RC Astro tools coming to Siril

Yes! I’ve been waiting for this (GPU acceleration with AMD) since march 25…awesome! Thanks Tony, I checked the news on Russell's Page yesterday and drew a blank!

1day 13hrs to go - counting down!

Tony Gondola · Jun 25, 2026 at 03:40 PM

I find the whole thing very interesting. BX has been the sole province of PI for so long and to be honest, a major reason for upgrading to PI at all…

Haven’t seen this discussed yet, but has anyone else tried the newest native PI noise reduction process, MLdenoise? It’s a machine learning version of noise removal that mimics NXT. Looking at the announcement it seems PI is going full machine learning to compete with the add-ons many people are using: BXT, NXT, and SXT.

I’ve tried MLdenoise. In my limited tests, my take is it’s much slower than NXT, while producing essentially the same result as NXT. My take is that PI is reacting to the popularity of the xXTs and as a defensive move creating their own version of this native to PI. Just curious what other’s think of this…

That is very nice to hear, thanks for sharing. Surprising to be honest as its been a while since BXT launched as a PI only tool.

It will be interesting to see how PI will react now that a significant PI only tool is available elsewhere. Especially considering that Siril is free (and excellent) and PI is quite expensive.

Tom Boyd · Jun 25, 2026, 05:58 PM

Oskari Nikkinen · Jun 25, 2026 at 05:24 PM

It will be interesting to see how PI will react now that a significant PI only tool is available elsewhere. Especially considering that Siril is free (and excellent) and PI is quite expensive.

They’re building their own native versions:

https://pixinsight.net/dev/index.php?ams/technology-preview-mldenoise-for-macos-arm64.19/

Good, competition driving development, just what I hoped to see.

Bill McLaughlin · Jun 25, 2026, 10:09 PM

My take is that there is a reason that there are so many plugins for PI and that is because there has been so little innovation in the native app. If it were not for the plugins, PI would have already gone the way of the DoDo. It is only their plugin architecture that has saved them.

I am still a big PI user and have been since very early days but they had better get moving toward the future or they will go the way of so many others that are now just history like CCD Stack, MaxImDL, and several others.

Maybe I am wrong but I think they may have had an “overly scientific” outlook and failed to truly embrace the fact that the vast majority of their users were not doing science.

I think you make some very good points. Many times I have had the experience of finding a process that I want to use only to find I have absolutely no idea about what the various sliders do. Then you click in the documentation icon and there’s nothing there. The tool tips will tell you what something does but not always in way that’s understandable.

That said, I really love PI as a workspace. In that regard it’s a thing of beauty, quick and effcient. I may stack in Siril and color correct in SAS but everything feeds into the PI workspace eventually.

It depends, if he was offering value for money, meaning things that I can’t accomplish elsewhere or not as easily elsewhere and if he did it at a fair price then it might be fine. It really all depends. Right now for various reasons I can get better results using PI (in conjunction with other tools) than without. If it gets to the point where that is not true then things need to be reassessed.

The idea that I can try, in free software, then buy these tools and continue using in free software, or, buy PI and continue to use them, is very attractive to me as a newbie to this hobby.

I hope Mr RC gets plenty of new sales for doing this :)

Bill McLaughlin · Jun 25, 2026, 10:09 PM

My take is that there is a reason that there are so many plugins for PI and that is because there has been so little innovation in the native app. If it were not for the plugins, PI would have already gone the way of the DoDo. It is only their plugin architecture that has saved them.

I am still a big PI user and have been since very early days but they had better get moving toward the future or they will go the way of so many others that are now just history like CCD Stack, MaxImDL, and several others.

Maybe I am wrong but I think they may have had an “overly scientific” outlook and failed to truly embrace the fact that the vast majority of their users were not doing science.

idk about that - there has been a lot of development in PI over the past 10 years. you might argue that some or a lot of it is “just keeping up with the competition” but nonetheless there have been a lot of new non-3rd party tools. i don’t know what other tools offer and whether PI was first on any or all of these, but in the last 10 years off the top of my head (and not in order)

BPP

WBPP

GradientCorrection

MultiscaleGradientCorrection (and the MARS project)

PSF-based subexposure analysis throughout the pipeline

PCC

SPCC

SPFC

MultiscaleAdaptiveStretch

i’m sure there are more…

obviously there have been other gradient correction tools like GraxPert that predated GC and MGC, and i think some form of SPCC is available in Siril? but you can’t really pretend that none of these things happened. MGC/MARS is a huge undertaking and is unique as far as i can tell. that alone is hugely important for people like me shooting from heavy light pollution.

IMO croman’s tools were just so useful that they seem to be the main innovation in pixinsight. but there’s a lot of other stuff that while not as “sexy” have really improved PI over the years.

the “scientific” mindset is baked right into PI, juan is never going to abandon that.

Tony Gondola · Jun 29, 2026, 03:03 AM

Yes but other applications have been evolving as well. It’s clear you haven’t used it but Siril has had SPCC for a long time now and Seti Astro Pro has the even more advanced SSSC. MARS has potential but in the images I’ve tried so far, it’s not a 100% solution.

The main point isn’t about PI keeping up with the standard solutions, it has, but rather about things that are very powerful and only available on the platform. That’s what makes the recent change with RC-Astro so important.

i hear you, i was just reacting to the statement that there had been no development or innovation in the PI core app and i don’t see it that way. for instance, i think PI was first out of the gate with spectrographic color calibration (2022 as far as i can tell). other programs may have caught up, but how is that not innovation on PI’s part? people have short memories, i guess.

at any rate i don’t know why russ’s tools were PI-exclusive at all or for so long, and it may very well affect PI sales if these tools are available as standalone programs. IMO PI is going to end when juan stops working on it anyway, so the future probably does belong to other tools.

I am in the trial version of RC noiseXT and in the Get Scripts section of Siril 1.4.2, there is a script for it already. It worked seamlessly on my images of the Crescent and Butterfly nebulae. I compared the effect of GraXpert to RC noiseXT side by side and RC wins, although GraXpert did okay considering that it is free.

Ruben Kier · Jul 5, 2026, 09:51 AM

I compared the effect of GraXpert to RC noiseXT side by side and RC wins, although GraXpert did okay considering that it is free

graxpert does noise reduction now?

Tony Gondola · Jul 5, 2026, 03:36 PM

Where have you been, it’s had it for ages.

well admittedly i have not had to use it for ages thanks to MGC and GC, but maybe 2 years ago i had it installed in PI. really had no idea!