More SGP woes

[Phillyo]

I cannot, for the life of me, get Sequence Generator Pro working correctly. I'm getting very disheartened and thinking I've wasted a small chunk of money on it. No matter what I try, I can't get plate solving working. I've tried Elbrus, I've installed Astrometry.net and all the associated files and I've tried the free version of pinpoint and it just fails, everytime. The auto focus routine takes it from close to being in focus to WAY out of focus before stopping and saying it's finished. Auto meridian flip doesn't work and it just sits there when it gets to the meridian too.

Anyone here manage to use SGP and think it's worthwhile?

Anyone suggest other software that can do it all that might work better?

Thanks,

Phil.

[Earl]

SGP rocks and is by far the best application for the price i have come accross.

Plate solving is a pain, and its paramont that you have image scale right to get it to work, also it needs to be dark enough i have found too, solving to early always failed for me. I got pinpoint also as the solving engine and this is more relaiable. try the 45 day demo.

Auto focus, you need to get your focus good first take a note of the step posistion so you can go back to it at will, then its very important to make sure the step size you are using is large enough to make a good v curve on the graph, this step size is specific for your setup. br a touch on the generous size with the value you input. ie if you work out its 25 make it 30.

Not tried Meridian flip as I have a habit of imaging to the meridian then chooising another target due to the mount I used to have.

There are some other available but some are not cheap.

Setting up SGP is very system specific and you cant just copy and paste someone elses setting you do need to work things out for your gear.

[dmahon]

I get great results with Elbrus, but I agree the settings are paramount. Works 9 times out of 10, and when it fails over to astrometry.net, that takes care of the 1 in 10.

Focussing, once I set it to take 8 second exposures, works fine too.

[Phillyo]

Thanks Earl,

I'm about 99% certain I've got the scale set up correctly? (Megrez 72 with 0.8 FR onto a QSI 583wsg = 3.219) I don't put an angle in as I don't know how to work that out? It still fails 9 times out of 10. I end up using Astrotortilla to platesolve through APT, then once I know I'm on target and guiding, I have to switch back to SGP to begin the imaging sequence. Which obviously isn't ideal?

dmahon,

I don't know how to change the exposure length when it's auto focusing? Manual I know how to do using sub frames etc but on auto it takes 20sec exposures so it takes an age

Thanks both,

Phil.

[Earl]

Thanks Earl,

I'm about 99% certain I've got the scale set up correctly? (Megrez 72 with 0.8 FR onto a QSI 583wsg = 3.219) I don't put an angle in as I don't know how to work that out? It still fails 9 times out of 10. I end up using Astrotortilla to platesolve through APT, then once I know I'm on target and guiding, I have to switch back to SGP to begin the imaging sequence. Which obviously isn't ideal?

dmahon,

I don't know how to change the exposure length when it's auto focusing? Manual I know how to do using sub frames etc but on auto it takes 20sec exposures so it takes an age

Thanks both,

Phil.

Have you solved a sub from the session after you have used Astrotortilla to platesolve, it might be a case of your plate solve sub not been long enough to get enough stars?

[Barry-Wilson]

Phil

Try not to despair, I do share your frustration when you're trying to get software to work and precious imaging time is wasted.

There is an option within the filters tab in the profile and Control Box to set exposure time for autofocus (you may even be able to set it within the autofocus options in the Control Box). I'm away from my pc to direct you exactly. I found the PDF manual to be very helpful in getting all of the basic settings: have you downloaded it?. I use a 10 s LUM sub binned 2x2 for autofocus with 15 steps for my FSQ and 20 steps form WO FLT. Both require backlash settings to give a symmetrical curve.

I use PinPoint which solves very quickly, before that I used Elbrus which solved about 80% of the time. You have to make sure that the file path to the GSC files is correctly set up as detailed on the web.

I do not have experience of other software except Nebulosity for capture and sequencing, so my opinion is somewhat limited: however, for what it is worth, I believe SGP is worth persevering. Its mosaic wizard is superb, as is the rest of its functionality. The forum is very helpful if you haven't yet drawn on its knowledge. There's plenty of folk willing to help and you can PM me and I'll help more when I'm back in the UK after my holiday. Keep at it and best wishes!

Barry

[jwellman]

Thanks Earl,

I'm about 99% certain I've got the scale set up correctly? (Megrez 72 with 0.8 FR onto a QSI 583wsg = 3.219) I don't put an angle in as I don't know how to work that out? It still fails 9 times out of 10. I end up using Astrotortilla to platesolve through APT, then once I know I'm on target and guiding, I have to switch back to SGP to begin the imaging sequence. Which obviously isn't ideal?

dmahon,

I don't know how to change the exposure length when it's auto focusing? Manual I know how to do using sub frames etc but on auto it takes 20sec exposures so it takes an age

Thanks both,

Phil.

Phil, 

Sorry you're having issues.  For Elbrus the angle must be populated or you can use the "Any Angle" solve which will take longer but it will attempt to figure out the angle for you.  

If you have internet access where you're at I would recommend using the Astrometry.net solver which is the easiest to setup. If you don't have internet access you can try using ANSVR which will using your existing Astrometry.net installation (as part of Astrotortilla).  

As for the auto focus exposure time you can set those values through the auto focus settings area.  Also I would recommend binning rather than sub frames here:

http://www.mainsequencesoftware.com/Content/SGPHelp/AutoFocus.html

Hope that helps,

Jared

[Phillyo]

Thank you all for your replies.

Barry,

I found the autofocus settings in the filters tab you mentioned and I have changed that, so thank you for that. I think my issues lie with calculating things like how many steps I need for the focus, how many points to do, and how on earth I figure out backlash settings!? This is all very new to a newbie imager. A lot to take in and as you say it's frustrating when it's a beautiful crystal clear night and I'm bashing my head against a wall trying to get the software to point where I want it to. Especially when I know that I can fall back on using something like APT with CDC and Astrotortillia and it all works ok. I just wanted something a little more integrated etc.

I've read through the manual before but I will go through it again that's for sure. I'm surprised at the lack of videos out there helping out, unless I'm looking in the wrong places (youtube mostly).

Jared,

I've tried Elrbus, the free trial of pinpoint and Astrometry.net and I have follwed online guides for installing and using them but they don't work. Ever. I do have internet access as I control everything from within the house once I have the mount polar aligned.

How do I calculate my angle? Is there anywhere I can upload a picture and it'll tell me the angle? Scale I'm guessing is the resolution and I get that data either using the formula or I use 12D string fov which gives me 3.22" (Megrez 72 with QSi 583 x0.8 FF/FR so FL for 346).

Thanks again everyone. I will get this sorted out!!

Phil

[ChrisLX200]

Using a calculated plate scale is no good, you need to solve a sub and use the figure from that. It is a pain that Elbrus is so inflexible about scale. I have several configurations (different scopes, with/without FF or Reducer) and had to go through this process for each, I actually used AstroArt to plate solve the image then transferred the data to Elbrus. It is very rare that Elbrus fails for me now but two things do still cause problems for it - a light background and mis-shapen (particularly) or poorly focused stars. I notice this more with the SCT, the refractor tends to produce impeccable star images but the old LX200 at longer focal length does not. You can get an idea how well Elbrus is performing from the 'confidence' figure it displays after a successful solve, it should be up above 70% if things are all good, if it drops below ~40% you are getting marginal results. I have tweaked a friends setup by changing the plate scale by increments of 0.01 arcsec/pixel and seen a 10% improvement in confidence value - it is that sensitive to the scale data - and this makes a big difference when sky conditions are less than perfect.

ChrisH

[Phillyo]

Hi Chris,

Thanks for explaining that. Do you know of a good way for me to solve an actual image rather than using a calculated one? I'm guessing AstroArt isn't free to use so that's out of the question. Would astrometry.net do it? Or Astrobin?

Phil

[ChrisLX200]

Hi Chris,

Thanks for explaining that. Do you know of a good way for me to solve an actual image rather than using a calculated one? I'm guessing AstroArt isn't free to use so that's out of the question. Would astrometry.net do it? Or Astrobin?

Phil

Yes Astrobin will do it, I just looked and the data it provides for scale appears to be accurate

ChrisH

[Phillyo]

Thanks Chris,

I'll go make myself an account then!

Phil

[Phillyo]

Ok so I've managed to get it solving an image I took a few weeks ago using Elbrus and Astrometry.net but pinpoint isn't working still. I'm guessing that's because I'm using the free version (trial version?) rather than the full one though.

My next question is how I increase the confidence level? It's always coming back at 50. The scale is 3.21 so i've set that as the scale under the camera settings. It's still giving 50...

Thanks for all the help so far!!

Phil

[ChrisLX200]

Ok so I've managed to get it solving an image I took a few weeks ago using Elbrus and Astrometry.net but pinpoint isn't working still. I'm guessing that's because I'm using the free version (trial version?) rather than the full one though.

My next question is how I increase the confidence level? It's always coming back at 50. The scale is 3.21 so i've set that as the scale under the camera settings. It's still giving 50...

Thanks for all the help so far!!

Phil

Ah. Caveat - if it uses Astrometry.net (either online or via a local installation) rather than Elbrus then it _always_ returns 50%. Don't ask me why I have no idea and it just does! So was this a blind-solve? How long did it take (might be an indication whether it used the Astrometry.net or not). Did you select Plate Solve or Blind Solve? Can you load another random image then right-click on it and select Plate Solve not Blind Solve (note that you will have to enter the image scale in the box because the scale would have been entered into the FITs header when it was taken and it will use that rather than the new figure of 3.21 - even if you have now got it entered in the settings sheet).

ChrisH

[Phillyo]

Ahh it might've been a blind solve as it took about 20-30 seconds maybe. I did click Plate Solve, but I have it set to fall back to blind solve if it fails. I'll turn that off and try again tomorrow.

On a side note, when using pinpoint as a stand alone program it solves in about 8 seconds. In SGP, it doesn't solve at all. How strange?

Phil

[ChrisLX200]

Ahh it might've been a blind solve as it took about 20-30 seconds maybe. I did click Plate Solve, but I have it set to fall back to blind solve if it fails. I'll turn that off and try again tomorrow.

On a side note, when using pinpoint as a stand alone program it solves in about 8 seconds. In SGP, it doesn't solve at all. How strange?

Phil

That would be about right, Pinpoint is quick - but so is Elbrus when it works, but if it fails then it drops back to a blind solve which will take longer (and 20-30 sec is correct for that).  If you have Pinpoint then you can get SGPro to work with that, seeing as it costs more ($149 IIRC) than SGPro itself not many users have it though.

ChrisH

[Phillyo]

I only have the 45day free trial of pinpoint...mostly because of the price!! If I can get Elbrus working then I'll use that. It's funny that Elbrus will solve when I use it as a stand alone program too, but won't work in SGP. I WILL get to the bottom of this!!

Thanks Chris,

Phil.

[ChrisLX200]

I only have the 45day free trial of pinpoint...mostly because of the price!! If I can get Elbrus working then I'll use that. It's funny that Elbrus will solve when I use it as a stand alone program too, but won't work in SGP. I WILL get to the bottom of this!!

Thanks Chris,

Phil.

Remember that the figure for plate scale is saved together with the sequence file, so if you load up an old sequence file which you have perhaps added more targets to then that will over-ride your new setting. Same with the equipment profiles. The first thing I do in a session is use Frame&Focus to download a 10sec image, I use that to check initial focus and then right-click on it and choose Plate Solve (making sure to check the image scale is correct for the hardware I'm using). Also ensure that in your filter configuration you have data entered for suitable exposure times for the plate solver and autofocus, otherwise it will use the same default exposure as for the Luminance filter (which will probably be too short to get a decent image for plate solving when using an RGB filter - and definitely so if using a NB filter). Note I'm thinking off the top of my head here without the program in front of me to check!

ChrisH

ChrisH

[Phillyo]

When I load a picture to plate solve it's normally taken from a while ago so it was taken with a different software package than SGP. When I then click on plate solve it doesn't give me any info at all. The RA and DEC say 0, Angle says 0 and Scale says 0. I manually put in the scale (3.18 to 3.22 varying it to see how it goes) then click on Plate Solve. It then takes ages and fails so falls back on a blind solve by the look of it.

Do I need to use a specific image?

Phil

[Phillyo]

Ah ha!! I've just tried it with one of the very few SGP images I have taken and it added the data automatically when I clicked the image and clicked Plate Solve. It solved it in under 10 seconds and has a confidence level of 66.

I think we're getting somewhere here!

Phil

[ChrisLX200]

Ah ha!! I've just tried it with one of the very few SGP images I have taken and it added the data automatically when I clicked the image and clicked Plate Solve. It solved it in under 10 seconds and has a confidence level of 66.

I think we're getting somewhere here!

Phil

Yes, without the RA and DEC info in the FITs header it will do a blind solve... SGPro populates the header when it takes the images.

ChrisH

[Phillyo]

Chris you're an absolute star!

I'll give this a try the next time it's clear, maybe sometime in 2015 Thanks again! Hopefully I will be able to start getting this all working together nicely now. I just need to figure out my eggy stars and I should be away!!

Phil

[ChrisLX200]

No probs, hope it all works (tomorrow night looks hopeful here if the winds don't ruin things...)

Note that when starting a new target, and if I haven't used SGPro's Framing and Mosaic Wizard (brilliant add-on that is BTW and I use it all the time now) then when you have plate solved the initial position and you're happy with the framing, then you will see the option in the success report dialog box to save the positional data (RA and DEC) for the target. Choose the right target if you have more than one in the sequence because it defaults to the top one! That will add the co-ordinates to the sequence and the next night you can just right-click the target in the sequence list and use Centre Target to get exactly back in postion. Change the default accuracy from 50 pixels to (say) 5 pixels, and change the number of iterations from the default 1 to 5.  SGPro will then be allowed 5 goes at nudging the scope to get it exactly right and I've always found 5 enough but YMMV. If you used the Wizard then the RA and DEC co-ordinates will be set for you, but you still need to change the default accuracy and iteration count. If you do this then when stacking the subs will fall on top of each other with very little needing to be cropped off the sides even after setting up on many different nights.

ChrisH

[Merlin66]

ChrisH,

I downloaded the trial SGP and must say it looked impressive.

Unfortunately the show stopper for me was it's inability to use the AstroArt plate solving. (It will use AstroArt guiding??!!)

I'm committed to AstroArtV5 for image acquisition, guiding and processing for the spectroscopy work....

[ChrisLX200]

ChrisH,

I downloaded the trial SGP and must say it looked impressive.

Unfortunately the show stopper for me was it's inability to use the AstroArt plate solving. (It will use AstroArt guiding??!!)

I'm committed to AstroArtV5 for image acquisition, guiding and processing for the spectroscopy work....

Use what you find works for you    I purchased AstroArt5 myself and use some of the processing tools it offers, but never got into it for image aquisition. The Plug-in features of AstroArt are good also.

ChrisH