RE: Problems with plate solving


Mar 2, 2014

 


----------------------------

#56122 Mar 2, 2014

I haven't changed anything except nowr I think I am using pinpoint 6. I still use GSC 1.1. But the plate solves take either forever or fail. The pointing error is no more than 10 to 20 arc mins off which routinely solved in a snap. What do I need to correct?



----------------------------

#56123 Mar 2, 2014

I've had the same experience, moving to Pinpoint 6



On Sun, Mar 2, 2014 at 5:05 AM, pgprendergast@...> wrote:

��I haven't changed anything except nowr I think I am using pinpoint 6. I still use GSC 1.1. But the plate solves take either forever or fail. The pointing error is no more than 10 to 20 arc mins off which routinely solved in a snap. What do I need to correct?



----------------------------

#56124 Mar 2, 2014

That requires an internet connection, correct? What we need is reliable fast plate solving when you are one or two FOV���s off from where you want to be. I keep hoping that someone will come up with a solution that can be used in the field where I don���t have reliable internet. For now when I do a long move from one side to the other I resynch along the way ��� not very elegant. Just one of the many non-joys of having to set up de novo every time.Rgrds-Ross ��From: MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of Tom MunneckeSent: Sunday, March 02, 2014 6:48 AMTo: maximdlSubject: Re: [MaxImDL] Problems with plate solving



----------------------------

#56125 Mar 2, 2014

On 3/2/2014 8:09 AM, Ross Salingerwrote:

Thatrequires an internet connection, correct?

Only if you want to use the all sky feature (if your way off orlost) as it uses Astrometry.net to find an initial solution thenuses that for a starting point for pinpoint to do its thing.



Tom



----------------------------

#56127 Mar 2, 2014

How does one turn or use on the all-sky feature of Pinpoint 6 from within Maxim DL?



On Sun, Mar 2, 2014 at 9:25 AM, Thouse look-up1@...> wrote:

��

On 3/2/2014 8:09 AM, Ross Salingerwrote:

Thatrequires an internet connection, correct?

Only if you want to use the all sky feature (if your way off orlost) as it uses Astrometry.net to find an initial solution thenuses that for a starting point for pinpoint to do its thing.



Tom



----------------------------

#56128 Mar 2, 2014

I believe you also need MaxIm 6 which is currently in beta test. ��Greg ��From: MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of Tom MunneckeSent: Sunday, March 02, 2014 1:34 PMTo: maximdlSubject: Re: [MaxImDL] Problems with plate solving



----------------------------

#56129 Mar 2, 2014

Gee I was really hoping that Maxim 6 would have all sky without astrometry.net being in the mix. Rgrds-Ross ��From: MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of Greg SalyerSent: Sunday, March 02, 2014 10:51 AMTo: MaxImDL@yahoogroups.comSubject: RE: [MaxImDL] Problems with plate solving



----------------------------

#56130 Mar 2, 2014

V5 does not use the net feature, you have to toggle it on or off inPP6 under the 'solve plates' tab and use pp6 to solve if you want touse the net.



Tom

On 3/2/2014 10:34 AM, Tom Munneckewrote:

��How does one turn or use on the all-skyfeature of Pinpoint 6 from within Maxim DL?



On Sun, Mar 2, 2014 at 9:25 AM,Thouse look-up1@...>wrote:

�� On 3/2/2014 8:09 AM, Ross Salingerwrote:

Thatrequires an internet connection,correct?

Only if you want to use the all sky feature(if your way off or lost) as it usesAstrometry.net to find an initial solutionthen uses that for a starting point forpinpoint to do its thing.



Tom



----------------------------

#56131 Mar 2, 2014

OK I'm not seeing an answer yet. ��I am by no means "lost", my aiming is well under one field of view. ��I can plainly see I am VERY close to center yet, plate solving takes forever. ��Any one have a solution????Peter



----------------------------

#56132 Mar 2, 2014

On 3/2/2014 6:50 PM, pgprendergast@... wrote: >

>

> OK I'm not seeing an answer yet. I am by no means "lost", my aiming

> is well under one field of view. I can plainly see I am VERY close to

> center yet, plate solving takes forever. Any one have a solution????





Which GSC are you using? The "corrected" v1.1, or the original? You

have to use the "corrected" one, IIRC.



How many stars is it detecting in the image? Perhaps too many or not

enough? You want anywhere from 15 to about 200, but no more. How about

how deep into the catalog? You don't need to use stars up to Mag 20

unless you are shooting VERY long and very deep. Most times, you can

put a cutoff at using catalog stars up to 18th mag, and speed up the

processing since there will be fewer stars against which to match.



Is the FOV set up correctly in MaxIM and PinPoint, so that the FITS

header has an ACCURATE starting point for pixel scale? If your image is

1.5 arc seconds/pixel, but the FITS header says 3.0 as/pp, it isn't

going to be able to solve. The scale in the header needs to be ideally

within about 10% of true, though you can get results if it is slightly

worse.



Jeff



----------------------------

#56134 Mar 3, 2014

One other thing to check, perhaps just too obvious, is whether the PinPoint 6 solve is starting with coordinates��near to what��you'd expect.���� I've had past issues with this (self-inflicted) with very similar symptoms to your own.- Rick McAlister



----------------------------

#56135 Mar 3, 2014

Peter is this only with MaxIm and notACP?

One other thing to check,perhaps just too obvious, is whether the PinPoint 6 solve isstarting with coordinates.near to what.you'd expect... I've hadpast issues with this (self-inflicted) with very similar symptomsto your own.







----------------------------

#56136 Mar 3, 2014

How is this even possible? When I slew to a location using Maxim, the FITS header contains the coordinates that Maxim thinks are in the FOV. You.d have to alter the input to Pinpoint in some way. Again, if it.s not in the FOV then PP tends not to solve. I for one don.t really understand why it can.t solve for something one degree (2 FOV.s in my case) off. I can always .synch hop. to avoid this. I keep wondering what Astronmetry.net has in its bag of tricks that lets it solve my image in 10 seconds or so that can.t be done on my laptop (comments welcome).Rgrds-Ross ��From: MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of Bob DennySent: Monday, March 03, 2014 7:00 AMTo: MaxImDL@yahoogroups.comSubject: Re: [MaxImDL] RE: Problems with plate solving



----------------------------

#56139 Mar 3, 2014

Ross,In order to match the stars in the image to the stars in thecatalog, PinPoint must read the stars in the star catalog for matching. Itcannot read the entire catalog, so it reads in the stars in an area set by aparameter in PinPoint. The RA and Dec in the FITS header are assumed to becorrect and PinPoint reads in stars near that position. It will cover an areaset by the "Search Area (as % of image)" parameter. You can set it up to 676%and PinPoint will try to solve the image. But the larger you set this parameter,the longer it will take to solve since you will have more catalog stars tocompare against.��If you are actually pointing outside the area covered by theSearch Area parameter, the solve will fail. There is an checkbox option for"Search outward spiral" that will allow PinPoint to search around the RA and Decin the FITS header in a spiral pattern. If you use this option, you do not needto set the Search Area parameter as large. I usually use 170%, but your mileagemay vary. Spiral searches are time consuming, but they do work.

��Astrometry.net uses an entirely different algorithm. Itdepends on a catalog of star pattern shapes that is already generated andmatches against them rather than searching the star catalog directly. Thiscatalog can be just as large or larger than the original star catalog, dependingon the number of different focal lengths you might try to solve. I use it allthe time, but do not know the details of the algorithm.��Hope this helps....Clear and dark skies!

-Bert

---------------BertStevensCo-DirectorDesert Moon Observatory, MPC #448Come visit us onthe web www.morning-twilight.com/dm448"Ihave loved the stars too fondly to be fearful of thenight."---------------

��



From: MaxImDL@yahoogroups.com[mailto:MaxImDL@yahoogroups.com] On Behalf Of RossSalingerSent: Monday, March 03, 2014 8:07 AMTo:MaxImDL@yahoogroups.comSubject: RE: [MaxImDL] RE: Problems with platesolving



Howis this even possible? When I slew to a location using Maxim, the FITS headercontains the coordinates that Maxim thinks are in the FOV. You���d have to alterthe input to Pinpoint in some way. Again, if it���s not in the FOV then PP tendsnot to solve. I for one don���t really understand why it can���t solve for somethingone degree (2 FOV���s in my case) off. I can always ���synch hop��� to avoid this. Ikeep wondering what Astronmetry.net has in its bag of tricks that lets it solvemy image in 10 seconds or so that can���t be done on my laptop (commentswelcome).Rgrds-Ross ��From:MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of BobDennySent: Monday, March 03, 2014 7:00 AMTo:MaxImDL@yahoogroups.comSubject: Re: [MaxImDL] RE: Problems with platesolving



----------------------------

#56140 Mar 3, 2014

Many of the details of Astrometry.net are elaborated in this paper:��adsabs.harvard.edu/abs/2010AJ....139.1782L... just FYI

~~~~~~~~~~~~~~~~~~~~~~~~~~~~Tim Stone



On Mon, Mar 3, 2014 at 1:16 PM, Bert Stevens blslcnm@...> wrote:

��Ross,In order to match the stars in the image to the stars in thecatalog, PinPoint must read the stars in the star catalog for matching. Itcannot read the entire catalog, so it reads in the stars in an area set by aparameter in PinPoint. The RA and Dec in the FITS header are assumed to becorrect and PinPoint reads in stars near that position. It will cover an areaset by the "Search Area (as % of image)" parameter. You can set it up to 676%and PinPoint will try to solve the image. But the larger you set this parameter,the longer it will take to solve since you will have more catalog stars tocompare against.��If you are actually pointing outside the area covered by theSearch Area parameter, the solve will fail. There is an checkbox option for"Search outward spiral" that will allow PinPoint to search around the RA and Decin the FITS header in a spiral pattern. If you use this option, you do not needto set the Search Area parameter as large. I usually use 170%, but your mileagemay vary. Spiral searches are time consuming, but they do work.

��Astrometry.net uses an entirely different algorithm. Itdepends on a catalog of star pattern shapes that is already generated andmatches against them rather than searching the star catalog directly. Thiscatalog can be just as large or larger than the original star catalog, dependingon the number of different focal lengths you might try to solve. I use it allthe time, but do not know the details of the algorithm.��Hope this helps....Clear and dark skies!

-Bert

---------------BertStevensCo-DirectorDesert Moon Observatory, MPC #448Come visit us onthe web www.morning-twilight.com/dm448"Ihave loved the stars too fondly to be fearful of thenight."---------------

��



From: MaxImDL@yahoogroups.com[mailto:MaxImDL@yahoogroups.com] On Behalf Of RossSalingerSent: Monday, March 03, 2014 8:07 AMTo:MaxImDL@yahoogroups.comSubject: RE: [MaxImDL] RE: Problems with platesolving



Howis this even possible? When I slew to a location using Maxim, the FITS headercontains the coordinates that Maxim thinks are in the FOV. You���d have to alterthe input to Pinpoint in some way. Again, if it���s not in the FOV then PP tendsnot to solve. I for one don���t really understand why it can���t solve for somethingone degree (2 FOV���s in my case) off. I can always ���synch hop��� to avoid this. Ikeep wondering what Astronmetry.net has in its bag of tricks that lets it solvemy image in 10 seconds or so that can���t be done on my laptop (commentswelcome).Rgrds-Ross��From:MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of BobDennySent: Monday, March 03, 2014 7:00 AMTo:MaxImDL@yahoogroups.comSubject: Re: [MaxImDL] RE: Problems with platesolving







----------------------------

#56141 Mar 3, 2014

I may try to fiddle with the parameters a bit and see what happens next time I get out to take pictures. My experience is that if my object is not in the FOV, then PP cannot locate where I am. This makes it pretty much useless other than as a way to center things. Of course once I do that, then I can move across the sky pretty well but it���s tedious. I find myself having to do a visual observation to start almost every evening when I���m trying to image at 1625mm and it just wastes time. If something could solve 2 or 3 FOV���s off that would make my life (but I guess not others) much easier.Rgrds-Ross ��From: MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of Bert StevensSent: Monday, March 03, 2014 11:17 AMTo: MaxImDL@yahoogroups.comSubject: RE: [MaxImDL] RE: Problems with plate solving



----------------------------

#56142 Mar 3, 2014

I found that occasionally Pinpoint would revert back to an old telescope focal length, or the FITS pixel info would not be in the image.�� I had to go back and reset the scope length again, for some reason.�� Never could figure out why the FITS data didn't always come over.



Can anyone explain why Pinpoint 6 is more brittle than 5 in terms of having the image nearly centered?�� Seems like this is a big step backwards in terms of software evolution.



On Mon, Mar 3, 2014 at 3:33 PM, Ross Salinger rgsalinger@...> wrote:

��I may try to fiddle with the parameters a bit and see what happens next time I get out to take pictures. My experience is that if my object is not in the FOV, then PP cannot locate where I am. This makes it pretty much useless other than as a way to center things. Of course once I do that, then I can move across the sky pretty well but it���s tedious. I find myself having to do a visual observation to start almost every evening when I���m trying to image at 1625mm and it just wastes time. If something could solve 2 or 3 FOV���s off that would make my life (but I guess not others) much easier.Rgrds-Ross��From: MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of Bert Stevens

Sent: Monday, March 03, 2014 11:17 AM

To: MaxImDL@yahoogroups.comSubject: RE: [MaxImDL] RE: Problems with plate solving



----------------------------

#56143 Mar 3, 2014

Perhaps you already stated this, but could you explain the environment

you are using PP6 in? Is it standalone? Are you using it in Maxim 5, or

in a beta of Maxim 6? Are you using it with an active Internet connection?



Chris



*******************************

Chris L Peterson

Cloudbait Observatory

www.cloudbait.com

On 3/3/2014 5:57 PM, Tom Munnecke wrote:

> I found that occasionally Pinpoint would revert back to an old telescope

> focal length, or the FITS pixel info would not be in the image. I had to

> go back and reset the scope length again, for some reason. Never could

> figure out why the FITS data didn't always come over.

>

> Can anyone explain why Pinpoint 6 is more brittle than 5 in terms of having

> the image nearly centered? Seems like this is a big step backwards in

> terms of software evolution.

>



----------------------------

#56145 Mar 3, 2014

Are you sure that PP6 is able to access astrometry.net when it's

embedded in Maxim 5? I'd think that feature might be inaccessible. Since

PP has always been poor at solving images where the coordinates weren't

already known roughly, it wouldn't surprise me if that performance

changed a little in the new version, possibly for the worse, given the

new ability to use astrometry.net. Perhaps Bob Denny can comment.



Chris



*******************************

Chris L Peterson

Cloudbait Observatory

www.cloudbait.com

On 3/3/2014 6:07 PM, Tom Munnecke wrote:

> I'm using latest version of Maxim 5, with latest version of PP6 (production

> versions only, no beta)

>

>

> On Mon, Mar 3, 2014 at 5:05 PM, Chris Peterson lists@...> wrote:

>

>>

>>

>> Perhaps you already stated this, but could you explain the environment

>> you are using PP6 in? Is it standalone? Are you using it in Maxim 5, or

>> in a beta of Maxim 6? Are you using it with an active Internet connection?

>>

>> Chris



----------------------------

#56151 Mar 4, 2014

I've had very limited experience with ACP. But with MaximDL it rarely works now



Sent from my iPad



----------------------------

#56153 Mar 4, 2014

I found the same thing for the past year or two. I switched to Sky X and the plate solve takes something on the order of 20 seconds. I've used both pinpoint LE and the full version. Never worked anywhere as good as Sky X plate solving.��Tom P



----------------------------

#56155 Mar 4, 2014

Yes, I've heard that the all-sky solving doesn't connect to Maxim 5... I have to wait for the new 6, I guess.�� And yes, I'm connected to the internet.�� I can get all sky plate solves via Visual Pinpoint.



Sometimes, I can get an image to solve in Maxim, but not in Visual Pinpoint... both use the same engine.�� I'm presuming that there is some settings glitch between the two.



On Mon, Mar 3, 2014 at 5:12 PM, Chris Peterson lists@...> wrote:

��Are you sure that PP6 is able to access astrometry.net when it's

embedded in Maxim 5? I'd think that feature might be inaccessible. Since

PP has always been poor at solving images where the coordinates weren't

already known roughly, it wouldn't surprise me if that performance

changed a little in the new version, possibly for the worse, given the

new ability to use astrometry.net. Perhaps Bob Denny can comment.



Chris



*******************************

Chris L Peterson

Cloudbait Observatory

www.cloudbait.com



On 3/3/2014 6:07 PM, Tom Munnecke wrote:

> I'm using latest version of Maxim 5, with latest version of PP6 (production

> versions only, no beta)

>

>

> On Mon, Mar 3, 2014 at 5:05 PM, Chris Peterson lists@...> wrote:

>

>>

>>

>> Perhaps you already stated this, but could you explain the environment

>> you are using PP6 in? Is it standalone? Are you using it in Maxim 5, or

>> in a beta of Maxim 6? Are you using it with an active Internet connection?

>>

>> Chris







----------------------------

#56156 Mar 4, 2014

Visual Pinpoint seems to require a very accurate pixel size value that must be manually set. When I've had a problem correctly setting this value has worked for me. ��Greg ��From: MaxImDL@yahoogroups.com [mailto:MaxImDL@yahoogroups.com] On Behalf Of Tom MunneckeSent: Tuesday, March 04, 2014 11:04 AMTo: maximdlSubject: Re: [MaxImDL] RE: Problems with plate solving



----------------------------

#56157 Mar 4, 2014

[I am the author of PinPoint]



Tom -



This is sort of a blanket statement... "[PinPoint] never workedanywhere as good as TheSky X". I would appreciate your furnishingsome engineering data (images, solve reliability and times) withwhich I could compare so I can see your experience and see whatmight be going on. This being the MaxIm DL list, we all should befocused on delivering the best experience to MaxIm DL users, noteffectively telling them they should go out and buy some othersoftware package based on a. rather thin blanket statement aboutplate solving. I am serious about my interest in engineering datawith which I can substantiate your experience and then see aboutmaking the changes needed to bring PinPoint up to yourexpectations.



. -- Bob



I found the same thing forthe past year or two. I switched to Sky X and the plate solvetakes something on the order of 20 seconds. I've used bothpinpoint LE and the full version. Never worked anywhere as good asSky X plate solving..Tom P



----------------------------

#56158 Mar 4, 2014

You said:

How does one turn or use on the all-sky feature of Pinpoint 6 fromwithin Maxim DL?



It is built into MaxIm 6. The feature was added to PinPoint afterMaxIm 5's design, therefore MaxIm 5's operator controls do notinclude it (as is with the newly introduced UCAC4 catalog feature).



�� -- Bob



----------------------------

#56159 Mar 4, 2014

This was my decision, and I knew upfront that it would not serve those users out in the field with nointernet service. If you are interested in the details of myengineering decision please feel free to call me at 480 396 9700and chat. It's to much (and probably boring to most) to write thewhole thing here. Suffice it to say that all-sky plate solving isdifficult, requiring lots of computing power and many gigs of diskspace.



There are two free local packages you can load up if you want:



Elbrus: www.astrosurf.com/pulgar/elbrus/elbrusin.htm

Astrometry.net: astrometry.net/use.html



Both do all-sky plate solving with a local database.



�� -- Bob



GeeI was really hoping that Maxim 6 would have all sky withoutastrometry.net being in the mix. Rgrds-Ross



----------------------------

#56160 Mar 4, 2014

Chris is right. PinPoint is an engine,and its features are exposed through whatever program hosts it, inthie case MaxIm DL. You need MaxIm DL V6 (in beta) to use all-skyand also the UCAC4 catalog.



�� -- Bob



PS: I will now post a separate article about UCAC4.



Are you sure that PP6 is able to access astrometry.net when it's embedded in Maxim 5? I'd think that feature might be inaccessible. Since PP has always been poor at solving images where the coordinates weren't already known roughly, it wouldn't surprise me if that performance changed a little in the new version, possibly for the worse, given the new ability to use astrometry.net. Perhaps Bob Denny can comment.

Chris

*******************************Chris L PetersonCloudbait Observatorywww.cloudbait.com

On 3/3/2014 6:07 PM, Tom Munnecke wrote:

I'm using latest version of Maxim 5, with latest version of PP6 (productionversions only, no beta)

On Mon, Mar 3, 2014 at 5:05 PM, Chris Peterson lists@...> wrote:



Perhaps you already stated this, but could you explain the environmentyou are using PP6 in? Is it standalone? Are you using it in Maxim 5, orin a beta of Maxim 6? Are you using it with an active Internet connection?

Chris





---------------

Yahoo Groups Links

*> To visit your group on the web, go to: groups.yahoo.com/group/MaxImDL/

*> Your email settings: Individual Email | Traditional

*> To change settings online go to: groups.yahoo.com/group/MaxImDL/join (Yahoo! ID required)

*> To change settings via email: MaxImDL-digest@yahoogroups.com MaxImDL-fullfeatured@yahoogroups.com

*> To unsubscribe from this group, send an email to: MaxImDL-unsubscribe@yahoogroups.com

*> Your use of Yahoo Groups is subject to: info.yahoo.com/legal/us/yahoo/utos/terms/



----------------------------

#56161 Mar 4, 2014

��VisualPinpoint seems to require a very accurate pixel size value thatmust be manually set. When I've had a problem correctly settingthis value has worked for me.



Actually the parameter is the plate scale, in arcseconds per pixel,for Visual PinPoint. This is determined from both the focal lengthand the pixel size. In MaxIm DL, you have the ability to specify thefocal length and get the pixel size from the camera, makingcalculation of the plate scale (for input to PinPoint) easier.

Sometimes, I can get an image to solvein Maxim, but not in Visual Pinpoint... both use the sameengine.�� I'm presuming that there is some settings glitchbetween the two.



Yes, each program is separately responsible for hosting the PinPointengine and for providing the parameters and settings. They both usethe exact same engine,, so any differences in behavior areabsolutely due to settings differences.



�� -- Bob



----------------------------

#56164 Mar 4, 2014

Re: " Never worked anywhere asgood as Sky X plate solving."��This assessment is perhaps a bit harsh. ��My experience has been that, once I got the configuration sorted out, PinPoint 6 worked quite well, as did LE before it. ��I am able to complete unattended MaxPoint auto-calibration runs of 50 or more stars quickly and flawlessly. ��Most plate-solves take on the order of a second or less, even in abysmal seeing conditions. ��I do recall that as I upgraded to PinPoint 6, the configuration parameters did not follow from LE (It also could have been an "operator error", don't know.). ��Once I figured that out, everything worked as before. ��So don't lose hope, just get the configuration right. ��

- Rick McAlister







Contact Us
This Site's Privacy Policy
Google's privacy policies

S
e
n
i
o
r
T
u
b
e
.
o
r
g