Auto sequencing data collection

Hi Rodd, you can use NINA or CCDCiel amongst others to image sequences automatically and in both pieces of software there is the ability to autofocus (assuming you have an ASCOM of INDI enabled focuser) then in the setup you can define different focus points for each filter. I have a reflector so haven’t come across such large differences in focus but this would be one solution to your problem.

Rob

Rodd,

N.I.N.A. is free and supports relative focus offsets between filters so as to avoid repeated refocusing if you are cycling filters between exposures.

The "Cuiv, The Lazy Geek" Youtube channel offers good N.I.N.A. tutorials (he contributes to the open source code),
https://www.youtube.com/channel/UC65vvpQDX5rymeqrYt-Bb1g

Here is the docs page at N.I.N.A's site that describes autofocus and filter focus offsets,
https://nighttime-imaging.eu/docs/master/site/advanced/autofocus/

You find a time of night when the temperature won't change much over the hour or so while running autofocus on every filter you commonly use and put the focus offset counts into N.I.N.A. You can specify a particular filter to autofocus with and then just apply the learned offsets for the others. Also you can setup per-filter exposure times to use during autofocus runs. Finally, there are ways to tell N.I.N.A. to refocus every so often; on pier flips; if measured focus drifts by x percent, etc.

Most of the time I image at f/4.9 to f/1.4 and don't find that any brand of filter is truly parfocal at fast f-ratios, and some vendors state as such.

Depending on the altitude angle and filter color, relative air mass you are looking through will also disperse blue light more than the red end of the spectrum. For this reason and since I have to change filters manually, and because my skies suffer a brighter greenish Las Vegas light dome to my east, I image first with red, then blue near zenith for minimum dispersion, and then green as the object goes into the darker western sky.

Jim

Rodd,

N.I.N.A. is free and supports relative focus offsets between filters so as to avoid repeated refocusing if you are cycling filters between exposures.

The "Cuiv, The Lazy Geek" Youtube channel offers good N.I.N.A. tutorials (he contributes to the open source code),
https://www.youtube.com/channel/UC65vvpQDX5rymeqrYt-Bb1g

Here is the docs page at N.I.N.A's site that describes autofocus and filter focus offsets,
https://nighttime-imaging.eu/docs/master/site/advanced/autofocus/

You find a time of night when the temperature won't change much over the hour or so while running autofocus on every filter you commonly use and put the focus offset counts into N.I.N.A. You can specify a particular filter to autofocus with and then just apply the learned offsets for the others. Also you can setup per-filter exposure times to use during autofocus runs. Finally, there are ways to tell N.I.N.A. to refocus every so often; on pier flips; if measured focus drifts by x percent, etc.

Most of the time I image at f/4.9 to f/1.4 and don't find that any brand of filter is truly parfocal at fast f-ratios, and some vendors state as such.

Depending on the altitude angle and filter color, relative air mass you are looking through will also disperse blue light more than the red end of the spectrum. For this reason and since I have to change filters manually, and because my skies suffer a brighter greenish Las Vegas light dome to my east, I image first with red, then blue near zenith for minimum dispersion, and then green as the object goes into the darker western sky.

Jim

*the problem is j gong know how to autofocus.  I do want to learn but when?

Bogdan Borz:
Hi Rodd,

You mean you are not actually using autofocus? It is much harder without it... I use Sequence generator pro. Basically, you have to test it before finding the right number of steps and the amplitude of each step. The program will start from the hypothesis that you are starting by being in focus or close to it. Then if you defined a step size of 10 steps per data point and 9 data points it will move the focuser out by 4x10 steps (the other 4 data points are in focus). It measures the star parameter (HFR) for every step by going in (it will gradually decrease) and once it starts to be too much in focus the HFR will increase again. It uses all the data points in the graph to trace a 'best fit' curve which gives the best focus at its apex. The difficulty is finding the good combo of step size and number of steps, it depends on your F ratio. So trial and error at first.

Your FWHM are really different, but right now the focus will shift a lot as the temperatures go down during the night and you have a significant difference in temperature between the start of the session and the end. So FWHM could be increasing even when using only one filter. In order to verify if they are parfocal you should have verified the focus for each RGB filter in the beginning of the session. My SHO Astrodons are not completely parfocal for ex.

Hope this helps. CS,

Bogdan

Well, focus can’t shift due to temperature in 20 sec.  I took 30 20 sec subs of each and the fwhm values of all the red were the same, the green were the same and the blue were the same (within .01-.3 arcsec). So it can’t be temperature. When a red sub is 5 arcsec and 20sec later a green sub is 7.5 arcsec then 20sec later a blue is 11 arcsec, and finally another red 30 sec later is agin 5 arcsec, I think that is proof it is the filters

Bogdan Borz:

Rodd Dryfoos:

Bogdan Borz:
Hi Rodd,

You mean you are not actually using autofocus? It is much harder without it... I use Sequence generator pro. Basically, you have to test it before finding the right number of steps and the amplitude of each step. The program will start from the hypothesis that you are starting by being in focus or close to it. Then if you defined a step size of 10 steps per data point and 9 data points it will move the focuser out by 4x10 steps (the other 4 data points are in focus). It measures the star parameter (HFR) for every step by going in (it will gradually decrease) and once it starts to be too much in focus the HFR will increase again. It uses all the data points in the graph to trace a 'best fit' curve which gives the best focus at its apex. The difficulty is finding the good combo of step size and number of steps, it depends on your F ratio. So trial and error at first.

Your FWHM are really different, but right now the focus will shift a lot as the temperatures go down during the night and you have a significant difference in temperature between the start of the session and the end. So FWHM could be increasing even when using only one filter. In order to verify if they are parfocal you should have verified the focus for each RGB filter in the beginning of the session. My SHO Astrodons are not completely parfocal for ex.

Hope this helps. CS,

Bogdan

Well, focus can’t shift due to temperature in 20 sec.  I took 30 20 sec subs of each and the fwhm values of all the red were the same, the green were the same and the blue were the same (within .01-.3 arcsec). So it can’t be temperature. When a red sub is 5 arcsec and 20sec later a green sub is 7.5 arcsec then 20sec later a blue is 11 arcsec, and finally another red 30 sec later is agin 5 arcsec, I think that is proof it is the filters

Right, I thought you were cyclying through each filter. It's clearly the filters. But parfocal filters does not mean that they will produce the same HFR for each filter, it means that for a given focuser position, each filter will be in focus. If you leave the Bahtinov mask in place after reaching focus for G and then you change for the R and B, they should be in focus. The HFR will be different between filters (blue bloats more than g for ex.).

That’s true.  But when I focus carefully while imaging I get decent fwhm for all filters. Blue is not typically as good, as you say, but usually I can get them within .5 arcsec of red and green. Or 1 on a poor night.   Not double of triple