Go Back   IceInSpace > Equipment > Software and Computers

Reply
 
Thread Tools Rate Thread
  #1  
Old 14-05-2014, 03:31 PM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
AstroTortilla Plate solving in the Office?

I hope there are some experienced users of AT that can help.
I'm just looking for a solution (pun?) to identifying stars in my slit guiding field of view. The FOV is normally 10 x 10arc min (or up to 25 x 25arc min, depending on the scope used)
I haven't tried AT "under the stars" as yet, but I do have an image which I'd like to solve...
Is there a way to use AT without the scope/camera connections?
Or should I just try my luck with Astrometry.net????
Reply With Quote
  #2  
Old 14-05-2014, 04:22 PM
pluto's Avatar
pluto (Hugh)
Astro Noob

pluto is offline
 
Join Date: Dec 2011
Location: Sydney
Posts: 1,982
Hi Ken,
Yes you can use AT without being connected to your gear.
Just ignore the "Telescope" selection and choose "File open dialog" under "Camera".
Then hit "Capture and Solve" and it will prompt you for an image to solve.

I found this guide helped me getting the settings correct:
http://lightvortexastronomy.blogspot...and-using.html
Reply With Quote
  #3  
Old 14-05-2014, 04:36 PM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Hugh,
Thanks for that.
I'm reading the suggest tut as I type....
I followed the same installation, BUT....
When it came to downloading the index files I chickened out after FIVE hours!!! It kept going on and on and on.......Something obviously went wrong!!!!
Let me finish the Tutorial and if I feel brave enough, try downloading the indices again.
Reply With Quote
  #4  
Old 14-05-2014, 04:55 PM
gbeal
Registered User

gbeal is offline
 
Join Date: Oct 2004
Location: New Zealand
Posts: 4,345
I managed to make Maxim plate-solve for a while, when I had a "real" camera, an SX IIRC, but since then the DSLR and/or mirror-less sort don't lend themselves to this, unfortunately.

I'd dearly love to re-try it, with AT (I downloaded the files ages ago, but never got past that), as plate-solving is a definite plus for pointing if imaging. Magic even.
Gary
Reply With Quote
  #5  
Old 14-05-2014, 10:07 PM
technofetishism's Avatar
technofetishism (Nick)
Registered User

technofetishism is offline
 
Join Date: Oct 2011
Location: Adelaide, Australia
Posts: 194
Downloading all the index files for astrometry.net in astro tortilla is a long long long exercise, i think all up theres over 30gb worth, and it doesnt help the astrometry server isnt exactly fast.

Just select the range you need as it suggests,and you can always add more by re-running the installer.

The other good thing about using astrotortilla with its astrometry.net install is if you use Sequence Generator Pro you can tell it to use a local astrometry.net install for plate solving rather than the internet version so you can still blind solve when away from a connection

http://adgsoftware.com/ansvr/INSTALL.txt
Reply With Quote
  #6  
Old 14-05-2014, 10:28 PM
mithrandir's Avatar
mithrandir (Andrew)
Registered User

mithrandir is offline
 
Join Date: Jan 2009
Location: Glenhaven
Posts: 4,161
Quote:
Originally Posted by technofetishism View Post
Downloading all the index files for astrometry.net in astro tortilla is a long long long exercise, i think all up theres over 30gb worth, and it doesnt help the astrometry server isnt exactly fast.
I've got a pretty much complete set and it's about 20GB.
Quote:
Just select the range you need as it suggests,and you can always add more by re-running the installer.

The other good thing about using astrotortilla with its astrometry.net install is if you use Sequence Generator Pro you can tell it to use a local astrometry.net install for plate solving rather than the internet version so you can still blind solve when away from a connection

http://adgsoftware.com/ansvr/INSTALL.txt
The version of astrometry.net that comes with AT is fairly old. If you want a current version you'll need to download the source from subversion at http://astrometry.net/svn/trunk/src/ (or git if Dustin has got around to moving it) and build your own copy with Cygwin.
Reply With Quote
  #7  
Old 14-05-2014, 10:42 PM
technofetishism's Avatar
technofetishism (Nick)
Registered User

technofetishism is offline
 
Join Date: Oct 2011
Location: Adelaide, Australia
Posts: 194
Good to know, looks like I'll be recompiling it tomorrow then :-)

Or I'll get the latest snapshot build. :-) http://astrometry.net/downloads/
Reply With Quote
  #8  
Old 05-07-2014, 10:35 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
Okay, I'm getting quite frustrated with Astrotortilla. It just will not solve, no matter how I tweak the settings (I have all the index files)!

I'm running on Win 8.1.

I've tried limiting the index files or using auto. I've tried the most liberal of blind plate solving parameters, or tuning to match my equipment as well as possible. Nada.

Yet, if I upload any of the images I've been trying with AT onto Astrometry.net, then BINGO, I get a successful solve in 5 to 30 sec.

What the heck? I thought AT was using the Astronomy engine? Getting very angsty now - suggestions most welcome!
Reply With Quote
  #9  
Old 05-07-2014, 10:50 PM
Eden's Avatar
Eden (Brett)
Registered Rambler

Eden is offline
 
Join Date: Jan 2014
Location: Melbourne, Australia
Posts: 399
AT can be a bit temperamental, when I first tried it I couldn't get it to solve out-of-the-box.

If you have downloaded the correct plate sources, the only parameters you really need to worry about are --sigma and scale max/scale min. Increasing the sigma speeds up solving time but at the expense of accuracy. Once you've got it solving, you can tweak the search radius and scale max/scale min to gain some speedup.

If you feel like uploading a sample of what you want to solve I'd be happy to run it through AT and have a look for you.
Reply With Quote
  #10  
Old 05-07-2014, 11:24 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
Thanks Eden, here is an example, saved in 3 different formats (FIT, TIFF and JPG). None of them work on my AT, all of them work on Astrometry.net

https://dl.dropboxusercontent.com/u/.../PlateTest.zip

The FIT file is what comes natively out of my CCD capture software.
Reply With Quote
  #11  
Old 05-07-2014, 11:28 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
I found something weird (to me, at least). If I try the FIT, TIFF or JPG files I get from my CCD software, then AT fails to solve.

However, if I upload any of these files to Astrometry.net, get it to do a solve, and then download the "new-image.fits" file it supplies, then my local copy of Astrotortilla can solve on this in <10 secs!

So is it that AT doesn't like my file formats, and yet Astrometry.net can somehow handle them? Very strange. At least I know my local copy of AT is working fine -- just not with my file types...
Reply With Quote
  #12  
Old 06-07-2014, 01:11 PM
Eden's Avatar
Eden (Brett)
Registered Rambler

Eden is offline
 
Join Date: Jan 2014
Location: Melbourne, Australia
Posts: 399
You're welcome. I spent a few hours trying to get Astrotortilla to come to the party but alas, no cigar (or enchilada, for that matter).

After some tweaking, AT eventually claimed to have solved the plate after 2,748 seconds. Apart from being highly impractical, I looked up the resolved co-ordinates in a star chart and found that the stars at that location were similar to the reference image, but it was incorrect.

I uploaded the file to Astrobin and it claimed to be able to solve it too, but look at the attached image. Delta Muscae is nowhere to be seen in the image yet Astrobin thinks that the reference image was captured at that location.

Can you please attach the output image you received from Astrometry.net, I'd like to inspect the FITS header. I have a feeling it has populated the header with the co-ordinates it believes it has resolved and that Astrotortilla is using those variables as a hint.

I have forwarded a copy of your image to Astrobin and the Astrotortilla group for comment. There is nothing wrong with the image or the format (I saved it out in both Maxim and AstroArt to be certain) though.
Attached Thumbnails
Click for full-size image (platetest.jpg)
128.4 KB28 views
Reply With Quote
  #13  
Old 06-07-2014, 01:55 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
Thanks Eden, this is immensely helpful. The FITS that Astrometry.net gave me is here (900 Kb):
https://dl.dropboxusercontent.com/u/...lateHadar.fits

I checked the FITS header (using Fitswork) and you are correct that all of the fitting information is supplied there, so I suspect this is indeed giving AT a head start.

Now although I'm trying a blind plate solve (from a software perspective), I actually know that I was pointing my astrograph just near Hadar (which is just of the FOV), so I am certain that the original Astrometry.net solve was in the correct position. Here is the location of the solve page:
http://nova.astrometry.net/user_images/344663#annotated

I also tried it blind on an image of M20 I took with the same CCD (high quality stacked 25 x 60 sec etc.) and this also works fine on Astrometry but fails on AT (unless I use the Astrometry.net FITS):
http://nova.astrometry.net/user_images/345065#annotated

My subsequent AT solve also puts it in the right position, and solves in 13 seconds. By the way, here are the AT settings I used for the successful solve on the Astrometry FITS - it is tuned to my setup.
Attached Thumbnails
Click for full-size image (ATSettingsAmar.JPG)
58.8 KB36 views

Last edited by Amaranthus; 06-07-2014 at 02:08 PM.
Reply With Quote
  #14  
Old 06-07-2014, 04:18 PM
Eden's Avatar
Eden (Brett)
Registered Rambler

Eden is offline
 
Join Date: Jan 2014
Location: Melbourne, Australia
Posts: 399
Nailed it.

I successfully managed to get AT to blind solve your original image in 7.9s. These settings don't leave much room for adjustment -- for example, if you increase --sigma to 3, it won't solve. To arrive at this point, I switched from my default scale of degrees width to arcseconds per pixel and after some fiddling around, it solved it in 3 minutes. I then worked it down from there.

Glad to be of assistance.

Try your M20 and let us know how you go.
Attached Thumbnails
Click for full-size image (7-9s.jpg)
126.9 KB51 views
Reply With Quote
  #15  
Old 06-07-2014, 05:15 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
Great work! I got a solution on the original (near-Hadar) image in 11.3 sec.

Unfortunately it's still being temperamental - the same parameters took 109 sec to yield a solution on my M20 image (based on a 60 sec sub in FIT format), and 94 sec on another shot I took near Rigel Kent.

So I think it's just a matter of me playing around with the settings further to balance the speed vs solving success trade-off. At least I know that it CAN work... Brilliant work Eden.
Reply With Quote
  #16  
Old 06-07-2014, 05:41 PM
mithrandir's Avatar
mithrandir (Andrew)
Registered User

mithrandir is offline
 
Join Date: Jan 2009
Location: Glenhaven
Posts: 4,161
Quote:
Originally Posted by Amaranthus
However, if I upload any of these files to Astrometry.net, get it to do a solve, and then download the "new-image.fits" file it supplies, then my local copy of Astrotortilla can solve on this in <10 secs!
The online version of astometry.net loads all indexes and runs them in parallel. You can do that too if you have a 64 bit *nix box with at least 32GB (maybe 64GB) RAM. I've tried building a 64bit version with Cygwin64 and it crashes 100% of the time because the external code it needs is not 64 bit clean.
Quote:
Originally Posted by Amaranthus View Post
Thanks Eden, here is an example, saved in 3 different formats (FIT, TIFF and JPG). None of them work on my AT, all of them work on Astrometry.net

https://dl.dropboxusercontent.com/u/.../PlateTest.zip

The FIT file is what comes natively out of my CCD capture software.
fitsverify says the DATE_OBS (2014-07-03 T08:22:30) has an illegal format.

fitscheck complains of no checksum in HDU #0.

Running on my Win7 64bit i7 box:
solve-field (SVN_24486) running completely blind without downsampling completed in just over 5 minutes.

Field: PlateTest.fit
Field center: (RA,Dec) = (210.4, -61.12) deg.
Field center: (RA H:M:S, Dec D:M:S) = (14:01:40.486, -61:06:58.688).
Field size: 58.6678 x 43.8176 arcminutes
Field rotation angle: up is -99.6558 degrees E of N

solve-field (SVN_24486) running completely blind with downsample 2 completed in 1 min 49

solve-field (SVN_24486) knowing the FOV is about 40x60 arcmin and within 5 deg of RA 14h Dec -61deg completed in 12 seconds.

If AT can get the position the mount claims to be pointing and that is within about 5 deg of the truth, and a reasonable approximation to the field size, by passing that to astrometry.net it solves much faster.
Reply With Quote
  #17  
Old 08-07-2014, 03:11 PM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 2,064
If you run your 64 bit *nix box off SSD, you can still get very fast solves with less than 32Gb and running in parallel. I have 24Gb and it works well.
Reply With Quote
  #18  
Old 09-08-2014, 04:16 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
So, I seem to be getting close to consistency with AT on my desktop now.

I started by setting the dimensions to arcsecpixels, setting a wide search range, and then seeing what AT reported as my FOV (once it eventually solved - I had to move around the sky a bit to get one). Now although I could calculate the FOV/scaling in theory for my imaging train, the reality is inevitably somewhat different due to spacing etc.

I had calculated 4.62 arcsec/pixel, for instance (with an ED80T, FF and SSG3), and AT estimated it as 4.675 (average of 10 solves). So once I had this figure, I set the tolerance for solving to 4.64 to 4.71 (giving it a narrow 3' margin on either side), and I got a mount-informed pointing (not blind) solve every time, always in <10 sec, and no more fails. A big win!
Reply With Quote
  #19  
Old 10-09-2014, 03:02 PM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
AstroArt Option??

back to looking at AT.....
I use a Lodestar guider under AstroArtV5, acquisition of spectra with an ATiK 314L.
The on-axis guide image is roughly 30 x 18 arc min...
BUT I can't see AstroArt as an option. Can't access the camera...
The AT discussion group requires membership of sourceforge which I'd prefer not to do...
Any ways around this one??
Reply With Quote
  #20  
Old 10-09-2014, 03:10 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
Can you do it as an ASCOM camera? The other options is the File/Open path - I often use this when stuck (e.g. camera flagged as busy).
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 08:58 PM.

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