Multiscale Adaptive Stretch MAS in PixInsight

From other tests of MAS verses VLS I’ve seen it looks like they both get you to very much the same place with VLS leaving behind a slightly nosier result.

Rainer Ehlert · Jan 1, 2026, 06:49 PM

Hi,

Two days ago I heard about MultiscaleAdaptiveStretch

I have been playing with it and as an advanced beginner I like it. It saves me a lot of pre-stretching time with HistogramTransformation (Sorry, Never understood GHS 😵‍💫?

I also tested VeraLux… but with MAS I got a result that I liked faster than with VL…

Below an example of one of my wide field of HT and MAS. Sure I need to refine the settings but as a first result for me it looks quite good. What I like is the possibility of defining the background value and so you always get the same? background value. Also the ability of stretching the single LRGB Ha OII and SII images is great as you get the same background value of all. Which was nearly quite impossible using HT. IMHO

What do you think about MAS? if you have tested it.

HT stretching and fiddling around + NXT

📷 image.png

MAS stretched and fiddling around + NXT

📷 image.png

For me, Veralux ‘paints’ the colour, making it very pleasing to the eye, but it does not respect the highlights, saturating the bright cores too much, such as the cores of M31 and M42, to give two well-known examples. I always use Masked Stretch for this reason, it stretches the image very well and protects the highlights excellently, however it ‘washes’ the colour too much for my taste, so I have found a way to combine the stretching of Veralux with the unbeatable highlight protection offered by Masked Stretch. I am working on a script that is a mixture of both, good colour and lighting, and proper protection of the highlights. Veralux does not use masks, so even if you add a mask to the unstretched image, Veralux does not take it into account, and that is what I have solved. As for MAS, it is very similar to Veralux, but it respects colour and contrast much more, although sometimes that is a matter of taste. I think we are in a competition of stretchers, from which only good things can come, as competition always produces benefits for everyone.

Ferran Bosch (S.A.C.) · Jan 2, 2026, 09:27 AM

not respect the highlights, saturating the bright cores too much

If you use the veralux from the Siril application, cores will not be saturated.

Robert Houghton · Jan 4, 2026, 06:07 AM

I’ve looked all over and cant find it . It says its installed .

Intensity Transf.

Cyril Richard · Jan 4, 2026, 07:03 AM

Ferran Bosch (S.A.C.) · Jan 2, 2026, 09:27 AM

not respect the highlights, saturating the bright cores too much

If you use the veralux from the Siril application, cores will not be saturated.

And why does it saturate in PixInsight?

I have played with MAS a number of times now and I have to say it is very impressive. Of course you can arrive at the same result, or close, with STF , GHS and Curves adjustments but this gets you there in one go. Depending on what you’re after in terms of a result some additional work in Curves or GHS will be needed but much of the early careful time spent in GHS and Curves is now not needed.

What is needed is a new workflow. What I have been doing is BXT(correct only)>SPCC>SPFC>MSGC>MAS. (Of course in order to do the early steps it is helpful/necessary to use STF) As I have said before, I think PixInsight is probably the best money I ever spent in astrophotography. In terms of what it has done for my end result and how straight up professional they are about updates and improved functionality it isn’t even close.

Arun H · Jan 4, 2026, 08:42 AM

For those who have used it - is it a game changer?

What is really a game changer?

I tested it and now I am using it and it saves me a lot of HT, HT, HT, … and I get to a better result faster which then I can tweak finer, maybe maybe not. I made some images and I did not even find it necessary to do any tweaking.

It is a question of R&D or Trial & Error…

¿game changer? I do not know but a Time Saver, Yes for me

George Hatfield · Jan 4, 2026 at 05:40 PM

I have yet to see anyone using MultiscalAdaptiveStretch for starless images. My routine is to always work on the stars separately, which gives me good control. For the starless image, I either use SetiAstro’s statistical stretch and/or GHS. So the new tool does not seem to be that useful, at least for me. Am I missing something?

I’m still using SA’s statistical stretch too, however; I did see that Frank has updated stat stretch for SAS, but will not be updating the PI script. I hope the PI script isn’t discontinued, but as a precaution, I will start experimenting with MAS.

George Hatfield · Jan 4, 2026, 05:40 PM

I have yet to see anyone using MultiscalAdaptiveStretch for starless images. My routine is to always work on the stars separately, which gives me good control. For the starless image, I either use SetiAstro’s statistical stretch and/or GHS. So the new tool does not seem to be that useful, at least for me. Am I missing something?

You are not missing anything IMHO

If what you do works for you nobody is going to force you to use a tool which does not fit into your workflow

It is just that easy 😉

Rainer Ehlert · Jan 4, 2026, 06:52 PM

Ferran Bosch (S.A.C.) · Jan 4, 2026, 03:07 PM

GiffS · Jan 4, 2026, 02:37 PM

I have played with MAS a number of times now and I have to say it is very impressive. Of course you can arrive at the same result, or close, with STF , GHS and Curves adjustments but this gets you there in one go. Depending on what you’re after in terms of a result some additional work in Curves or GHS will be needed but much of the early careful time spent in GHS and Curves is now not needed.

What is needed is a new workflow. What I have been doing is BXT(correct only)>SPCC>SPFC>MSGC>MAS. (Of course in order to do the early steps it is helpful/necessary to use STF) As I have said before, I think PixInsight is probably the best money I ever spent in astrophotography. In terms of what it has done for my end result and how straight up professional they are about updates and improved functionality it isn’t even close.

Why do you calibrate the colour first instead of the flow? Can you explain why?

.

¿¿¿¿

I wrote that maybe that could have been a typo. First I need to check if it is possible to SPCC before SPFC…

¿OK?

… and I have not yet found how to delete an answer. Looks like it is not possible ¿?

If I remember correctly, I believe SPFC is needed before using MSGC. I’m not sure that using SPFC before (or after) SPCC does anything useful. But since I’m still on the middle slopes of my personal processing learning curve, I could be entirely wrong about this….