It would be incredible if you could “deblur” star movement in astrophotography and make the stars circular.
Man, how amazing would that be, that would make Topaz leaps ahead
Hello Team,
As an avid amateur astrophotographer, I frequently use your software to enhance my celestial images, including moon landscapes, solar captures, planetary shots, and deep-sky photographs. Your software has been instrumental in my hobby, helping me refine and improve the quality of my images. However, I’ve noticed that a one-size-fits-all approach to sharpening and denoising doesn’t always yield the best results across the different types of astrophotography subjects.
To this end, I’d like to suggest the introduction of dedicated settings or profiles within your software specifically tailored for various astrophotography categories. For instance, separate optimization features for lunar surfaces, solar images, planetary details, and deep-sky objects could significantly enhance the user experience and outcomes for astrophotography enthusiasts like myself. Such a feature would not only improve the software’s versatility but also empower users to achieve more precise and stunning astronomical imagery.
Thank you for considering this suggestion. I am looking forward to future updates and am excited about the potential of enhancing my astrophotography with your software.
1 Like
I’ve not had any luck with lunar images with it; it always blows out the highlights, even with Autopilot turned off. The resulting DNG ends up severely clipped and unusable vs the original NEF.
1 Like
Thanks for the idea and suggestion! Make sure that you do not forget to vote for your idea 
3 Likes
I’ve tried to use Photo AI as well with my astrophotography without much success. I feel like this is because so much time is spent on specific subject photographic sources, where it seems to shine. For astrophotography, the AI training would be very different to get the best results. I could see there being almost an entirely different product just for astrophotography. Would be even better if they could learn from existing stacking software out there to create a Topaz AstroPhoto AI product, that could do it all, incorporating AI during the stacking process.
2 Likes
Hi All,
I have been playing with Astrophotography of late. I have a good mount but still only using my DSLR and various lenses.
While I LOVE Topaz in so many ways, I feel that in this space, the AI tends to create things that are not in the data when it comes to Deep Sky Objects like nebula.
Is there any plans to make de-noise solution specifically for this space?
Either way, it wont prevent me from using Topaz for all the other shooting I do because of the stunning job it does. I have even been picking out old images and re-shopping them just so I can give them the Topaz treatment.
For enhancing size try hifi only without all the options, if you have noisy though you are SOL because the algo considers dot detail to be noise, even though they are stars. They really really need to train the other algorithm on natural patterns, as it easily wipes things out or hallucinates.
1 Like
I am not much of an Astro guy but I did take some photos of a lunar eclipse a few months ago and found interesting differences in results based on the cameras. I think my Panasonic G7 seemed to produce better results than my Nikon Z6II that was for stars at least. For the moon, both were not great with lots of made-up craters etc. The moon I would have thought should be something extremely easy to even just fake like Samsung does with their mobile photos do as there is only 1 moon that you need to train on.
1 Like
I agree. Actually the newer models in Topaz Photo AI are actually WORSE for astrophotography.
1 Like
Photo AI isn’t necessarily trained for astrophotography but it looks like this is becoming an increasingly popular use case for the app so we’re definitely going to be adjusting our models accordingly.
1 Like
Hey Tim, that’s awesome news.
The best part is that when you do your modelling, you can compare raw data with images taken from images taken with incredibly powerful telescopes to ensure the model works properly.
I would be happy to help if possible and even share some of my data.
Astrophotography seems to be a blossoming field photography, especially at the pro/am line.
I speak, critique and judge at photography groups and I have seen many amateurs who are not geeks enjoying this field.
1 Like
A gentle push to see if there are any updates on plans to improve Photo AI for astrophotography. I have both the ‘old’ DeNoise AI (V3.7.2) and the newer Photo AI (1.3.12) for macOS.
I get better results when using DeNoise AI compared to Photo AI: much ‘cleaner’ and less artifacts. Can this model not be ported to Photo AI?
So, I am prompted for upgrading to Photo AI V2.4.0. (something I would only consider if it meant improvements for astrophotography as it comes at a cost).
There is a large community of photographers who focus on astrophotography and would be thrilled to be able to process .FIT files in Topaz Photo. Any chance of adding that functionality in the near future?
I’ve gone ahead and moved this thread to our Ideas topic where your input can be reviewed by the appropriate development team.
Make sure that you do not forget to vote for your idea
Since I’m currently involved in astrophotography, I noticed that the PHAI denoisemodels may contain astrophotos for training.
I find this as such critical since these naturally flow into the final result and thus falsify it, because the photographic genre as such are very different among themselves and have different image processing goals.
Also, it could be that in the training data are images from old
cameras that had a problematic noise.
Problematic in the sense that these were often simply cut off in the shadows when editing.
For example, it was known in the past that Canon cameras before 2016 were significantly worse in terms of dynamic range than Sony and Nikon cameras, because both of the latter used Sony sensors, which were superior at the time.
So the training data from these “problematic” sensors would have to be treated separately.
It would make sense to have four models that are each trained with the specific data.
Astro (Tiff?RAW?)
Normal (RAW Cameras 2017+)
Old (RAW Cameras 2017-)
Processed images (tiff, jpeg etc.)
You forgot film photography, which would need models per physical size standard. 35mm, 120mm, etc.
You can forget about a model for astrophotography, I just started doing that and the last thing I need is something to de-noise.
Maybe that will change but I suspect not now.
I have a suggestion for an extra topic inside Sharpening. Why don’t you add a button for Astro Blur? That could help to get rid of the star motion and helps to improve astro photographi when you don’t have a tool to follow the track of the stars.
Thanks for sharing this. Since astrophotography is a small subset of our users, I don’t think this can be prioritized at the moment. However, we’re seeing growth in this type of editing done with Photo AI so it could be something we explore in the future.
I can see how this is slightly different from just simply removing noise because sometimes the stars get confused for noise and vice versa.
