Astro Tortilla no longer working ! Help !

[halli]

Hi

I own a 130pds and have recently bought a GSO RC8.  I have really enjoyed using Astro Tortilla with the 130 pds and it has worked really successfully.  I followed the Light Vortex Astronomy tutorial which was excellent to set it up initially.  

I now want to use it with my RC8 which has a smaller field of view, so I downloaded all the necessary files between the smallest an largest field of view dimensions ie the largest for the 130pds and smallest for the RC8 and also applied the new field of view for the RC8 in the solver config.  So now I have a settings file for the 130pds and also the RC8 depending on which scope I use.  The only real difference being the field of view settings for each scope.  

However when using it with the RC8 it simply doesn't want to solve !     

Im pretty sure the star shapes are ok in the RC8 images and it appears to be able to extract plenty of sources during the solving but fails to match them.  

I recently uploaded an RC8 image to Astrometry.net and it was solved by the web site fairly quickly.  Also when using it with the 130pds with larger field of view solver settings (which is the only difference) it works fine.  

It is really baffling and I was hoping someone out there has had a similar experience and perhaps can point me to something obvious I am doing wrong !    

I simply cant live without Astro Tortilla !

Thanks in anticipation

Ian

[Thalestris24]

Hi

It would be useful if you could include the focal length of your scope, details of your camera and your calculated fov. Plus the settings you're using with Astrotortilla.

Louise

[halli]

Hi Louise - thanks for your response.

The GSO RC8 is 203mm F8 @1624 mm focal length - camera is a Canon 1100D.

Calculated field of view from sky at night FOV calculator is 46.84' x 31.23' ie  0.78 X 0.52 degrees.

My solver settings are 

 

I took this screen shot  from the Light vortex tutorial page, the only difference is that my  FOV settings are scale max O.78 and scale minimum 0.52  - the Vortex guy also has a GSO RC8 coincidentally but with a different camera !

 the custom options setting  is therefore -

-sigma 70 --no-plots -N none -H 0.78 -L 0.52 -r --objs 100

Hope this makes sense !

Cheers

Ian

[Thalestris24]

Hi

Looks like your max should be about 1 deg and min about 0.1 deg but I'll have a closer look tomorrow. Also, a 10sec exposure seems rather long... Usually, something in the range 1 - 5s is ok.

Louise

[Thalestris24]

Hi

Looks like your max should be about 1 deg and min about 0.1 deg but I'll have a closer look tomorrow. Also, a 10sec exposure seems rather long... Usually, something in the range 1 - 5s is ok.

Louise

Sorry - that was rubbish! The best thing is to use an online plate solver to get the fov for an actual image and you can enter those values for the max and min. However, you need to calculate the larger and smaller values to determine which index files to download. I'll look again tomorrow!

Louise

[kalasinman]

The instructions say to subtract 20% from min and add 50% to max to start with.

[Thalestris24]

The instructions say to subtract 20% from min and add 50% to max to start with.

But if you have a platesolved image you should be able to use the actual values straightaway, I think.

[beamer3.6m]

You said you minimum was 0.52 but you have your astrotortilla settings at 0.525. Surely you shoukd have set it to wider than 0.52 so 0.45 etc.

Make sure you have the index files to cover this.

[Thalestris24]

Hi again

Hopefully, my brain is working a bit better today!

Anyway, here is screen capture of my current AT settings for a 130pds and 1100d, with an effective focal length of 740mm as I'm using a GSO coma corrector:

Now this is after AT has solved an image and the config file has been saved. As you can see, AT is using a scale min of about 10% less of the actual horizontal fov, and a scale max of about 10% greater than the actual fov.

I know in the past, I've put 0 in the scale min, and slightly greater than the actual fov for the scale max and it's solved ok. Now I also have the scale refinement set to .1, i.e. 10%. Presumably, if I had this set to zero then AT would reflect the exact fov for a solved image. However, tbh, I've never actually tried it as my settings work fine and I leave them alone! I sometimes fiddle with the sigma value depending on sky conditions. If it's very clear then sigma can be low or set higher to limit the number of stars detected and reduce the effects of noise in the image. Where I am in Glasgow, I think sigma = 30 is about the lowest I've set it. Search radius is set to 10 since I'm confident my gotos are good It can be up to 45deg, I believe, though that would only needed if you have a big goto error.

I'm not sure how long exposures effect the solvability. I guess if you're tracking very well and there's little apparent star movement over the exposure time then maybe 10s is ok. The -c parameter allows for a very small amount of distortion from roundness.

You haven't mentioned a coma corrector? (or is it field flattener with an RC?). I'm not sure how having a distorted field will affect solvability.

Hth

Louise

Edit: Of course, having the right settings is useless unless you also have the necessary set of index files. You can turn the log output on to see what AT is doing when it's trying to solve

ps the -c parameter should have a value of 0.02 next to it (I suppose windows messed that up!)

[halli]

Hi Louise - thanks very much indeed for the info - appreciated  - and to all who responded.  

Im pretty sure I did do the 20% substraction etc when I downlaoded the reference files Kalasinman- so I should have enough index files to cover what is needed.  I have also ensured my FOV settings are covering the right range Beamer 3 - thanks for the comments guys.

I have confirmed the field of view for my RC8 from a plate solved at astrometry.net as you suggested previously Louise (good idea !) and it is not too far away from the calculated FOV so I dont think this is the issue.  I have fine tuned the FOV settings as a result  and I have managed to plate solve an image from my RC8 using Astro Tortilla bit it took 3915 seconds !  So this is useless in practice.  Im pretty sure I have downloaded all the required reference files.  I think the problem is that it finds too many sources over 1500 to do the solving and cant be looking at the correct batch to start  - not sure why this is - I thought it might look in the ones  with the correct FOV first !  

The information you provided about your settings Louise will mean I can experiment a bit more in the light of your experience to see if I can spot the problem and get things to improve - I am hopeful this will lead to a solution - thanks a lot..

I am not using a focal reducer with the RC8 yet  - thought I would try it native first !

I will try and apply my brain this afternoon and experiment further - after I have seen England losing to Australia in the cricket !

Anhow thanks for the valuable input and I will let you know how I get on

Best wishes

Ian

[Thalestris24]

Hi Louise - thanks very much indeed for the info - appreciated  - and to all who responded.  

Im pretty sure I did do the 20% substraction etc when I downlaoded the reference files Kalasinman- so I should have enough index files to cover what is needed.  I have also ensured my FOV settings are covering the right range Beamer 3 - thanks for the comments guys.

I have confirmed the field of view for my RC8 from a plate solved at astrometry.net as you suggested previously Louise (good idea !) and it is not too far away from the calculated FOV so I dont think this is the issue.  I have fine tuned the FOV settings as a result  and I have managed to plate solve an image from my RC8 using Astro Tortilla bit it took 3915 seconds !  So this is useless in practice.  Im pretty sure I have downloaded all the required reference files.  I think the problem is that it finds too many sources over 1500 to do the solving and cant be looking at the correct batch to start  - not sure why this is - I thought it might look in the ones  with the correct FOV first !  

The information you provided about your settings Louise will mean I can experiment a bit more in the light of your experience to see if I can spot the problem and get things to improve - I am hopeful this will lead to a solution - thanks a lot..

I am not using a focal reducer with the RC8 yet  - thought I would try it native first !

I will try and apply my brain this afternoon and experiment further - after I have seen England losing to Australia in the cricket !

Anhow thanks for the valuable input and I will let you know how I get on

Best wishes

Ian

Hi Ian

You want to increase the sigma value - probably to at least 30, maybe higher. The chances are that many/most of the 1500 'stars' are likely just noise... Also objs 50 seems to be sufficient for a good solve. So, as long as you detect 50 or so genuine stars you should get a solve, if you have the right index files. Clearly, if you haven't got a solve within a reasonable time then something's wrong so you might as well abort.

Louise

[halli]

It works at last..............!

Not sure what did the trick in the end but changed sigma value to 100,  objs to 50,  put -c as 0.02 and made sure the FOV was accurate from a plate solved at Astrometry.net.  I ran out of patience to do one change at a time to establish which it was !  However I think they all acted together towards the same end to reduce noise and make the solving more efficient.  

I had a few hours clear sky tonight and just tried it out in anger and it solved within a minute or so for most cases.

I can now set up my RC8 as quickly as I can with my 130pds !

Next thing is to get my new OAG finding guide stars and guiding the RC8 - quite fiddly to set up and needs some more work yet.

Anyhow thanks for all your input and support.

Special thanks to Louise for the detailed information which helped considerably !

Cheers

ian