My tutorial for installing, setting up and using AstroTortilla

[SnakeyJ]

Cheers Gib007,

I've tried the file permissions, resetting them to allow all users full control of files within my capture and cgywin folders, though I'm still getting the same error.

I've even licensed my neb3, though this was long planned

It won't even work on the open file dialogue selected - I did find some similar threads on CN and other forums, which advised re-installing the cgywin folder - though this is also unsuccessful. Guess I'll remove and install from scratch again.

Sorry to hijack your thread, but do appreciate the assistance and again thanks for a great tutorial - it'll be fantastic if I can get this working and sure it's only a matter of time and persistance!

[Demonperformer]

Thanks for this, Kayron. Lots of useful information on the settings to use. I can see why I have been getting some majorly long times to plate-solve!

Have you tried using AT polar alignment option? When I have tried using it, I have got some seriously screwy results: http://stargazerslou...riously-screwy/. I wonder if adjusting the settings as you have suggested would improve the PA results (although I can't see why it would, other than speed of processing)?

[oly2b]

Does anybody have any sample images which I test my AT installation?

I tried one jpg from another forum but AT could not obtain a solution.

It would be useful to have samples images and settings which are know to work so you play when the sky is not playing.

[Gib007]

Cheers Gib007,

I've tried the file permissions, resetting them to allow all users full control of files within my capture and cgywin folders, though I'm still getting the same error.

I've even licensed my neb3, though this was long planned

It won't even work on the open file dialogue selected - I did find some similar threads on CN and other forums, which advised re-installing the cgywin folder - though this is also unsuccessful. Guess I'll remove and install from scratch again.

Sorry to hijack your thread, but do appreciate the assistance and again thanks for a great tutorial - it'll be fantastic if I can get this working and sure it's only a matter of time and persistance!

Don't worry, you're not hijacking the thread! That is certainly very odd. It should definitely work if you are asking AstroTortilla to read an image file you provide manually! Indeed it might be a good idea to remove everything - uninstall AstroTortilla and then delete the cygwin folder as well. Re-install from scratch and test. Other than that, all I can advise is perhaps contacting the developers to see if they have specific advice as I have not encountered this issue before, sorry!

Thanks for this, Kayron. Lots of useful information on the settings to use. I can see why I have been getting some majorly long times to plate-solve!

Have you tried using AT polar alignment option? When I have tried using it, I have got some seriously screwy results: http://stargazerslou...riously-screwy/. I wonder if adjusting the settings as you have suggested would improve the PA results (although I can't see why it would, other than speed of processing)?

You're more than welcome! In reply to your question, no, I have not personally used the AstroTortilla polar alignment tool. This is mainly because EQMod's polar alignment routine is excellent and together with an iteration of Alignmaster, produces very good results in my experience. I may give it a try but at this very moment, I'm no authority figure to provide support for it, sorry! Looking at it though, it does not seem to give a real-time indication of the correction process as you turn your mount's altitude and azimuth bolts. For example, when Alignmaster is used, the program purposefully misaligns your second chosen star to the same amount as the polar alignment error, asking you to correctly centre the star but using your mount's bolts. Since you are looking at the star in question through an eyepiece with crosshairs, you can see how you are correctly the polar alignment. Further iterations then show you how much closer you are. AstroTortilla seems to only state how far off you are in each direction, but leaves you to your own devices in order to correct it by just the right amount in each axis.

Does anybody have any sample images which I test my AT installation?

I tried one jpg from another forum but AT could not obtain a solution.

It would be useful to have samples images and settings which are know to work so you play when the sky is not playing.

A way I like to test AstroTortilla at home is using the ASCOM simulator and Stellarium to monitor the telescope's (simulator) position. Go to Tools and Goto image. Here you can select an image file and AstroTortilla will plate solve it and slew the telescope (simulator) to the object(s) in the image. This is how I test at home. Settings in AstroTortilla will obviously alter how effective the plate solving is even for this feature.

[oly2b]

A way I like to test AstroTortilla at home is using the ASCOM simulator and Stellarium to monitor the telescope's (simulator) position. Go to Tools and Goto image. Here you can select an image file and AstroTortilla will plate solve it and slew the telescope (simulator) to the object(s) in the image. This is how I test at home. Settings in AstroTortilla will obviously alter how effective the plate solving is even for this feature.

This is what I have tried doing but using CdC, but I get the same messages on the log window

Any clue to what I may be doing wrong?

[kirkster501]

A great tutorial and I learned some refinements such as the -r and objs options. Its a really great program, you can home in on stuff like very faint galaxy clusters tht you woudl struggle to do without AT.

Anyone know what the "no plots" option does? I think there is an option somewhere where when it solves on, say M31 AT actually tells you this.

[andyo]

This is a great tutorial I used the -r and objs but it kept saying no solution so I removed it and it works again.However I change the downscaling option to 3 and this helps reduce the solve time considerably,not sure how it works though I think it has something to do with rescaling the image

[Gib007]

This is what I have tried doing but using CdC, but I get the same messages on the log window

Any clue to what I may be doing wrong?

This is odd and looks to me as if AstroTortilla is struggling to communicate with your camera (or imaging software). However, you say this is supplying an image to AstroTortilla for "Goto Image", right? If that's the case, then this shouldn't be happening. In fact, it looks as if it fails after 0 seconds and that should simply not be the case unless the image is corrupt or cannot be read properly. If it fails, it fails are some time has passed for finding a solution. Maybe something to do with the AstroTortilla installation or astrometry index files? It could warrant a complete reinstall. If you choose to do this, uninstall AstroTortilla, delete its folder and also delete the cygwin folder before proceeding to reinstall from scratch.

A great tutorial and I learned some refinements such as the -r and objs options. Its a really great program, you can home in on stuff like very faint galaxy clusters tht you woudl struggle to do without AT.

Anyone know what the "no plots" option does? I think there is an option somewhere where when it solves on, say M31 AT actually tells you this.

My thanks for your comments! I agree, AstroTortilla is absolutely fantastic and I no longer go out imaging without using it. When software performs plate solving, it identifies objects in the image that are recognisable objects from catalogues (e.g. M31, NGC2204, etc). This can take some time to do and because astrophotographers like to use laptops in the field, some slower than others (e.g. I use a netbook), it can worsen plate solving speed. AstroTortilla therefore has "--no-plots" as a default option to prevent it from wasting time in identifying the objects visible.

This is a great tutorial I used the -r and objs but it kept saying no solution so I removed it and it works again.However I change the downscaling option to 3 and this helps reduce the solve time considerably,not sure how it works though I think it has something to do with rescaling the image

That's interesting. In theory these options are a good idea because they sort the stars in "brightest first" and then of course you restrict how many you want to include. Perhaps you can leave "-r" by itself then, which sorts by "brightest first" but does not restrict how many stars are used. Whatever works best for you! Regarding downscaling, yeah, this is to make the images smaller and quicker to deal with. If you use a CCD camera, you can also just enable 2x2 binning for your plate solving image and this naturally downscales your image by 2, with the added advantage of capturing more stars in less time.

[kirkster501]

AT is one of the most useful programs there is for AP in my opinion. I am lost without it. It can place a small object absolutely bang on centre. M87 here is a good example.

Its even more powerful than that! You don't need an EP anymore to get aligned. Just make a rough note on your focuser where your focus point is, connect up and put focuser at that point, take an exposure, plate solve, then slew to a brighter star - Vega, Capella or Rigel are my faves - and get perfect focus. Then slew to photo object!! Simples! It is superb software, anyone not so using you are seriously missing out.

[jnc71106]

i get this error

2013-09-09 12:57:44,608 - astrotortilla - INFO - Solving...

2013-09-09 12:57:46,312 - astrotortilla - INFO - Reading input file 1 of 2: "50"...

2013-09-09 12:57:48,062 - astrotortilla - INFO - ERROR: Image type not recognized: Could not determine file type (does the file exist?): 50

2013-09-09 12:57:48,280 - astrotortilla - INFO - augment-xylist.c:585:backtick Failed to run command: /usr/lib/astrometry/bin/image2pnm.py --sanitized-fits-outfile /tmp/tmp.sanitized.EaKmJ4 --fix-sdss --infile 50 --uncompressed-outfile /tmp/tmp.uncompressed.4VdElZ --outfile /tmp/tmp.ppm.rmQhTa --ppm

2013-09-09 12:57:48,500 - astrotortilla - INFO - ioutils.c:605:run_command_get_outputs Command failed: return value 255

2013-09-09 12:57:48,500 - astrotortilla - INFO - No solution in 3.9s

i use the file open dialog as i use artemis for capture

any ideas ?

[jnc71106]

i removed the -r and objs and it does at least start the solving but fails to solve the image will have to go over the instal tute again to see if i've done something wrong

i'm using a 200mm lens with my atik and my FOV details are 2.57° x 1.92° can you confirm the indexes i need please

my scale min is 0.96 and scale max 3.90 is this correct ?

Edited September 9, 2013 by jnc71106

[cfpendock]

Just as a matter if interest, I recently started using this excellent software. But I couldn't make it work for 2 days - tried changing sigma, took out the space between -r and -objs 50 (big improvement - it started to solve), absolutely nothing worked - error messages, no image found etc. Although I have no idea what it does, I eventually changed "downscaling" to 4. Works a treat now, but why, I don't know. I had previously tried it on a number of very different images (different FOV etc) without success. Now, providing I adjust sigma to get between 30 and 100 objects, I get success nearly every time. Most importantly, using the "go to image" tool, I can easily continue imaging from when the last cloud stopped play. (I have to dismantle the scope and put it away each time....)

Chris

[RogerTheDodger]

Great tutorial - thanks for taking the time and trouble to write it.

Just one point that had me scratching my head for a while is that I cut and pasted the -r and -objs 50 text into my Custom Setting and kept getting a "Cannot find file "50"" error. This went away when there was no space between the -r and the -objs as follows "-r-objs 50"

Thanks again

Steve

try using --objs (that's with two minus signs)

[Gib007]

try using --objs (that's with two minus signs)

My thanks for pointing out this omission. Indeed the objs command is meant to have two minus signs, e.g.: --objs 50

Following reading the developer's recommendations, I have also updated the tutorial to recommend setting the restriction to 100 stars, rather than 50. The developers generally recommend that AstroTortilla is allowed to plate solve with between 100 and 200 stars, recommending you adjust your exposure time and sigma values in order to help with this as well.

i get this error

2013-09-09 12:57:44,608 - astrotortilla - INFO - Solving...

2013-09-09 12:57:46,312 - astrotortilla - INFO - Reading input file 1 of 2: "50"...

2013-09-09 12:57:48,062 - astrotortilla - INFO - ERROR: Image type not recognized: Could not determine file type (does the file exist?): 50

2013-09-09 12:57:48,280 - astrotortilla - INFO - augment-xylist.c:585:backtick Failed to run command: /usr/lib/astrometry/bin/image2pnm.py --sanitized-fits-outfile /tmp/tmp.sanitized.EaKmJ4 --fix-sdss --infile 50 --uncompressed-outfile /tmp/tmp.uncompressed.4VdElZ --outfile /tmp/tmp.ppm.rmQhTa --ppm

2013-09-09 12:57:48,500 - astrotortilla - INFO - ioutils.c:605:run_command_get_outputs Command failed: return value 255

2013-09-09 12:57:48,500 - astrotortilla - INFO - No solution in 3.9s

i use the file open dialog as i use artemis for capture

any ideas ?

I must admit I cannot personally diagnose the source of your problem from this log. The developers may be able to help more specifically though!

The log does seem to make reference to "50". Is that your filename or something? Try filenames without spaces (use underscores _ or dashes - instead, in that case). I doubt this "50" being referred to is from the objs command but if it is, ensure it is set with a double minus: --objs 50

Perhaps remove it to try and see what happens.

i removed the -r and objs and it does at least start the solving but fails to solve the image will have to go over the instal tute again to see if i've done something wrong

i'm using a 200mm lens with my atik and my FOV details are 2.57° x 1.92° can you confirm the indexes i need please

my scale min is 0.96 and scale max 3.90 is this correct ?

Considering your maximum FOV is 2.57° and your minimum FOV is 1.92°, we can say that your maximum FOV plus 50% is 3.90° and then 20% of your minimum is 0.38° (22.8'). You therefore seem to need astrometry index file 4207 for Narrowest level and astrometry index file 4213 for Widest level. Those two and the ones in-between cover your entire FOV amply.

Just as a matter if interest, I recently started using this excellent software. But I couldn't make it work for 2 days - tried changing sigma, took out the space between -r and -objs 50 (big improvement - it started to solve), absolutely nothing worked - error messages, no image found etc. Although I have no idea what it does, I eventually changed "downscaling" to 4. Works a treat now, but why, I don't know. I had previously tried it on a number of very different images (different FOV etc) without success. Now, providing I adjust sigma to get between 30 and 100 objects, I get success nearly every time. Most importantly, using the "go to image" tool, I can easily continue imaging from when the last cloud stopped play. (I have to dismantle the scope and put it away each time....)

Chris

Do you use a CCD camera or a DSLR? If you use a CCD camera, use binning instead of down-scaling. Binning effectively does the same thing to the plate solving image - it down-scales the resolution by the binning amount. However, with binning comes significantly increased CCD sensor sensitivity so you will pick up many more stars and make them all brighter, with the same or even with lower exposure time. This way you get the benefit of down-scaling but with the added plus of binning's increase in CCD sensor sensitivity. If you use a DSLR, this won't be an option. Please note down-scaling makes plate solving faster because the image is effectively halved or quartered in resolution so it is much faster to deal with. With down-scaling though also comes reduced resolution (in terms of picking up stars and details) so use down-scaling with care.

[Aaron Small]

Great tutorial. Solved my problem after a few days of frustration not being able to work on images that would work on astrometry.net. Finally figured out it was an extra space (I had "-no -plots" instead of "-no-plots"). Glad to have the -r and --objs options as they cut my solves in half. Some of the syntax on the command lines could be better, but who am I to question the programmer of such a powerful tool. Thanks again, Aaron

[Gina]

Great software, great tutorial and a great thread Pity the options are so critical - I've written plenty of software where there are option strings like this and parsing to allow for extra spaces etc. is relatively easy. I guess the author of AT is concentrating on the difficult bits and has certainly done a good job but the user interface can make all the difference as to whether the software gets used and particularly newbies to get to grips with it quickly.

[cfpendock]

My thanks for pointing out this omission. Indeed the objs command is meant to have two minus signs, e.g.: --objs 50

Following reading the developer's recommendations, I have also updated the tutorial to recommend setting the restriction to 100 stars, rather than 50. The developers generally recommend that AstroTortilla is allowed to plate solve with between 100 and 200 stars, recommending you adjust your exposure time and sigma values in order to help with this as well.

I must admit I cannot personally diagnose the source of your problem from this log. The developers may be able to help more specifically though!

The log does seem to make reference to "50". Is that your filename or something? Try filenames without spaces (use underscores _ or dashes - instead, in that case). I doubt this "50" being referred to is from the objs command but if it is, ensure it is set with a double minus: --objs 50

Perhaps remove it to try and see what happens.

Considering your maximum FOV is 2.57° and your minimum FOV is 1.92°, we can say that your maximum FOV plus 50% is 3.90° and then 20% of your minimum is 0.38° (22.8'). You therefore seem to need astrometry index file 4207 for Narrowest level and astrometry index file 4213 for Widest level. Those two and the ones in-between cover your entire FOV amply.

Do you use a CCD camera or a DSLR? If you use a CCD camera, use binning instead of down-scaling. Binning effectively does the same thing to the plate solving image - it down-scales the resolution by the binning amount. However, with binning comes significantly increased CCD sensor sensitivity so you will pick up many more stars and make them all brighter, with the same or even with lower exposure time. This way you get the benefit of down-scaling but with the added plus of binning's increase in CCD sensor sensitivity. If you use a DSLR, this won't be an option. Please note down-scaling makes plate solving faster because the image is effectively halved or quartered in resolution so it is much faster to deal with. With down-scaling though also comes reduced resolution (in terms of picking up stars and details) so use down-scaling with care.

Using the (CCD) camera is no problem - it was solving existing images which I had a problem with, but now the problem is "solved".....Thanks very much for the info anyway, I'll try switching off downscaling for when I am using the camera.

Chris

[Gina]

I use binning with AT - at least 2x2 and often more with the wide view of lenses. This way I can get solves down to 5-10 seconds.

[kirkster501]

I understand Annti the Finnish developer has a new version in progress as well.

[Gib007]

Great tutorial. Solved my problem after a few days of frustration not being able to work on images that would work on astrometry.net. Finally figured out it was an extra space (I had "-no -plots" instead of "-no-plots"). Glad to have the -r and --objs options as they cut my solves in half. Some of the syntax on the command lines could be better, but who am I to question the programmer of such a powerful tool. Thanks again, Aaron

I'm glad it helped you out!

Indeed it is a very powerful tool and an excellent piece of work.

Great software, great tutorial and a great thread Pity the options are so critical - I've written plenty of software where there are option strings like this and parsing to allow for extra spaces etc. is relatively easy. I guess the author of AT is concentrating on the difficult bits and has certainly done a good job but the user interface can make all the difference as to whether the software gets used and particularly newbies to get to grips with it quickly.

Yeah. I'm pretty confident however that are time passes and new versions are released, the developers will re-create the user interface to have a menu whereby you can see all available options and set them accordingly. All things considered, that's probably the easiest thing the developers could do in terms of actual programming and user interface design. I bet they are simply concentrating on making it work and so that it does its job well.

Using the (CCD) camera is no problem - it was solving existing images which I had a problem with, but now the problem is "solved".....Thanks very much for the info anyway, I'll try switching off downscaling for when I am using the camera.

Chris

I'm glad you have got it sorted!

I use binning with AT - at least 2x2 and often more with the wide view of lenses. This way I can get solves down to 5-10 seconds.

Me too - 2x2 binning is excellent as it down-scales the image for faster processing and at the same time, makes everything brighter with the same exposure time, picking up more stars and making them more well-defined.

I understand Annti the Finnish developer has a new version in progress as well.

Good to hear. It's certainly a must-have program for those of us controlling the mount via ASCOM.

[jnc71106]

I've got it working, although playing around with the ascom simulator it it doesn't reslew to my target after it has solved the image so will have to keep trying

Thanks for all the help getting to this stage

[Gib007]

Just to add to this thread that I have been using the settings proposed in the tutorial for several nights now and using 2x2 binning and a 10 second exposure (with a Luminance filter), I get very fast plate solving times of under one minute. This is on my netbook, too! It has a slow CPU, 2GB RAM and a 64GB SSD. 

[kirkster501]

I am starting to image with APT as my capture application.  It also works very well with AstroTortilla with my Atik 460 via the ASCOM drivers.

APT also supports dithering with PHD too.

[jimao22]

I am very envy on all of you because you did manage to use AT. I tried to do the same last few days and I am very frustrated because I cannot do it.

My set up is based on an EQ6 mount/EQMOD, an ATIK314L+ camera with TRUETECH filter-wheel, I am using MaximDL 5.22, on W7/64 and whatever I am doing, I receive the same message from AT - "no solution in 0.0 sec".

I installed AT in drive C:\\, Cygwin was properly configured to be downloaded, I re-installed AT several times, one time from the scratch, but same message over and over again.

I even sent an e-mail to http://sourceforge.net/p/astrotortilla/home/Home/ on mailing list, but no answer till now.

I will be very grateful if someone will tell me what to do because I'm stuck.

Thank you...

[daz]

Jimao22 - not all of us are having success .... 

Anyway, looking at your screenshots, I would say that AT is not able to connect to the camera to take the image....

Have you tried without using Maxim, and just connecting directly to the ASCOM driver for the camera?