I have been away from the scope for some time - so no new images but I have from time to time gone back to this image to try improve it. This is my final effort, until next year.
For full info and higher resolution, visit my web page:
3 Likes
Nice one Kinch! Very dramatic. I wish I could get stars nice and sharp like that.
Very nice blending of NB filter data.
cm
Thanks guys - glad to hear your comments.
@buzz: Initially stars were really oblongā¦it took a lot of work to āpull them inā. I have seen many images where people make a final NB image without starsā¦but it really is worth doing the work to keep them in and if possible, better still, put them in, in RGBā¦which unfortunately I cannot do for this image. I do however always try to reduce the influence of the stars in a NB imageā¦we are of course more interested in the nebula.
@chasmiller46 : Interesting pointā¦the blending. I love working NB in PixInsight and especially the possibilities in Pixel Math for blending the SHO. Normally I try this that & the other in the blend and at the end of the session run with one that pleases meā¦and end up having no idea how I blended. On this occasion I came up with a good idea; I actually did a screen shot of how the image basics were blended. I share it here for anyone interestedā¦
Very interesting. The PI guys have a couple of mixing formulae (besides straight SHO) that I always experiment with. One of the problems is the same formula will produce very different results depending on the relative strengths of the three channels, so one size does not fit all.
I have not tried one similar to what you are showing. It never occurred to me to add in the Ha data as a synthetic luminescence channel. Iām not sure I understand what that gets you but it is something I will need to try.
Charlie
imo instead of replacing the L of the SHO image with Ha only, you should make a Lum master by integrating your S, H and O masters together, and use that for the L in LRGBCombination. youāre muting all kinds of stuff by using only the Ha as the L.
rob
@pfile A great idea Rob, that I have never done. I will definitely give that a try. In general my Ha is much cleaner & sharper than SII or OIII, so for me there is always a definite improvement in the overall image quality by adding it in as Lum (as long as it is not overdone). However, I take your point that a lot of data is getting mutedā¦so looking forward to trying as you suggested.
@chasmiller46 The beauty of PixelMath is that you can play around with the mixing all day long until you find something that pleases you. For me, there is no right end result in NB imagesā¦I just aim to get what appears to make an interesting image showing the nebula as best I can. @pfile has now given me something else to try.
@Kinch When Iām doing NB, I take a few hours of RGB too and put in the realistic star color. Your stars are very crisp (but white) - mine are colorful but more diffuse. I think some of it is down to the difference between a RCT and a high quality refractor but it seems that the stretching process in PI, regardless of the precise tool or method (or use of deconvolution beforehand) makes matters worse in my images. Maybe I need to start thinking of removing the stars from the NB image altogether and add them back in later.
@buzz This conversation is somewhat off SGP topics but I am sure I can get a few more lines in without raising heckles too much.
The filter wheel used for this image only has NB filters - so although I always want RGB for stars I donāt always get to do a run with the RGB filters in. In fact in this instance, for various reasons, I did not even get the amount of NB data that I wantedā¦and in fact, I did not even get any flats done.
On my last image I did manage to get RGB starsā¦the image at the very bottom of this page: NGC 6820 & NGC 6823 - KinchAstro
I am not sure, but in that image the stars are probably not so ātightā as in the NB only imageā¦so perhaps I have āoverdone itā in the Eagle Nebula. (But I never take them out during the processing).