Adding H-apha to OSC data?

[ChrisV]

I'm currently using an OSC camera which I'm happy with (ASI071). I have no intention to go mono/filters for a while yet.

But I do have an H-alpha filter lying around. Would it be worthwhile getting some H-alpha shots with this camera - for targets with strong H-alpha. Then combining the H-alpha with the OSC shots, or maybe splitting into the OSC shots into RGB channels and replacing the R-channel with the H-alpha?

[billdan]

Hi Chris,

I have never tried it, so I'm only speculating here.
If the object has strong Ha emission your OSC will get it all anyway. Like Eta Carina or Rosette.

The Ha filter would be useful where there is weak Ha emission and is swamped by the broadband signal. Like in galaxies etc.

From what I have read you could also image with the Ha filter under a bright moon.

Cheers
Bill

[glend]

Chris it can have some benefit. I used to shoot Ha with my Canon 450D - however it was modded to be more Ha sensitive. Any OSC could probably do it, but you may need longer subs for sure. You won't be sky limited with an Ha filter on your camera, so 300s subs could be useful, assuming you have good guiding. Do your normal OSC subs, and then put on the Ha filter and shoot something like 12 x 300s subs. Then stack your OSC subs in DSS or whatever you use, and do the same for the Ha subs. That will give you two registered and stacked Tiff files. You will then need to register these two files, to align them for combining. DSS can do it easily, just put those two files in as lights, nothing else needed. In the parametres you want to select "register akready registered files" this will align these two (OSC and Ha) files against each other. In DSS intermeadite files, you can tick save a .reg file, if you want, but the input files will be registered against each other anyway - its just a way to avoid naming confusion.
Then layer your OSC and Ha registered files, and colour Ha as you please. When Ha is layered with RGB layers, i always make it a slightly different shade when i colourise it, this insures that Ha nodules, in targets like M83 standout from the typical RGB image. You can also try boosting the saturation or brightness. Hope that helps.

[Atmos]

This is a quick comparison that I've done. The first is a straight RGB (not an astro RGB or modded DSLR) and the second is a HaRGB.

I've been doing a lot of playing around with the best ways of blending it all together. So far the best I've found in PixInsight:
- LinearFit all the frames together against the channel with the lowest noise. In my case, this was the Ha.
- Using pixel math I did a Max(R,Ha) for the red channel and standard for the G and B.

Looking at these comparisons you can see that the nebulosity has lost some of it's pink hue and has gone mostly red. I am still looking for a good way of blending some of the Ha with the blue channel to simulate the hydrogen beta emission.

[ChrisV]

Thanks for the tips everyone, especially on how to combine the H-alpha & colour data - I wasn't even thinking about that stage yet!

I was also concerned that the GB might be relatively noisy so adding (what I presume would be) lower noise H-alpha might be pointless - and I'd be better off just getting more OSC data. But Colin's pics says a thousand words. So there's my project for the next few months.

[multiweb]

I used to debayermy QHY8 to Red only for Ha, then combine green and blue for Oiii.
The mix with RGB was to put 50% Ha in the Red channel and ~15% Ha in the blue channel. Seemed to work for most parts.

[ChrisV]

I tried the H-alpha filter on the ASI071 OSC camera the other night. Wow the background is low - with 5 - 10min subs !

Is it worthwhile using a higher gain for the H-alpha shots so that its easier to swamp the lower read noise at the higher gain?
I've been shooting at unity gain which has read noise 2.6e-. At higher gain RN is ~ 2.3e-. Maybe that's not much better?

[Atmos]

There isn’t much of a difference between 2.6 and 2.3 Although it is a 25% decrease in needed exposure time so it could be worthwhile giving it a go.

You also need to remember that only 1/4 of the pixels are detecting the Ha emission.

[ChrisV]

Thanks colin. Yes that's not a huge decrease in RN. And 25% of the pixels for Ha sucks. Maybe it's all telling me to go mono ...

[RickS]

Come over to the grey side :lol:

NB does require long subs. Unless you have a very fast system 1800 second subs are good if you can manage them.