›

Wonderful star reduction method, but stars become too flat

Vivian Budnik
in PixInsight
Hello Adam,
I keep coming back to your lessons, which are an awesome resource. One problem I am finding with your star reduction methods, is that stars become mostly "flat". In general beautiful stars, even small ones have a bit of glow around them that gives them volume. This becomes lost with star reduction. I have tried to apply different gradients to the edges, with the hope that this would resolve this problem, but I have not found something convincing enough. I even see this in your own spectacular images, so I assume that perhaps it is too difficult  to accomplish. It kind of needs "a local mapping" of the glow around each star like in your contour  masks, but catching a little of the star glow, but then this would have to be shrunk in a way that I cannot figure out. Of course this would need to be done selectively for different star sizes.

Any tip or suggestion would be much appreciated. There is somebody in astrobin that achieves this terrific volume effect, so it can be done ( I can send the link if needed). I just don't know how

all the best,


Vivian

Comments

  • Vivian Budnik
    I meant: "I jut don't know how to do this (ie, give volume to stars)"
  • BlockHeadBlockHead
    Yes, both by substitution of pixels as I show or by erosion (Morphological Selection) this will happen. Actually Morphological selection will leave little halos more than my technique.It was actually the main reason I disliked MS- because you can always tell someone used it.

    WIth my technique, you can always make the halo mask darker by changing its brightness or blurring it more. This will keep your original glows. If the concern is that the glow is not maintained on the larger stars ...but the smaller ones are OK- you should include the larger stars in the "big star" mask and subtract them so that are not part of the halo mask.

    -the Blockhead 
  • BlockHeadBlockHead
    Sure...send the example link.
  • Vivian Budnik
    Thank you, Adam. Here is an example of what I mean. Both the large and the really tiny stars in this image have a beautiful 3D like shape....Of course, he might not be doing any kind of substitution. I will try your original pixel substitution some more.

  • Vivian Budnik
    another: https://www.astrobin.com/full/4q5m9i/0/?mod=&real=

    look at that little star cluster (NGC2100). You can define each star through its volume. I don't know how to describe.
  • BlockHeadBlockHead
    I am not certain what process he is using... he says:

    "found I got more of the blue in the APP version and the more glowy APP stars I use to take the "edge" off the blown cores of stars coming from my Startools outputs.."

    So it sounds like he is getting "flat" stars from normal processing- and then doing something to make them "glowy" which is what you like. 

    The other image is similar. Both of the examples you choose are narrowband- which naturally have "de-emphasized" stars since stars are not as bright (compared to the nebula) in narrowband images. 

    So I see their results- but not yours. I think you need to post your example and circle the issue.

    My guess, without looking, is that you would probably like the results of maskedstretch on your images.

    -the Blockhead
  • Vivian Budnik
    Ahaa! I did not notice that comment of his. He uses the Star tools program, which perhaps has a way to make this effect "happen"by some knowable or unknowable magic. I don't know that program. What I really like about PI is that one can know (more or less) what you are doing with the image.

    Here is one of my images. I was pleased with it, except for flat stars: https://www.astrobin.com/full/48ryub/B/?mod=&real=. I used the Morphological selection method.

    But you have another good point: it is to a great extent about the esthetics one likes..

    Thank you, Adam.
  • BlockHeadBlockHead
    Yeah... I think in your case, having looked at the image, part of the issue is likely how you are stretching the image. If you want "softer" stars that maintain a radial brightness profile you will want to employ multiple midtone stretches (gamma) using CurvesTransformation or the midtones adjustment of HistogramTransformation. As I mentioned, MaskedStretch can also be employed...but this might be too much.

    -the BLockhead
  • Vivian Budnik
    Thank you for the suggestion. I will play with this.

    Vivian
Sign In or Register to comment.