SGP - Platesolving nuisance

[swag72]

22 minutes ago, souls33k3r said:

......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?

I would do a blind sync from your target position......if you do it from home (assuming that's weights down and pointing north) I have found issues with that in the past.

[souls33k3r]

2 minutes ago, swag72 said:

I would do a blind sync from your target position......if you do it from home (assuming that's weights down and pointing north) I have found issues with that in the past.

Roger that mademoiselle 

[symmetal]

57 minutes ago, souls33k3r said:

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?

Do the blind sync after slewing to your target. Syncing at the pole is not recommended. RA 'errors' in particular may be reported as large when they are very tiny especially if you have a small FOV, as at the pole itself any RA value is 'correct'. ?

Alan

Ah!. Sara beat me to it. Didn't notice it had gone to page 2. 

[tooth_dr]

I always do a blind sync at the Pole home position, but the above point explains why this may not be accurate! Thanks for that ??

[han59]

I noted that setting the image size in arc minutes 1% or 2% higher or lower  then actual (can't remember which way, lower?) makes Platesolve2 more reliable. PlateSolve2 is very critical on image size. You could try that on a  saved image. It will still not help solving an image near the pole.

 

 

 

[Jkulin]

I always used to have a problem in SGP plate solving, so what I do first of all is slew within CdC to my target, then use Astrotortilla to platesolve which 99.9% of the time it does perfectly, this was with my AZ-EQ6-GT.

I then go back to SGP and tell it to centre on the object and all works fine.

However....first light with my new iOptron CEM60EC SGP platesolved perfectly straight off and put the image smack in the middle of the frame, couldn't have asked for better.

I haven't tried it since, but when I next image then will check it out.

As Ray said you need to have your parameters set correctly in your framing and mosaic dialogue box.

HTH

[souls33k3r]

2 hours ago, han59 said:

I noted that setting the image size in arc minutes 1% or 2% higher or lower  then actual (can't remember which way, lower?) makes Platesolve2 more reliable. PlateSolve2 is very critical on image size. You could try that on a  saved image. It will still not help solving an image near the pole.

 

 

 

I do apologise, where do i set this value?

 

3 minutes ago, Jkulin said:

I always used to have a problem in SGP plate solving, so what I do first of all is slew within CdC to my target, then use Astrotortilla to platesolve which 99.9% of the time it does perfectly, this was with my AZ-EQ6-GT.

I then go back to SGP and tell it to centre on the object and all works fine.

However....first light with my new iOptron CEM60EC SGP platesolved perfectly straight off and put the image smack in the middle of the frame, couldn't have asked for better.

I haven't tried it since, but when I next image then will check it out.

As Ray said you need to have your parameters set correctly in your framing and mosaic dialogue box.

HTH

My laptop is sluggish as hell so trying to avoid any such software (CdC or Stellarium). I guess i will have to do something to make this work.

[Jkulin]

Seriously. have a play with CdC, it is so simple to use.

When I first started using it I couldn't believe how easy it was, but underneath it are some serious tools that I now use and it is brilliant, PM me if you get stuck and I'll talk you though over the phone.

I looked at Stellarium and found it clunky, so didn't persevere.

HTH

 

[stash_old]

1 hour ago, Jkulin said:

iOptron CEM60EC

Bit of a diff in price so one  would hope it is better but I take note for future upgrades  ?

[souls33k3r]

1 hour ago, Jkulin said:

Seriously. have a play with CdC, it is so simple to use.

When I first started using it I couldn't believe how easy it was, but underneath it are some serious tools that I now use and it is brilliant, PM me if you get stuck and I'll talk you though over the phone.

I looked at Stellarium and found it clunky, so didn't persevere.

HTH

 

Top man, cheers mate. Will do

Let me have a cheeky read about the software while i'm work (As you do) lol

[han59]

5 hours ago, souls33k3r said:

>> I noted that setting the image size in arc minutes 1% or 2% higher or lower  then actual (can't remember which way, lower?) makes Platesolve2 more reliable. PlateSolve2 is very critical on image size. You could try that on a  saved image. It will still not help solving an image near the pole.

I do apologise, where do i set this value?

 

Platesolve2 has problem if the image size specified is a little too large. I don't use SGP except giving some assistance to club members, but while assisting in adding PlateSolve2 to the CCDciel program, I found this behavior. For this  a factor of 98% was added to the code.

As far I know there are in SGP two places to set is as indicated in the screenshots below.  The final PlateSolve2 windows should show a image size equal or a little smaller then actual size in arc minutes. If it is 1 or 2 arcminutes too large compared to actual it could fail.  Astrometry.net will tell you the actual image dimensions.

If it is all working and you're brave and have some IT skills you could try my ASTAP as PlateSolve2 substitute. It should be faster for larger offsets if setup correctly and has no problems with the image scale. But PlateSolve2 should work reliable and there is no need to replace it.

Han

 

[stash_old]

36 minutes ago, han59 said:

Platesolve2 has problem if the image size specified is a little too large. I don't use SGP except giving some assistance to club members, but while assisting in adding PlateSolve2 to the CCDciel program, I found this behavior. For this  a factor of 98% was added to the code.

As far I know there are in SGP two places to set is as indicated in the screenshots below.  The final PlateSolve2 windows should show a image size equal or a little smaller then actual size in arc minutes. If it is 1 or 2 arcminutes too large compared to actual it could fail.  Astrometry.net will tell you the actual image dimensions.

If it is all working and you're brave and have some IT skills you could try my ASTAP as PlateSolve2 substitute. It should be faster for larger offsets if setup correctly and has no problems with the image scale. But PlateSolve2 should work reliable and there is no need to replace it.

Han

 

There seems to be only a 64bit Windows version ?????? Is there a 32bit Win vers as per Linux ???

[stash_old]

1 hour ago, han59 said:

The final PlateSolve2 windows should show a image size equal or a little smaller then actual size in arc minutes. If it is 1 or 2 arcminutes too large compared to actual it could fail.  Astrometry.net will tell you the actual image dimensions.

Perfect - changed the settings,as per your instructions, in Platesolve2 as an experiment and guess what it worked - see proof attached.

Many thanks for that info

[han59]

1 hour ago, stash_old said:

There seems to be only a 64bit Windows version ?????? Is there a 32bit Win vers as per Linux ???

Just uploaded a win32 version. This was already overdue some time.  Most users are now on 64 bit.

I'm glad that PlateSolve2 is working. Now a field test to see if it stays that way.

Han

 

[stash_old]

1 minute ago, han59 said:

Just uploaded a win32 version. This was already overdue some time.  Most users are now on 64 bit.

I'm glad that PlateSolve2 is working. Now a field test to see if it stays that way.

Han

 

Many thanks I use both 32 and 64bit machines so I will give it a go. Yes a field test is needed perhaps tonight while I am sun bathing in 20degs heat. Again thanks for that info most interesting/useful

[han59]

1) Just an other remark. Your screen shots indicated  PlateSolve2 version 2.28.  This is the version you can download from PlaneWave. With SGP normally PlateSolve2 version 2.29 is supplied. The difference I noticed is that version 2.2.8 writes the solution  to a small file on disk using the decimal separator (dot or comma) as set in Windows.  In version 2.2.9 this is fixed with a dot.  Combined with a comma as separator, reading the PlateSolve 2.28 solution from disk becomes  a nightmare if the decimal separator is set as a comma in Windows. I assume version 2.2.9 was cooked for SGP to fix this.

Above is a problem for programmers. What I try to say is that for SGP the PlateSolve2 in C:\Users\you\AppData\Local\SequenceGenerator\  should be version 2.2.9  unless you have a old SGP I'm not aware off.

2) The PlateSolve2 image size found and set is for your image is 2.42 x1.46 degrees. If it works it works. However if you face problems it will also work with 2.4 x1.4 degrees.

Han

 

 

 

 

[stash_old]

6 minutes ago, han59 said:

1) Just an other remark. Your screen shots indicated  PlateSolve2 version 2.28.  This is the version you can download from PlaneWave. With SGP normally PlateSolve2 version 2.29 is supplied. The difference I noticed is that version 2.2.8 writes the solution  to a small file on disk using the decimal separator (dot or comma) as set in Windows.  In version 2.2.9 this is fixed with a dot.  Combined with a comma as separator, reading the PlateSolve 2.28 solution from disk becomes  a nightmare if the decimal separator is set as a comma in Windows. I assume version 2.2.9 was cooked for SGP to fix this.

Above is a problem for programmers. What I try to say is that for SGP the PlateSolve2 in C:\Users\you\AppData\Local\SequenceGenerator\  should be version 2.2.9  unless you have a old SGP I'm not aware off.

2) The PlateSolve2 image size found and set is for your image is 2.42 x1.46 degrees. If it works it works. However if you face problems it will also work with 2.4 x1.4 degrees.

Han

 

 

 

 

Yes I use APT so we use the "free" version 

I used 2.42 x 1.46 as this was the values returned by Astronomy upload and solve but its good to know you dont have to be spot on. ?

Just tried ASTAP but so far it says "No Solution found" but thats my fault probably entering the Ra/Dec. I shall continue with the experiment.

Thanks for the Info

[han59]

I tried to solve your M31 image, but it fails. ASTAP doesn't like stretched or stacked images. It works best with original images, 5, 10 or 600 seconds exposed.

If your happy with PlateSolve2 stay with it.

If your want to make is work, the approximate center position is required if used manually.  It will find an offset of 10 degrees in a few minutes.

Secondly the limiting magnitude should set once equal or higher of limiting magnitude of the image to solve. If you don't know it set it at maximum 17 but it will be slow. A too high value will slow down solving.

And finally the field of the camera should be roughly set within range.

If the input is JPG, set the binning to x2 or x4 for conversion to FITS.

Han

[stash_old]

24 minutes ago, han59 said:

I tried to solve your M31 image, but it fails. ASTAP doesn't like stretched or stacked images. It works best with original images, 5, 10 or 600 seconds exposed.

If your happy with PlateSolve2 stay with it.

If your want to make is work, the approximate center position is required if used manually.  It will find an offset of 10 degrees in a few minutes.

Secondly the limiting magnitude should set once equal or higher of limiting magnitude of the image to solve. If you don't know it set it at maximum 17 but it will be slow. A too high value will slow down solving.

And finally the field of the camera should be roughly set within range.

If the input is JPG, set the binning to x2 or x4 for conversion to FITS.

Han

Ok didn' see your reply so I ried it via APT by pointing to Platesolver2 (aka renamed Asatap.exe) - it called the program ok and seemed to pass the correct settings (approx) but APT just timed out afer 180's. I will have a play taking into account your settings as shown above. Note here is no Icon shown in he tray while ASTAP is running but Task manager shows its running the correct Platesolver2 from the Astap location. APT log showed below - note I am not using C2a but something called HNSKY ? as the planetarium 

2018/07/25 19:19:55 (UT 2018/07/25 18:19:55)    OpErr    Cannot detect camera
2018/07/25 19:19:56 (UT 2018/07/25 18:19:56)    Op    Dithering System initialized (free mode).
2018/07/25 19:19:56 (UT 2018/07/25 18:19:56)    Op    Connected to C2A (Computer Aided Astronomy).
2018/07/25 19:19:59 (UT 2018/07/25 18:19:59)    Op    Connected to EQMOD ASCOM Simulator
2018/07/25 19:20:34 (UT 2018/07/25 18:20:34)    DBG    Preview loading Canon SDK : 3
2018/07/25 19:21:03 (UT 2018/07/25 18:21:03)    DBG    PointCraft: Ra 19:59:36, Dec 22:43:15, FOVx 2.01226759, FOVy 1.34451973
2018/07/25 19:24:04 (UT 2018/07/25 18:24:04)    OpErr    PointCraft: Solving stopped with timeout!
2018/07/25 19:26:31 (UT 2018/07/25 18:26:31)    DBG    PointCraft: Ra 20:18:45, Dec 38:26:24, FOVx 2.01226759, FOVy 1.34451973

 

I was using an old CR2 (canon image) of m27 to see if it handled RAW images ok

 

[stash_old]

OK called ASTAP manually and set some of the parameters you suggested - see result

[stash_old]

ok don know how or maybe your settings change but APT worked - a lot slower than Platesolver2(real) but that might be that it annotated the new fits file -  I am truly impressed Hans now for a full field test .

Sorry Souls33k3sr for hijacking your thread with ASTAP stuff

[han59]

Okay we are making progress. 

Lets make it faster:

1) By limiting the maximum magnitude the speed will increase. Solve the image manually and read the image range in the log. E.g. 14-15  Then set the maximum magnitude at 15 or maybe one step higher ( is slower). Or just try empirical till is doesn't solve.

2) Un-check "convert to FITS" (with solution) See screen shot in my previous mail.

3) Un-check "produce deepsky annotated jpg"

4) Experiment with FITS binx2 or binx4.

If this is all working, ASTAP will loose some time by making a FITS from  the JPG received from APT but with increasing offset, ASTAP should find faster the solution. Try an offset of 5, 10 or even 20 degrees and see how it performs. (Set the maximum search range accordingly)

There is a little ASTAP icon at the screen bottom, right side when it is running.  The hint indicated how far it is in degrees from the start point.

Using jpg as an input and the PlateSolve2 command is minor disadvantage. Using FITS or the dedicated command line , ASTAP knows the image exposure time so can set the settings automatic correctly. Maybe one day Ivo (of APT) will add a dedicated command line specific for ASTAP and this will be solved. ?

I don't use APT or SGP so have only limited test experience with these. Feedback will be nice. My normal operation is with CCDCIEL and FITS files and it has never disappointed me. Again the JPG file input could be a disadvantage.
 

I'm also sorry hijacking this topic.

Han

[souls33k3r]

19 hours ago, stash_old said:

ok don know how or maybe your settings change but APT worked - a lot slower than Platesolver2(real) but that might be that it annotated the new fits file -  I am truly impressed Hans now for a full field test .

Sorry Souls33k3sr for hijacking your thread with ASTAP stuff

No worries mate, hijack away. If it help you then why not

[souls33k3r]

So last night was the first night i was able to go out and image. 

I followed all the advice from @swag72 @RayD and also changed the parameters in platesolve 2 @symmetal

Initially slewed to my intended target, did a Blind solve and then did the center on target and platesolve worked the first time. I now had hope that all will be good so couldn't wait until pier flip. Now the moment of truth, pier flip happened, slewed to the target but again the silly platesolving did not work even after 4-5 tried. I re-ran the platesolve again and this time it worked the first time.

[RayD]

1 minute ago, souls33k3r said:

So last night was the first night i was able to go out and image. 

I followed all the advice from @swag72 @RayD and also changed the parameters in platesolve 2 @symmetal

Initially slewed to my intended target, did a Blind solve and then did the center on target and platesolve worked the first time. I now had hope that all will be good so couldn't wait until pier flip. Now the moment of truth, pier flip happened, slewed to the target but again the silly platesolving did not work even after 4-5 tried. I re-ran the platesolve again and this time it worked the first time.

Are you using EQMOD and, if so, do you definitely have your alignment method set to dialogue mode?  Also make sure you are not hitting your limits as this will prevent you getting back to the right spot.  Finally I would run a check on cone error to make sure you don't have huge amounts.

Just make sure that everything is clear manually.  I struggle with my TAK on my AZ-EQ6 when using the EFW3 as the cables hit the tripod on both sides of the flip, so I can't leave it to run on its own.