How far do you stretch?

[MikeWiles]

As I continue to work on getting better at the processing half of the equation....my biggest "struggle" these days is simply figuring out when to quit with the image stretching.  As a simple solution I have lately resorted to simply dragging the STF function onto the Histogram Transformation and applying it for an anxiety free stretch. 

I've also been using the Masked Stretch script...but I'm unsure where to set the target median and it seems to rob the images of contrast when I use it. 

So my question is...how do you do a non-linear stretch in your image?  And...how do you determine the point where you've stretched "enough" without stretching too much??

Thanks,

Mike

[Philip de Louraille]

It is a combination of noise (small details) in the background and a "washed out" image (large details) for me.

[sreilly]

The longer image, read more total imaging time, gives you a higher signal/noise ratio which allows a more aggressive process. A loss of contrast can be remedied using curves after the HST.

-Steve

[marekc]

Hi Mike,

I don't really have an answer, but I think you've asked a really good question. I, too, would love to have some sort of guideline for how far to push the non-linear stretch. As has been pointed out, a high-SNR image can be pushed farther. It would be nice, though, to have some quantitative way of applying a `strong, but not-too-extreme' stretch, based on something like the SNR of one's background sky.

For right now, I do something like this:

1) Deconvolution on the bright (i.e. high-SNR) part(s) of my object.

2) MMT denoising on the dim (i.e. low-SNR) part(s) of my object.

3) nonlinear stretch, to a point where I don't seem to be bringing out the remaining background noise (that I was unable to totally eliminate in step 2).

- Marek

[sreilly]

There are other considerations as well. Depending on the object, you might want a really aggressive stretch to pull out faint background details or not so much if you like darker background. I posted a new image yesterday that, for me needed a good deal stretching more to bring out the dark nebula component that my normal stretching would have only hinted at. See http://www.astral-imaging.com/NGC1333-LRGB.htm as an example. I also did the same for this image, again posted yesterday, to show more dark nebula that I didn't know existed. See http://www.astral-imaging.com/vdB-142-LRGB.htm

-Steve