Still can't get Astro Tortilla to solve

[oldpink]

not sure where I'm going wrong
I tried it again last night since I had a decent clear spell
centered on Arcturus which should have been an easy target but twice I tried and twice it failed to solve

yet if I upload to Astrobins or check online it plate solves the same image easily and correctly

anyone have any tips or advice on possible problems or things to try

[continuum]

It's probably only a matter of finding the right settings for your equipment - trial and error I'm afraid.

Have you looked through this tutorial - http://lightvortexastronomy.blogspot.co.uk/2013/08/tutorial-imaging-setting-up-and-using.html.

Well worth studying.

Ron

[IanL]

Not sure if I've pointed you at this before, but this is probably the best step-by-step tutorial for AT:

http://lightvortexastronomy.blogspot.co.uk/2013/08/tutorial-imaging-setting-up-and-using.html

First and foremost, I'd suggest you take and save a fair few images so that you can test during the day rather than being limited to precious imaging time.  Sounds like you already have some to work with than can solve, but here are a few tips to create a set of test data:

- Make sure you have the same set-up as you would normally use - scope/camera/reducer or corrector/filters/etc.

- Save them as JPEG (or shoot them as RAW and convert then to JPEG later).

- Take sets of images of different parts of the sky.  Use the maximum ISO setting for these as that's what you'll typically do when using AT for real.  Try different exposure lengths (5 seconds, 10 seconds, 30 seconds, one minute - unless you have excellent polar alignment and tracking don't bother going any longer than this as the stars will probably get trailed a bit and may not be identified by AT later).

- Make sure you have some identifiable star in the frame each time so that when testing you can figure out if AT did indeed solve the plate properly.  Name the images for that star so that you're not guessing what the target is when testing.

- Check your focus is spot on by examining some of the images in detail.  Don't wait till the next day as poor focus will also cause AT not to identify the stars properly.

By all means continue to test AT that evening if you want, but now you are in a really good position to continue testing during the day:

- You can ask AT to do a 'blind' solve just by using the 'file dialog' camera to open one of your saved images.  Bear in mind that if you do this, AT has no idea where the image is in the sky so it has to search all of the index files and that can take a long time.

- I'm assuming you are using EQMOD for mount control?  If so it would be better to use the simulator for testing.  So you'd fire up all your software as if you were connected to the mount, but instead of connecting to the actual EQ5 mount (in the various mount chooser dialogs in each piece of software) you choose the simulator option instead.  So fire up your planetarium (or stellariumscope if using stellarium), connect to the simulator and the usual EQMOD interface will fire up, but it will not be talking to the mount, just simulating one.  You then connect AT and any other software you need to the simulator.

- Now you should slew the 'mount' to the target of your test image - AT should show 'Slewing' and then 'Tracking' once on target.  Now use the file dialog camera to open your test image and have it solve.  This will be a lot quicker since AT has some idea where in the sky to start looking for the solution.  You don't need to be spot on the target, just within the number of degrees specified in the 'search radius' setting in AT. The smaller the radius, the quicker it will solve (or fail!) but conversely the more accurate your initial point has to be (This matters when using AT in anger, particularly if you are using AT to sync EQMOD alignment points as you go - make sure the search radius is big enough to cope with the error following the first slew of your scope when you have no alignment points).

Okay so now hopefully you are in good shape to do plenty of testing when the pressure is off.  I'd really suggest working step-by-step through the tutorial above as it is very thorough and covers all the bases.  Things that are likely to trip you up are:

- Having an image that can't be solved due to too few stars, too many stars or noise/focus problems.  I'd try uploading some images to http://nova.astrometry.net/upload .  This is exactly the same back-end software used by AT, so if you can get an online solution through it then you can be confident that you're not wasting your time as AT should also work once you get the setup right.

- Are you getting enough stars detected?  Check the AT status bar as it is solving - if you only get a few 'sources' detected (i.e. stars) then it might be that you need a longer exposure image.  I'd say you need at least 20 stars detected; it can work with fewer but not always.  I have no problems with my 80ED in this regard, but using my SCT I often only get 7 or 8 stars detected which can take a long time to solve or fails altogether, so I have to expose for at least a minute to have some chance of success.

- Bad star shapes is also a reason why you might get few sources detected - coma or longer exposures and poor tracking may cause this.  There is a star shape relaxation parameter detailed in the tutorial above I think, so try tweaking that a small amount to make it detect more stars.

- Conversely if you are getting hundreds or thousands of stars detected you will also have problems.  Unless you have very fast optics/long exposure/dense star field the cause is usually noise being picked up as stars.  You might try the downscaling option to kill off some of the noise.  N.B. If you are downscaling an using arcseconds per pixel (the "app" setting on the additional parameters) to specify the scale of your image to AT then you have to adjust it to compensate, if you're using 'degwidth' to specify the scale then you won't need to do so.

- The other option is the sigma setting which can reduce the number of stars to a manageable number, perhaps combined with the 'brightest sources' setting.  All this is detailed in the tutorial so I'm not repeating it here.

- Once you have a realistic number of stars detected and being used (say 30 to 100) then after that you just need to look at the following:

- Do you have the right index files installed for your image scale as discussed in the tutorial? Also double-check what astrometry.net solutions say about the radius of your image.  Try selectively adding or removing index files (by copying them out of the index folder).  If you don't have the right scale index files it won't solve, but equally having too many index files of the wrong scale will just slow things down unnecessarily.  Make sure you are using the up to date index files - at least one old set were released that had problems but the current AT installer should fetch the correct ones.

- Tweak the parameters in AT as described in the tutorial.  Make sure you are telling AT what the correct image scale is, otherwise it will fail.

If you can post any examples of when it is failing (images, status messages, logs, etc.) plus your AT settings and what index files you have installed, we might be able to give you some specific pointers as to why.

[oldpink]

I think it may be the star count
when processing the images in DSS it was reporting 18 stars though I could see more with my eye in the finished Tif

I will try something with more stars to see if that works

[Dave_D]

probably a daft question but are you focussed properly? i accidentally nudged my focus just a smidgen one day and AT had a merry fit at not solving.

[oldpink]

yeah focus was pretty sharp
Arcturus was flaring pretty bad due to the magnitude but the surrounding stars were pretty sharp

[frugal]

If I use the "downsample 2" parameter the number of stars it finds to match goes up from about 20 to about 300. If I do not down sample the image it never solves.

[oldpink]

thanks I'll look at that next time I'm out

[oldpink]

Finally got it to solve last night
I upped the sub time to 20 seconds that seemed to do the trick

thanks for all the help and advice