Quote:
Originally Posted by gregbradley
Impressive small stars and stars in general, plus detail in the neb and the dust lanes. But the Ha blend has gone a bit salmon. This is from using Ha as a luminance probably.
Ha is best added as its own layer in screen or soft light blending mode and as a hue/saturation layer clipping mask (google astrodon screen mode for blending narrowband images).
Even then I find I have to trial and error a lot with various blending modes and saturations/lightnesses.
If you have a nice build up of colour you can get away with using a bit of Ha as luminance layer but usually not more than about 20% opacity where the increased detail is offset with the onset of the salmon colour shift.
Others use a blend of Ha with red in the red channel but you don't have any instant control over that blend. I suppose users must know from experience what amount to blend in to get it right.
Greg.
|
That's funny you mention salmon 'cos when I was processing, that was somewhat what I was aiming for
.
To be frank I really didn't know what I was doing in the processing. I got as far as creating the masters fine (up to the Decon/LinearFit level), but after that I kinda got lost in which path to take - I was trying to follow Chris Foster's PI Workflow but it got a little convoluted for me.
So I tried to build a SuperLum made out from RGB's extracted Lum and Ha combined, then put that new SuperLum back into the RGB. It was probably overcomplicating things. I'll google up your astrodon screen mode when I reprocess this again.
As for the detail for the pix, I'd like to think it being super selective in the subs. This is where I learnt a lot about my rig (despite owning it for over 5 years!). I shot everything in Bin1 with my ASI6200MM even at my scope's native focal length of 2500mm, which gave me a 0.31"/px resolution. I'm also guiding via OAG rather than a guidescope.
Thereafter trying to get my mount to guide as close as possible to that figure or lower was a real PITA - it was really in the end up to seeing conditions. My best was mid March with the Red filter - that one evening I had 0.28"->0.32" for most of the evening with the best at 0.24", and the worst at 0.37". I learnt that with my setup combination, if I got more than 0.45" RMS errors, my HFR goes up and I got pregnant stars - whilst still with low eccentricity and very round, they were pregnant nevertheless. So in the end for most channels I ditched anything from 30% -> 50% of my subs, keeping only those with nice tight stars, and good eccentricity/SNR.
As an aside, I also learnt a lot about my backyard
. Despite having a reasonable wide field of view, my only useful imaging window is something like 30deg on either side of the meridian - maybe 35deg either side at best. Despite having a wider view, whether it's heat plumes coming from rooftops from the east, or Fremantle skyglow on the west, I can see my RMS change (and therefore HFR change) as tracking goes up toward meridian, then flip and back down. So again i have to ditch subs of where the target is 55deg or lower.
I also learnt that I don't like PI's Subframe Selector's FWHM as a measurement in the weight calculation. I wish I could calculate HFR/Ecc/SNRWeight instead of FWHM/Ecc/SNRWeight in the formula. So I manually selected most images (with Subframe Selector as a guide).
Interesting to read about your location, sky conditions, rooftops and where you choose to image in the sky. All these considerations certainly paid off. Imaging at long FL is taxing on the gear and unforgiving.
