92 replies to this topic

[ButterFly]

Here is v1.1 of Plate Align to add alignment points to CPWI from platesolving instead of centering stars:

 

Connect CPWI to your mount however you ordinarily do so.  Everything is the same up until you need to center stars.  Make sure you know how to use CPWI properly.  Use CPWI with an eyepiece a few times if you need to make sure.

 

Open Plate ALign, and connect CPWI and a camera.  Check to make sure your latitude, longitude, and system time are correct or the mount will move to incorrect places.

 

Find focus, exposure, and gain settings that permit a plate solve.  It is better to use your telescope rather than a guide scope.  Try the settings far from the Milky Way to make sure it can solve everywhere. 

 

Double click on the chart to add an alignment point at that alt/az location.  Add as many as you wish, but remember that it goes in order.  Try to minimize the slewing distance between points.  When ready, run the alignment.

 

Revisions Plate Aign v1.1:

Added Focuser Support
Added Focal Length update after first Solution for faster solving
Added units to slew speed
Added checking ASTAP's default location

Fixed window focus changing when validating text boxes
Fixed to prohibit SIMBAD lookup and goto during alignment
Fixed to check whether Epoch is one of J2000.0 or JNow in Settings file
Fixed to slew to Az Alt based on selected Epoch
Fixed to only write settings for connected components
Fixed to write FITS Data asynchronously in blocks of 28800 bytes
Fixed to use Haversine formula with small angles

Changed constructors for ASCOM classes to permit reading properties before connecting
Changed icons of child windows
Changed ASCOM panel layouts
Changed time precision to include milliseconds

 

 

Attached Files

•  PlateAlign_v1.1_Installer.zip   374.93KB   281 downloads

[ButterFly]

Here is the source:

 

 PlateAlign_v1.1_Source.zip   170KB   89 downloads

 

I'm hoping to add Polar alignment, but that will take quite a bit more testing.  That will be v2.0.  Let me know if there's something else you would find useful.

[MarMax]

I've just read through your two threads on PlateAlign and wondered if I would benefit from using it. I have a CPC 1100 that I'm very familiar with along with CPWI and I'm also running SharpCap and Stellarium and sometimes use PHD2 if guiding. My normal startup with CPWI is a two-star alignment, after which I run things with Stellarium and SharpCap, using plate solving within SharpCap.

 

Does PlateAlign replace the two-star alignment and will it take less time for alignment?

[ButterFly]

Choose quick align in CPWI so there is tracking.  You can add as many points as you want, and I don't know how long it takes you to center stars now, so I can't say whether it will take you longer.  You can do other things though, while it's capturing and solving.

 

I'm fine with 5s captures anywhere in the sky at 900mm focal length and 2.4 micron pixels.  I get about 3-4 points per minute, with slewing taking the most time.  Choose a path for your set of points that avoids unnecessary slewing back and forth to minimize that.

[MarMax]

Thanks for the response. I'd say a two-star align from start to finish is about 5 minutes, maybe a bit less. I've never used quick align and don't really know what tracking does. My apology for the dumb questions but does Plate Align connect to the camera via an ASCOM driver. I've only used SharpCap for running cameras (and PHD2 for guiding) so I'm wondering if Plate Align integrates with the other apps.

 

My current workflow is to start CPWI and initiate a two-star alignment. Then start SharpCap and center the first star and go back to CPWI and repeat for the second star. At this point I don't use CPWI any more. CPWI is always running in the background but I use Stellarium with SharpCap once alignment is complete.

 

Is the difference with Plate Align that I don't use SharpCap initially? I just start CPWI then start Plate Align and select quick align?

[ButterFly]

You would use this before Sharpcap. It connects to both CPWI and your camera, both through ASCOM. After you add your alignment points, you're done with this program and you close it. CPWI then has an alignment, gets minimized, and forgotten as usual. If you need Sharpcap first to polar align or focus, disconnect the camera when you're done in Sharpcap to avoid two things telling the same camera to do different things.

Five minutes is enough time for a lot of alignment points! Add at least ten and see how CPWI does. You can watch the progress in CPWI alignment model. It updates as you add new points. If you don't improve much between points six and ten, maybe you don't need ten points. How many and which model terms CPWI adds depends on your mount type, and the number of points you add. If you add more points, it can add more model terms, which can mean better tracking. Play around with it and see how many points works for your goals. Once you find a set of points that works well enough, you just run it without doing much else.

I get good RA and Dec tracking with a 25 point model with my EQ mount in under five minutes. The difference between the model and the plate solution is in the arcsecond level by then. It's five second exposures after a five second settling time. Solving is under a second. Then slewing on to the next.

I have a set of points that work well, and I don't snag cables of anything like that with them. I collimate, focus, align my finder, then run the points. That's when I head inside, take off my coat, ..., and set up my inside stuff while it's aligning. When it's done, I start guiding, touch up focus in Sharpcap, and get to it. When doing satellites, I'm usually outside with an eyepiece.

[MarMax]

Thank you ButterFly for the excellent explanation. Clearly you created Plate Align because you want the best possible tracking.

 

If it ever clears up I'll try it with the CPC 1100. I have a CGEM II that I picked up used recently and other than testing it with CPWI I've not used it under the stars. I'm familiar with the SharpCap polar align tool since I use it with the G11G but have not done a polar align with the CGEM II. I'm going to continue with some more questions for clarification and then make a set of workflow notes.

 

Workflow CPC 1100

- start Plate Align

- start CPWI (does start order matter?) > select quick align

- use Plate Align to add alignment points

- quit Plate Align, start SharpCap

 

Is the "alignment model" you refer to this dialog box with Select Target and Pointing Model Information?

 

 

Workflow CGCM II

- start Plate Align

- start CPWI > select quick align

- use Plate Align to add alignment points

- quit Plate Align, start SharpCap and run Polar Align tool

 

I've never run CPWI with an EQ mount so have not used ASPA. From the CPWI manual it sounds like SharpCap is a quicker and easier tool to use. Do you use ASPA or SharpCap for polar align?

 

 

[ButterFly]

CPWI needs to be connected to the mount for anything else to connect to it through ASCOM.  You can start them in any order, but you have to connect CPWI to the mount before connected to CPWI through ASCOM.  If you don't connect to the mount first, there is nothing to connect to through ASCOM!

 

I polar align using Sharpcap first because I can.  It is a lot quicker and easier than centering a star.  You do need to see the celestial poles though, and you need to use a wide field camera, such as on your guide scope.  Sharpcap can't polar align with small true fields (so no main scope) and you need to be pointed at the pole (so no random pointing).  If you can't do either of those things, Sharpcap's polar align will fail.  You would have to use ASPA at that point.  I put crosshairs in so one could, but Sharpcap is quicker and easier if you can use it.

 

If you send alignment points to CPWI before polar alignment, you should redo the alignment points after polar aligning.  The more you change the polar alignment, the more off the star alignment can get.  That will affect RA+Dec tracking, which needs an actual measurement of the polar error. 

 

If you have to use ASPA, do a first run of six stars that are well spaced, with the last point around meridian and the celestial equator.  CPWI will then have enough to know your polar error well enough, and you'll be close to the optimal position to center a star.  If you have never used ASPA before, center stars using the handcontroller or CPWI a few times until you are familiar with the process.  Doesn't need to be perfect centering in the eyepiece (or sensor), but enough for you understand how it's done.  If your polar alignment is far off, the star be well off the chip if using a small sensor, which you then have to center using the EQ mount's alt az knobs.  Having a separate guide scope helps there to bring it back to center-ish.  An OAG can be hopelessly lost.

[MarMax]

CPWI needs to be connected to the mount for anything else to connect to it through ASCOM.  You can start them in any order, but you have to connect CPWI to the mount before connected to CPWI through ASCOM.  If you don't connect to the mount first, there is nothing to connect to through ASCOM!

 

OK, got it and I do connect to the mount first, it's assumed when I say "start CPWI".

 

I polar align using Sharpcap first because I can.  It is a lot quicker and easier than centering a star.  You do need to see the celestial poles though, and you need to use a wide field camera, such as on your guide scope.  Sharpcap can't polar align with small true fields (so no main scope) and you need to be pointed at the pole (so no random pointing).  If you can't do either of those things, Sharpcap's polar align will fail.  You would have to use ASPA at that point.  I put crosshairs in so one could, but Sharpcap is quicker and easier if you can use it.

 

If you send alignment points to CPWI before polar alignment, you should redo the alignment points after polar aligning.  The more you change the polar alignment, the more off the star alignment can get.  That will affect RA+Dec tracking, which needs an actual measurement of the polar error. 

 

OK, got it, polar align first with SC then start CPWI and use Plate Align.

 

If you have to use ASPA, do a first run of six stars that are well spaced, with the last point around meridian and the celestial equator.  CPWI will then have enough to know your polar error well enough, and you'll be close to the optimal position to center a star.  If you have never used ASPA before, center stars using the handcontroller or CPWI a few times until you are familiar with the process.  Doesn't need to be perfect centering in the eyepiece (or sensor), but enough for you understand how it's done.  If your polar alignment is far off, the star be well off the chip if using a small sensor, which you then have to center using the EQ mount's alt az knobs.  Having a separate guide scope helps there to bring it back to center-ish.  An OAG can be hopelessly lost.

 

I'll not use ASPA. I really like the SC polar align tool.

Thank you for clearing this up. I'm only familiar with using the G11G with SC polar align. With the G11G after polar alignment I just slew to the first object and plate solve. I never build an alignment model. It's always worked well. Maybe GeminiTelescope.net works a bit differently than CPWI. I'm just an end user and have no clue what's going on behind the scenes.

[ButterFly]

Some bug fixes:

 

Fixed Focal Length update when plate scale is negative (diagonal mirrors)
Fixed sending alignment point when slew aborted

 

 PlateAlignSetup_v1.1.1_Installer.zip   363.04KB   280 downloads

[ButterFly]

Source:

 ASCOMPlateAlign_Source.zip   169.89KB   110 downloads

[A. Viegas]

This looks very interesting.   Arguably, the biggest PITA problem for building an alignment model with CPWI is the need to use your keyboard to slew the scope and then do fine centering.   Usually, because my mount is left in the same spot, I just use a last alignment and center on one star and then plate solve going forward with SGP or Nina.    I am looking forward to using your program to build a larger model.  Thanks for doing this

[spz56]

Thanks! Just need a few clear nights. I’ll get it installed and give it a try when I can. 

[spz56]

Here's a quick status update on my progress with PlateAlign. I've only had a couple hours at the most over the last few nights and I was dodging high clouds the entire time. The first night out, I couldn't get the plate solver to work at first. It connected to CPWI just fine after a quick align and I was able to slew around with the app.

 

I did a quick manual align and an All Star Polar Align on a AVX mount to work on guiding (DEC is pitiful with this mount). I've only used this for daytime solar stuff prior to the last week or two so I never bothered with CPWI or guiding before. It does not work great with CPWI - the mount can slew off in any random direction, could not complete a StarSense autoalign because of this but I was able get a manual StarSense model if I slewed the AXV from east to west. By the way, I completely tore down the mount last week, cleaned, polished, and regreased it with Super Lube. This did nothing to fix the problems with CPWI or PHD2 guiding.

 

I came back to PlateAlign after the clouds starting moving in and tried to plate solve single points; plate solver failed (input FL was correct at 346mm). Tried to solve FIT files; still didn't work. Went to ASTAP stand alone and FIT file solved fine when I adjusted the image heigth to >2 degrees. I went back to PlateAlign and tried solving individual points and it working great. I tried few 3 point alignments and the app worked great. Gave up at this point because there were too many clouds to dodge.

 

Last night I tried again with a few 8 point alignments and the app worked great. Here's a couple of screen images. A few points failed (not sure why, clouds?) but the model was OK and I went back to fiddling the AVX guiding (DEC was no better).

 

Thanks for the great app!

 

Steve

 

 

Attached Thumbnails

• 
• 

[ButterFly]

Play with the exposure and gain before the ASTAP settings.  180 degree search radius is all sky.  With a "Quick Align", it's close enough that the default 15 degree radius is fine, as long as your time and location are correct, and you did a reasonable job lining up the index marks and polar aligning (under ten degrees or so).  My first "star" is off by a few degrees at most.

 

It shows an HFD of around 5, so you can turn up the Star radius from the default of 1.5.  That will help keep hot pixels from interfering - then turn up the gain to get more stars.

 

If you want to play with the ASTAP settings, you can drag and drop the file you are having trouble solving.  Change the ASTAP settings (star radius first), then drag and drop.

 

Yes on the Dec axis of the AVX.  After about 20 points or so, I notice that most of the error is in Dec.

 

I came back to PlateAlign after the clouds starting moving in and tried to plate solve single points; plate solver failed (input FL was correct at 346mm). Tried to solve FIT files; still didn't work. Went to ASTAP stand alone and FIT file solved fine when I adjusted the image heigth to >2 degrees.

This could mean that your focal length is off?  ASTAP prefers +/- 5%.  Look at the fits headers to make sure your camera is reporting the correct pixel size and chip sizes.  The latest versions (v1.1+) adjusts that focal length after the first solve to make the other ones quicker.  If you're having trouble, send me a PM with your header's pixel size and chip size, as well as the ini file from ASTAP.
 

[spz56]

Thanks for the tips! The focal length is correct: 346mm. Also, the camera settings look good. Here is a a fits header file:

 

 

Image #1
    SIMPLE  =                    T / file does conform to FITS standard
    BITPIX  =                   16 / number of bits per data pixel
    NAXIS   =                    2 / number of data axes
    NAXIS1  =                 3704 / length of data axis 1
    NAXIS2  =                 2778 / length of data axis 2
    EXTEND  =                    T / FITS dataset may contain extensions
    COMMENT   FITS (Flexible Image Transport System) format is defined in 'Astronomy
    COMMENT   and Astrophysics', volume 376, page 359; bibcode: 2001A&A...376..359H
    BZERO   =                32768 / offset data range to that of unsigned short
    BSCALE  =                    1 / default scaling factor
    DATE-OBS= '2023-04-01T22:13:06' / date of the observation
    DATE    = '2023-04-01T22:13:06' / date of file creation
    EXPTIME =                  60. / Exposure time in seconds
    EXPOSURE=                  60. / Exposure time in seconds
    GAIN    =                 1000 / Analog Gain
    XPIXSZ  =                 4.63 / Pixel Width in microns
    YPIXSZ  =                 4.63 / Pixel Height in microns
    XBINNING=                    1 / Binning level along the X-axis
    YBINNING=                    1 / Binning level along the Y-axis
    INSTRUME= 'SkyRaiderDS10CTEC'  / Camera Model
    OBJECT  = 'SH2-235 '           / name of observed object
    CCD-TEMP=                   0. / actual measured sensor temperature in degrees C
    SET-TEMP=                   0. / sensor temperature setpoint in degrees C
    IMAGETYP= 'Light Frame'        / type of image
    CREATOR = 'MALLINCAMSKY V4.11.21645.20221103' / the name of the software task th
    PROGRAM = 'MALLINCAMSKY V4.11.21645.20221103' / the name of the software task th
    SWCREATE= 'MALLINCAMSKY V4.11.21645.20221103' / string indicating the software u
    COLORTYP= 'GBRG    '           / Bayer mosaic
    BAYERPAT= 'GBRG    '           / Bayer pattern
    BAYOFFX =                    0 / Bayer offset X
    BAYOFFY =                    1 / Bayer offset Y

[ButterFly]

This header didn't come out of anything I wrote.

 

Check the header Plate Align is writing.  That's what is being sent to solve, and that is what ASTAP is reading.  Confirm that the pixel sizes, and the naxis 1 and 2 are the correct.  The output pixel scale in degrees is in the solved ASTAP files ending in .wcs or .ini, in the same directory as the written file, under CDELT1 and CDELT2.

[spz56]

Oh - I see. This is the header from the camera file only. I could not find any saved fits from PlateAlign. I've been playing with ASTAP, and it will fail every time until I enter the correct image height (alignment tab, stack menu) or use "AUTO" selection. Once I select the AUTO height, ASTAP will blind solve anything I throw at it (native long SCT FLs to wide angle fast set-ups). Once a solution is found, ASTAP will default to it and solve it consistently for the same FOV. The pixel math 2 tab solves for FL using pixel size data from the fit header. I'll use a settings file for each of my set-ups and be sure to load it in before using ASTAP. Here is a solved file header. Does this look correct to you?

 

Thanks again! I very much apprecate the effort you put into developing the app.

 

Steve

 

 

SIMPLE  =                    T / file does conform to FITS standard            
BITPIX  =                   16 / number of bits per data pixel                 
NAXIS   =                    0 / number of data axes                           
EXTEND  =                    T / FITS dataset may contain extensions           
COMMENT   FITS (Flexible Image Transport System) format is defined in 'Astronomy
COMMENT   and Astrophysics', volume 376, page 359; bibcode: 2001A&A...376..359H
BZERO   =                32768 / offset data range to that of unsigned short   
BSCALE  =                    1 / default scaling factor                        
DATE-OBS= '2023-04-01T22:30:20' / date of the observation                      
DATE    = '2023-04-01T22:30:20' / date of file creation                        
EXPTIME =                  60. / Exposure time in seconds                      
EXPOSURE=                  60. / Exposure time in seconds                      
GAIN    =                 1000 / Analog Gain                                   
XPIXSZ  =                 4.63 / Pixel Width in microns                        
YPIXSZ  =                 4.63 / Pixel Height in microns                       
XBINNING=                    1 / Binning level along the X-axis                
YBINNING=                    1 / Binning level along the Y-axis                
INSTRUME= 'SkyRaiderDS10CTEC'  / Camera Model                                  
OBJECT  = 'SH2-235 '           / name of observed object                       
CCD-TEMP=                   0. / actual measured sensor temperature in degrees C
SET-TEMP=                   0. / sensor temperature setpoint in degrees C      
IMAGETYP= 'Light Frame'        / type of image                                 
CREATOR = 'MALLINCAMSKY V4.11.21645.20221103' / the name of the software task th
PROGRAM = 'MALLINCAMSKY V4.11.21645.20221103' / the name of the software task th
SWCREATE= 'MALLINCAMSKY V4.11.21645.20221103' / string indicating the software u
COLORTYP= 'GBRG    '           / Bayer mosaic                                  
BAYERPAT= 'GBRG    '           / Bayer pattern                                 
BAYOFFX =                    0 / Bayer offset X                                
BAYOFFY =                    1 / Bayer offset Y                                
HFD     =  5.821939648747E+000 / Median Half Flux Diameter
STARS   =  3.200000000000E+001 / Number of stars detected
CTYPE1  = 'RA---TAN'           / first parameter RA  ,  projection TANgential  
CTYPE2  = 'DEC--TAN'           / second parameter DEC,  projection TANgential  
CUNIT1  = 'deg     '           / Unit of coordinates                           
EQUINOX =               2000.0 / Equinox of coordinates                        
CRPIX1  =  1.852500000000E+003 / X of reference pixel                          
CRPIX2  =  1.389500000000E+003 / Y of reference pixel                          
CRVAL1  =  8.528855813873E+001 / RA of reference pixel (deg)                   
CRVAL2  =  3.572501320678E+001 / DEC of reference pixel (deg)                  
CDELT1  = -7.678342146638E-004 / X pixel size (deg)                            
CDELT2  =  7.679400796794E-004 / Y pixel size (deg)                            
CROTA1  =  8.413525896584E+001 / Image twist X axis (deg)                      
CROTA2  =  8.412043323104E+001 / Image twist Y axis (deg) E of N if not flipped.
CD1_1   = -7.845760114097E-005 / CD matrix to convert (x,y) to (Ra, Dec)       
CD1_2   = -7.639002636741E-004 / CD matrix to convert (x,y) to (Ra, Dec)       
CD2_1   = -7.638152826643E-004 / CD matrix to convert (x,y) to (Ra, Dec)       
CD2_2   =  7.866608631869E-005 / CD matrix to convert (x,y) to (Ra, Dec)       
PLTSOLVD=                    T / Astrometric solved by ASTAP v2022.10.22.      
COMMENT 7  Solved in 0.3 sec. Offset 0.0".
COMMENT cmdline:"C:\Program Files\astap\astap.exe"  -f "C:\Users\Steve\Desktop\2
COMMENT 30401223021.fit" -r 15 -z 0 -t 0.007 -m 1.5 -sqm 256 -analyse2 15
END

[ButterFly]

345mm is right. Uncheck the "delete files" box under the solver settings to keep the files after closing out. Or, just open the file while the program is still running. Check the header for the correct pixel and chip sizes.

Some of them solve while others don't and your focal length is right, so it's sounding more like exposure gain at different parts of the sky. Try turning up the gain - it's over to the left of the slider.

[spz56]

I finally had a clear 90 minutes last night (not enough to image though; third try for the Leo Triplet). I must say that PlateAlign worked very well with no plate solve fails. I ran PlateAlign, ASPA, and PlateAlign again. I was good to go with PA under 2". Love the app - thanks again for developing it!

 

Steve

[Scottless]

@ButterFly would you be open to publishing your source code to github to enable collaboration? I'd love to contribute with automated alignment routine. E.g. click a button, go drink coffee, come back and have 25 point system done.

[ButterFly]

@ButterFly would you be open to publishing your source code to github to enable collaboration? I'd love to contribute with automated alignment routine. E.g. click a button, go drink coffee, come back and have 25 point system done.

No need for Github.  The source is available here subject to the included license.  This already is an "automated alignment routine. E.g. click a button, go drink coffee, come back and have 25 point system done" that works with CPWI and a few other mounts.  If you have any specific suggestions, I would love to hear them.  If it's something I like, I'll implement it, if I can.  You are free to make permissible changes and/or distributions under the license.  The license says what you need to do.

 

Again, I'm always open to suggestions.  Use it for a while, then see what else you need.  Other software does other things such as focusing, polar aligning, and collimating much better, so I really don't see the need to add much of anything else here.  I would like for it to work with more mounts, but it's up to the mount manufacturer to provide that ability, so it's out of my hands really.

 

If you really want to go ape****, you can just accept whatever RA/Dec coordinates the mount sends out as "encoder counts".  Write your own TPoint™-like model and send the mount the RA/Dec coordinates it needs to actually get to where it wants to go.  That would be fully extensible to every ASCOM mount - and a lot of work!

[Scottless]

@ButterFly you are right - initially I thought that SavePoints functionality saves the points as Ra/Dec, which obviously dip below horizon at different times of year. However, since it expressed them as Alt/Az, I suspect those pairs will work fine year-round, but translate to different Ra/Dec coordinates.

 

Using this tool definitely improved my pointing accuracy, how if only CPWI could use PXP model for tracking too.

[ButterFly]

Trees and buildings are what I had in mind. The Milky Way is much easier to solve in because there are more stars there, so different times of the year can mean exposing a little longer.

CPWI does have RA+Dec tracking available. Get enough points so that your polar error is stable. If it changes a lot when you add point ten, either add a few more, or get rid of point ten. With my AVX, the errors are largely in Dec once I get to twenty points. That's honestly the real limit of the software model - the hardware.

[iceguy]

Thanks for this!  My StarSense camera decided to stop working and this filled in perfectly.