feb 2011
Location: Kilmore, Australia
Posts: 3,342
The settings assistance in ASPS. You will need an image (Preferably FITS) taken with your setup, you first pull some data from the image in the settings assistant, then it sends you to solve the same file with astrometry.net online, after which you copy some data from that solved image into the settings assistant to calculate your precise focal length. Then you can use THAT as your focal length in APT or other imaging software for when it calls on ASPS.

I have not used APT for some time but I am sure you can also do a plate solve (Blind solve or near solve) and sync the mount to that from within APT.
Reply With Quote
  #19  
Old 21-10-2020, 08:56 PM
Craig_
Registered User

Craig_ is offline
 
Join Date: May 2020
Location: Sydney
Posts: 307
Quote:
Originally Posted by The_bluester View Post
The settings assistance in ASPS. You will need an image (Preferably FITS) taken with your setup, you first pull some data from the image in the settings assistant, then it sends you to solve the same file with astrometry.net online, after which you copy some data from that solved image into the settings assistant to calculate your precise focal length. Then you can use THAT as your focal length in APT or other imaging software for when it calls on ASPS.

I have not used APT for some time but I am sure you can also do a plate solve (Blind solve or near solve) and sync the mount to that from within APT.
Thanks, I will look into the settings assistance component. Cheers

And yes, you can platesolve within APT, but APT itself doesn't do it, it just farms out to your chosen program. Those have given me mixed results, hence this thread
Reply With Quote
  #20  
Old 03-11-2020, 08:42 PM
MarkInSpace (Mark)
Registered User

MarkInSpace is offline
 
Join Date: Jul 2020
Location: Melbourne
Posts: 219
Hi Craig,
I use APT and just swapped over computers and had to get the platesolving set up again. Eventually, I got everything ok.
Here are my settings for Pointcraft:
ASTAP - downloaded and installed, pointing to the install location, for me: c:\Program Files\astap
ASTAP does not require any special config.
Use ASTAP for blind solve is NOT selected

PlateSolve2
downloaded and installed, pointing to the install location, for me: c:\platesolve\platesolve2.28
I installed APMM and UCAC3PSdirectly underneath platesolve2.28
Those directories had to be set in Platesolve2.exe
Use PS2 for near solving is NOT selected

ASPS - downloaded and installed, pointing to the install location, for me: c:\Program Files (x86)\Platesolver
You must run platesolver.exe to download the correct directories for your focal length

I hope that helps!!
cheers
Mark
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 07:57 AM.

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