StarTools

Assuming I have a processed RGB stack and a processed HA stack that is based on the extracted red channel taken with a ha filter and the same camera how should I combine the two in startools?

That's a good question with no one good answer

IMO, the best method, which will make sure you retain RGB coloring and will give you good luminance detail, is to create a synthetic luminance from your R, G, B and Ha, process that, and combine that with a lightly processed RGB.

You can have a look here for a tutorial that goes into the creation of a Synthetic luminance frame.

Combining RGB and Synthetic luminance can alternatively done like so (I prefer that technique myself).

Hope this helps!

admin wrote:That's a good question with no one good answer

IMO, the best method, which will make sure you retain RGB coloring and will give you good luminance detail, is to create a synthetic luminance from your R, G, B and Ha, process that, and combine that with a lightly processed RGB.

You can have a look here for a tutorial that goes into the creation of a Synthetic luminance frame.

Combining RGB and Synthetic luminance can alternatively done like so (I prefer that technique myself).

Hope this helps!

Kind of. The problem is that solution will force me to give up resolution so I would prefer an option where I get to keep the interpolated RGB resolution.

midwayexpress wrote:Kind of. The problem is that solution will force me to give up resolution so I would prefer an option where I get to keep the interpolated RGB resolution.

Hmmm... Could you elaborate on why this gives up resolution?

Could you provide some acquisition details (number of subs, exposure times and bin1x1 or bin2x2) for each channel/band?

Thanks!

I'm using a one shot color color camera so I don't get a separate R/G/B channels unless I extract them separately. When I do that I get an image of the size that is effectively 2x2 bin. So for the image in question I have 2.5 hours of RGB data and 3 hrs of Ha data. As I took the ha data with a OSC color camera and extracted the red channel it will have to be resampled to the full RGB frame size.

midwayexpress wrote:I'm using a one shot color color camera so I don't get a separate R/G/B channels unless I extract them separately. When I do that I get an image of the size that is effectively 2x2 bin.

Ah. That is true, but is true regardless of the method you'd want to employ (and is still true for the R and B channels converted to luminance, while the G channel ends up being a 1.41x1.41 bin for the luminance).

The debayering matrix is always something to mindful of, however, you will have shot multiple subs (I imagine), causing a slight drift, which will have bayer drizzled the subject (especially if you dithered); chances are that at least one real sample for every red pixel is available.

Your stacker will (hopefully) have a bayer drizzle option (DSS does and so does PixInsight). This should allow you to obtain a full resolution Ha data set (with the caveat that it will be bit noisier).

You could examine your data first and see whether the reduced resolution is even a problem; if it turns out detail is 'smeared out' over multiple pixels anyway (due to atmospheric conditions or due to optics limitations), then maintaining native resolution is a bit of a waste; it would be put to much better use by binning it (and this reducing the noise in your image) so you can push the data harder.