Slewing and Focusing with an H-Alpha Filter

I’m relatively new to Sequence Generator Pro, but I absolutely love it. I have it doing automatic slewing, photography, and focusing. I am using a filter-modded Canon T3i camera as my imaging camera. While I save the RGB frames for dark sites, I enjoy taking additional hydrogen-alpha frames using an Astronomik 6nm clip filter.

The bad thing about using an H-Alpha filter is that it significantly reduces the brightness of the frame and, as a consequence, the number of visible stars. This has led to two significant issues:

  1. When doing plate solving, I must do much longer exposure times in order to capture enough stars. Even 90 seconds at ISO 3200 hasn’t been enough. And since SGP doesn’t offer the option to guide while taking a plate solve frame, sometimes my stars are trailed.

  2. During the focus routines, longer exposures are necessary as well, but I seem to be limited to 100 seconds. It does seem to be working somewhat, but I can’t really tell if it is focusing on stars or noise due to the high ISO setting.

Can anybody make any suggestions on settings to use in this configuration? I lost three hours of sleep last night trying various things (unsuccessfully), and ended up shutting everything down and just doing dark frames. I’m hoping someone on this forum uses the same or similar setup and can offer some guidance.

Thank you!

Just learning SGP myself (lots of experience with other SW) and as far as I know there is no built-in way to do this with manual filters. With a software controlled filter wheel you just use a focus offset and focus with a clear or no filter.

You could go to a really bright set of stars (something like the Pleiades maybe) and, using control panel, run focus on that with the Ha. Then run a focus at your object w/o the filter. The difference in focus point would be your offset. From then on just focus w/o the filter first, then put the filter in and manually apply the offset. Since this is a manual filter change (if I understand correctly), it could clearly not do this by itself, however, so focus would drift eventually due to temp (although if you have a temp comp focuser that would help).

Maybe there is a way to have it pause while you insert the filter, apply the offset, and continue but even that sort of defeats the purpose of automation. I would have to reluctantly suggest you need a filter wheel that can be controlled by SGP (most can).

There are other automation programs which use FocusMax for focussing that can do what you want because FocusMax works differently and will slew to a bright star as well as adjust focus exposure. Those programs mostly will cost more than a filter wheel, however!

I have to disagree with you on this. I have already done multiple sequenced nights with an Oxygen-III filter. Though I had to increase the exposure time to get enough data, I had no problem getting the slewing and focusing to work. With the H-Alpha filter being more restrictive, I would imagine that my exposures need to either be at a higher ISO or longer exposures - there has got to be a way to make it work. Manually putting the filter in is not an option, because it would require me to remove the camera from the scope. My camera is inside of a DSLR cooler so it needs to stay sealed in there.

I don’t believe there is a 100 second limitation on AF frames. But ill verify that this evening. I know I’ve done 3 minute AF exposures before when setting my offsets. But something could have changed since then.

You are correct that there is no way to guide during some solver frames. We can see about adding this option. However it might be better to build an accurate model and forgo plate solving when shooting narrowband with a dslr. Otherwise centering and slewing will take a long while as the guider has to be started and settled prior to each frame and slew.


100 seconds sounds crazy long. Even with my 3nm Astrodon filters, I only
need 30 second exposures. And to compensate for long exposure time, use
temperature compensation.

1 Like

I use 30 second exposures as well. DSLR versus CCD of course.


Can you say something about the DSLR cooler you are using.



Please stay on topic here… unless the cooler specifically has something to do with SGPro or the OP’s issue, please take this off-line or over to CN. If you are asking because the cooler might affect or help to resolve the OPs problem, please… by, all means, continue.

Well, I did not say it could not be done, just that there was no built-in way to accommodate narrowband filters and still be automated without a filter wheel. I have also done slewing, solving and focusing with (3nm Astrodon) narrowband filters, but not with SGP (and not with a DSLR). It just takes (as you said) long focus and pointing exposures. This means it is not exactly efficient, of course, but if one is willing to accept that, then it will work assuming one can do the required length exposures.

I don’t care about automating the filters, as I use the same filter night after night after night. I’ve plate solved successfully using Astrotortilla in the past with the H-Alpha filter; I just like using SGPro because it means I can set an object to start well after my bedtime without having to stay up well after my bedtime :slight_smile:

To Jared - I tried entering 120s for my focusing time and when I saved the changes it reverted to 100s. I tried this a couple of times.

I didn’t have a chance to check the auto focus time but it sounds like it’s an issue. We’ll resolve that.

You can use astro tortilla (or more correctly with SGP. You just need to install ANSVR and have SGP point to ANSVR rather than So if you’re having success with this might be a work around for plate solving for you. There is more info on ANSVR here:

Also what plate solver are you using? Elbrus or PinPoint?



I’ve been using Cygwin this entire time. I used to use Astrotortilla by itself back when I was using BackyardEOS to do my image capture. Now that I do automation, I use Cygwin with ANSVR. I will admit, however, that the Cygwin/ANSVR route does have some major limitations that I feel hurt processing. The biggest one is that you can’t select any of the custom options other than downscaling. In AstroTortilla, I use the options “-r” and “–objs 100”. The “-r” tells the plate solver to sort the stars to use by brightest first, generally speeding up plate solving, and the “-objs 100” tells it to use the brightest 100 objects. It also allows a “–sigma x” option which allows the plate solving engine to ignore some of the noise and x is the variable.



I’d be glad to add additional custom options to the ansvr config page. For now though, since those options are not there, you can open ansvr in a text editor and add the options you want to the SOLVER_EXTRA_ARGS definition on line 71. For example, change this:


to this:

my $SOLVER_EXTRA_ARGS = "-r --objs 100 --sigma x";

In addition to -r, --objs, and --sigma, are there other options you think we should add?


Since those are the only additional options that I use, I don’t know of any other that should be added.

I just ran a couple of test plate solves. Before adding the extra arguments, the solve failed every time. After adding the extra arguments, however, the solve succeeds every time. This has definitely made a positive difference!

It’s also worth noting that, being able to adjust the sigma value, I now don’t have it downsample at all and it gets many more sources. I’ve also been able to reduce my plate solve exposure time to 30 seconds with success. I certainly hope this continues - for now it is just taking junk frames to keep the autofocus going, but at 11:45 it should stop and head for the Eastern Veil nebula.

I will post results tomorrow morning and let you know how it worked.

Scott, I just uploaded a new version of ansvr, ansvr-0.12. This version has options on the ansvr config web page corresponding to the solve-field -r, --objs, and --sigma command-line options.

I posted a new topic in the Plate Solving category announcing version 0.12 and providing download/installation instructions.



Well, the changes to the Ansvr configuration definitely helped. Almost all things plate solve succesfully - and much more quickly - with a 30-second exposure. I still had a little trouble but when I increased the time to 60 minutes even the difficult areas began to plate solve. Looks like it’s finally working!

The only remaining difficulty is tracking during the exposures. If it’s possible to add a feature to have it run the autoguider during plate solve frames, that would be extremely helpful so I don’t have to babysit the system during plate solving.


We can take a look at adding that option. This will likely add a good deal of time to the plate solve task though…just FYI.


I understand. However, given the two options:

(1) Wake up and sit outside for a while, trying multiple times to get a good plate solve when the mount keeps suffering from bad tracking, losing sleep

(2) Sleep right through, knowing the autoguiding will add some time yet make it so it plate solves the first time.

Not too difficult of a decision to make, in my opinion.

I can certainly understand that…just setting expectations.