SGP - Platesolving nuisance

[souls33k3r]

Hi All,

I’ve been having this for a while but last night it just got to me so thought I should see if someone has some guidance to solve this issue.

In simple words, platesolving doesn’t work the first time (4/5 failed attempts) but then while my scope is pointing in the same position in the sky where it failed and I run the platesolving again, it works.

I did change the setting to go from binning 1x1 to 2x2 and the exposure time from 5s to 10s but this didn’t make any difference either.

Are my setting correct? 

Please advise.

Thanks in advance.

[Dr_Ju_ju]

These are my settings, which generally work ok, if it fails anywhere, is that it's still too light to get enough stars, & funnily enough after a meridian flip, it can sometimes get its knickers in a twist...

[drjolo]

You may try to increase Search regions number. 

[souls33k3r]

2 minutes ago, Dr_Ju_ju said:

These are my settings, which generally work ok, if it fails anywhere, is that it's still too light to get enough stars, & funnily enough after a meridian flip, it can sometimes get its knickers in a twist...

That made me chuckle because it did have a fit when I last tried to do meridian flip. 

Wouldnt 4x4 will make the area smaller any way? Is that really advisable when what you’re really trying to do is platesolve to figure out where you are by presenting the most available sky?

It was fairly dark when i do my target searches so must be some setting tweaking I reckon.

Search region is a good shout mate. Cheers

[Davey-T]

Can't be much help as I've had a stab at SGP a couple of times then gone back to what I know, Maxim.

As I recall there are more settings somewhere for resolution and maybe scope focal length and it needs to get the approximate coords from somewhere.

Using the Star71 3 seconds binned 2x2 is along enough exposure, PinPoint solves in a couple of seconds so presumably SGP/Platesolve should do the same.

Dave

[souls33k3r]

20 minutes ago, Davey-T said:

Can't be much help as I've had a stab at SGP a couple of times then gone back to what I know, Maxim.

As I recall there are more settings somewhere for resolution and maybe scope focal length and it needs to get the approximate coords from somewhere.

Using the Star71 3 seconds binned 2x2 is along enough exposure, PinPoint solves in a couple of seconds so presumably SGP/Platesolve should do the same.

Dave

I was a very much Maxim user prior to taking the plunge in to the world of SGP. Yeah never had this issue with Pinpoint.

Like you said, i'm sure there are settings elsewhere that i might need to tweak. Just need to now find out what to tweak.

[RayD]

Yep as noted above mate go for max regions, and also double check that you have the right catalogues installed.

As Dave rightly points out, you need to ensure your focal length and pixel scale is set up correctly under the camera tab in the equipment profile (do this using the equipment profile manager otherwise it won't save) as that's where it picks up this information from. 

Finally, if this is your first solve of the evening you may need to blind solve first and then do a normal solve.

[souls33k3r]

3 minutes ago, RayD said:

Yep as noted above mate go for max regions, and also double check that you have the right catalogues installed.

As Dave rightly points out, you need to ensure your focal length and pixel scale is set up correctly under the camera tab in the equipment profile (do this using the equipment profile manager otherwise it won't save) as that's where it picks up this information from. 

Finally, if this is your first solve of the evening you may need to blind solve first and then do a normal solve.

I will surely double check the catalogs. This is the only scope that i have set up so i'm sure when i first installed the catalogs, i did them right.

Focal length and pixel scale is definitely set up correctly.

I will however change the Max region and with regards to blind solve, how do you best suggest i do this? Slew to the target > take a 10s exposure > right click on the image > select Solve image > choose blind solve and then do the normal solve?

The same issue happens when it does meridian flip like @Dr_Ju_ju mentioned. Ideally i would like to make the process seamless if i possible can.

[swag72]

Personally I'd give it more than 4 goes to centre......... My EQ8 takes 4 goes to get it in the right place (to an accuracy of 4 px) Another thing that cold be tripping you up is EQMOD...... You need to make sure that is set to 'dialogue based' 

I also do a blind sync and solve when I first slew the scope so that it knows where it is.

[Davey-T]

Can you fill in the approximate coords for the first solve as per Maxim / PinPoint so it doesn't need to blind solve ?

Dave

[RayD]

18 minutes ago, souls33k3r said:

I will surely double check the catalogs. This is the only scope that i have set up so i'm sure when i first installed the catalogs, i did them right.

Focal length and pixel scale is definitely set up correctly.

I will however change the Max region and with regards to blind solve, how do you best suggest i do this? Slew to the target > take a 10s exposure > right click on the image > select Solve image > choose blind solve and then do the normal solve?

The same issue happens when it does meridian flip like @Dr_Ju_ju mentioned. Ideally i would like to make the process seamless if i possible can.

Do you use CdC?  If so, all I do when I start my imaging session off is to slew to the target area (usually using the slew now command from the sequence I'm using, or, if it's a new target just using CdC slew command), and then do a blind solve using the button on the Scope Centering module.  I think you achieve the same doing what you suggest, but clicking the blind solve button takes the image for you based on your frame and focus settings.  Once this is done you shouldn't need any more blind solves.

[RayD]

3 minutes ago, swag72 said:

You need to make sure that is set to 'dialogue based' 

Good call.  That caught me out with my AZ EQ6!

[souls33k3r]

1 hour ago, swag72 said:

Personally I'd give it more than 4 goes to centre......... My EQ8 takes 4 goes to get it in the right place (to an accuracy of 4 px) Another thing that cold be tripping you up is EQMOD...... You need to make sure that is set to 'dialogue based' 

I also do a blind sync and solve when I first slew the scope so that it knows where it is.

Cheers Sara, what number would you recommend i put in to center? Is the "Dialog based" the one under "User Interface" in EQmod? At the moment i have it set to "Append on Sync".

1 hour ago, RayD said:

Do you use CdC?  If so, all I do when I start my imaging session off is to slew to the target area (usually using the slew now command from the sequence I'm using, or, if it's a new target just using CdC slew command), and then do a blind solve using the button on the Scope Centering module.  I think you achieve the same doing what you suggest, but clicking the blind solve button takes the image for you based on your frame and focus settings.  Once this is done you shouldn't need any more blind solves.

No i don't use CdC, i do however have Stellarium but it's super sluggish and only use it for planetary objects. Saying this, when i do use the slew option within SGP, it's almost always pointing around the target any way so it does get me close enough. This blind solve button, is this the one on "Scope Centering" docking module? If so then all what i remember seeing on there are "Solve & Sync" and "Blind Sync" options. Any thing to minimize the process.

1 hour ago, Davey-T said:

Can you fill in the approximate coords for the first solve as per Maxim / PinPoint so it doesn't need to blind solve ?

Dave

When you do the frame and focus, that puts in the coordinates of the target if i understood your question correctly mate.

 

[swag72]

I would recommend you use whatever accuracy in the plate solving that you are happy with...... I'm not happy with 50, so set it to 4! The EQ8 takes a couple more goes to get there than the Mesu, but they both work faultlessly.

Append on  sync may well have been your issue..... yes dialogue based under the user interface ??

[souls33k3r]

3 minutes ago, swag72 said:

I would recommend you use whatever accuracy in the plate solving that you are happy with...... I'm not happy with 50, so set it to 4! The EQ8 takes a couple more goes to get there than the Mesu, but they both work faultlessly.

Append on  sync may well have been your issue..... yes dialogue based under the user interface ??

Cheers Sara, i will change it to the append on sync first and then start messing about with the numbers on centering.

As always, you lot a legends ... a massive help.

[swag72]

Change it to dialogue based...... 

[souls33k3r]

8 minutes ago, swag72 said:

Change it to dialogue based...... 

Sorry sorry that's what i meant. I'm like nemo these days.

[chrisshillito]

8 minutes ago, souls33k3r said:

Is the "Dialog based" the one under "User Interface" in EQmod? At the moment i have it set to "Append on Sync".

Yes.

Append on sync wil at a new point to your pointing model every time SGP syncs. This can be a pain when using plate solving at yo end up with clusters of closely grouped alignment points (where as alignment points really want to be space out).  You have two options:

1. Stay with "append on sync" but set the proximity range filter so that only the last sync point added within a proxiity range is actually used by EQMOD (any other sync points within the proximity range are automatically deleted from the model). If you have a permanmently sited mount doing this allows you to have platesolving build up/refine that, as you visit different targets, will improve the accuracy of any "non platesolved"  gotos you may subsequently make.

2. Change to "Dialogue Mode"  (orginally this was the mode used for manual, dialog led alignment). With Dialogue mode any sync received by EQMOD only acts to shift its current point model to align with that sync point - no new points are added. This is the more traditional implementation of a sync. Each time a new sync is made the correction shifts in RA and DEC (shown as DxSA and DxSB in EQMOD) will be recalculated based only on the current sync/posiiton offset.

Most folks just go the route od option 2 - probably beacuse this iswhat was settled upon by a consensus of opinion on the SGP support forum many years ago (but that's not to say there aren't folks sucessfully using option 1).

Chris

 

[souls33k3r]

10 minutes ago, chrisshillito said:

Yes.

Append on sync wil at a new point to your pointing model every time SGP syncs. This can be a pain when using plate solving at yo end up with clusters of closely grouped alignment points (where as alignment points really want to be space out).  You have two options:

1. Stay with "append on sync" but set the proximity range filter so that only the last sync point added within a proxiity range is actually used by EQMOD (any other sync points within the proximity range are automatically deleted from the model). If you have a permanmently sited mount doing this allows you to have platesolving build up/refine that, as you visit different targets, will improve the accuracy of any "non platesolved"  gotos you may subsequently make.

2. Change to "Dialogue Mode"  (orginally this was the mode used for manual, dialog led alignment). With Dialogue mode any sync received by EQMOD only acts to shift its current point model to align with that sync point - no new points are added. This is the more traditional implementation of a sync. Each time a new sync is made the correction shifts in RA and DEC (shown as DxSA and DxSB in EQMOD) will be recalculated based only on the current sync/posiiton offset.

Most folks just go the route od option 2 - probably beacuse this iswhat was settled upon by a consensus of opinion on the SGP support forum many years ago (but that's not to say there aren't folks sucessfully using option 1).

Chris

 

Ah, cheers for taking the time out explaining this Chris.

I do not have a permanent setup but i do roughly set up in the same location (a foot or two away). It's not to say that that plate solve doesn't work, it does but it takes a few tries before it does work.

I can try option 2 and see how this goes. Hopefully this will help me get over my issues that i've been having with plate solving so far.

[RayD]

3 hours ago, souls33k3r said:

This blind solve button, is this the one on "Scope Centering" docking module? If so then all what i remember seeing on there are "Solve & Sync" and "Blind Sync" options. Any thing to minimize the process.

That's the one mate, blind sync.  It just takes the image and sync's it with CdC or Stellarium.  Personally I like CdC as, whilst on the face of it it is a bit nuts and bolts, it just works exceptionally well as a star chart, which is all I need it to do for an imaging session.  Stellarium is nice, and I do sometimes use it to plan sessions as the imagery etc. is much better (it's a planetarium), but not for scope control.

[Dr_Ju_ju]

I fully agree, Stellarium is a wonderful thing to look at, and to plan a session, but when it comes to the "grunt work", I much prefer CDC....

[symmetal]

If you have trouble with Platesolve2 solving with SGP even when you know the scope is pointing in the right direction, try adjusting these Platesolve2 parameters:

Click on 'Settings' on the Plate Solve tab in Control Panel:

 

Click on Parameters... from the 'View' menu:

Change these parameters to the settings as shown:

I used to have trouble when solving in the Milky Way sometimes but it solves every time with these settings. With the default settings it seems it can omit too many bright stars from the image when solving, particularly when there are thousands of stars in the image. Changing the exposure time would probably have the same effect but with these settings it seems very tolerant of exposure values and solves anywhere with the same exposure.

Also thanks to Sara and Chris for the explanation about the 'Dialogue Mode' eqmod sync. I'd run into problems sometimes with the SGP  AutoCentering where it would refuse to go closer than 500 pixels or so. Deleting the 'sync' points in eqmod would cure it but setting 'Dialogue Mode' seems a better option. ?

Alan

[souls33k3r]

1 hour ago, RayD said:

That's the one mate, blind sync.  It just takes the image and sync's it with CdC or Stellarium.  Personally I like CdC as, whilst on the face of it it is a bit nuts and bolts, it just works exceptionally well as a star chart, which is all I need it to do for an imaging session.  Stellarium is nice, and I do sometimes use it to plan sessions as the imagery etc. is much better (it's a planetarium), but not for scope control.

I did have a look at CdC when i was first venturing in the world of planetarium software but the look of it just threw me off and went straight to Stellarium. I must admit, because my laptop is a very old thinkpad, i only use it for image acquisition. If i run Stellarium, the whole thing becomes super sluggish. Also what i noticed while using Stellarium is that if i have this app in full screen mode and i use Teamviewer, i won't be able to click on any thing even though if you look at the laptop, my mouse would be moving. By pressing F11, it comes out of the full screen window and then i'm able to use my mouse.

I guess i can take another look at CdC but will try to avoid it or Stellarium. I do have stellarium on desktop just to plan my sessions but that's just about it.

43 minutes ago, symmetal said:

If you have trouble with Platesolve2 solving with SGP even when you know the scope is pointing in the right direction, try adjusting these Platesolve2 parameters:

Click on 'Settings' on the Plate Solve tab in Control Panel:

 

Click on Parameters... from the 'View' menu:

Change these parameters to the settings as shown:

I used to have trouble when solving in the Milky Way sometimes but it solves every time with these settings. With the default settings it seems it can omit too many bright stars from the image when solving, particularly when there are thousands of stars in the image. Changing the exposure time would probably have the same effect but with these settings it seems very tolerant of exposure values and solves anywhere with the same exposure.

Also thanks to Sara and Chris for the explanation about the 'Dialogue Mode' eqmod sync. I'd run into problems sometimes with the SGP  AutoCentering where it would refuse to go closer than 500 pixels or so. Deleting the 'sync' points in eqmod would cure it but setting 'Dialogue Mode' seems a better option. ?

Alan

Cheers for the values Alan, i'll make sure i change them. 

My body is knackered but if it is surely going to be clear up until 1am then i might have another session tonight. More Ha data on the current project.

[Andyb90]

I had an issue with Platesolve2 a while back and reducing the exposure worked for me, so worth a try. I do 'Solve and Sync Blind' first for alignment using Astrometry.net (local version). Normally I only have to do that once and can then go to center on target in SGP and Platesolve2 successfully solves after 1 or 2 attempts.

Andy.

[souls33k3r]

2 minutes ago, Andyb90 said:

I had an issue with Platesolve2 a while back and reducing the exposure worked for me, so worth a try. I do 'Solve and Sync Blind' first for alignment using Astrometry.net (local version). Normally I only have to do that once and can then go to center on target in SGP and Platesolve2 successfully solves after 1 or 2 attempts.

Andy.

Cheers Andy. I was initially trying 5s but bumped the exposure to 10s to see if that made any difference which unfortunately it did. 

There's a plethora of useful advice in this thread for me (and for any further generations of going down the same AP rabbit hole). Proof will be in the pudding if i get to test this all tonight.

Do you think i should slew to the intended target first and then do "Blind Sync" or do the "Blind Sync" when my scope is in home position?