GraXpert 3.0 DENOISE - Astrophotography Processing made in Germany
ฝัง
- เผยแพร่เมื่อ 6 ต.ค. 2024
- GraXpert 3.0 DENOISE was just released - I show you how to install it, how to use it natively and from PixInsight with a new process and how good it is compared to NoiseXTerminator.
Link to install GraXpert: www.graxpert.com
Link to add into the PixInsight Repository: pixinsight.dee...
Join my Patreon site for getting cutting edge news about Astrophotography software and equipment, early access without commercials and tons of supporting documents: www.patreon.co...
If you buy any equipment you may consider these three shops - by using the links below you support the channel:
Agena Astro: agenaastro.com...
High Point Scientific: www.highpoints...
Astroshop.eu: www.astroshop....
#astrophotography #pixinsight
------------------------------
Music credits:
ORBITAL_StriKe by B E T T O G H | / bettogh
bettogh.bandca... | open.spotify.c....
Music promoted by www.free-stock...
Creative Commons Attribution 3.0 Unported License
creativecommon....
Based on the comments below it look like GraXpert Denoise is running very fast on Intel/AMD chip sets. The slow performance that I observed seems to be an issue of GraXpert working with Apple M chip based computers!
As always, Great work Sasha!
I’m going to take this moment and complement you on the quality and technical detail that you have in your videos. You are truly a step above the rest of the channels that deal with astrophotography. You are clear analytical, and accurate. As an aerospace engineer I appreciate this. Thank you. I find all your videos very helpful with my own work.
Wow, thank you! Very happy to hear that! 😊
TBH the quality of the result is worth the time taken
I'm using Nvidia CUDA - both NoiseXTerminator and new GraXpert Denoise are more less same fast. Around 5 to 10 sec depends on image.
You confirm my suspicion - they seem to have an issue with Apple M chips….
Had a quick look and was fast for me, will test later if it runs ok within PI.. but very nice info and explanation i also i see 2 scripts in util/toolbox as well as the grax process.. in PI
intel 14900k and 10900k
thankyou SO much for doing this; vielen danke!
Tested with AMD GPU and it's equally as fast as NoiseX. Might work better in PI with a mask and a higher setting as it blurs image quite a bit. I find masks useful using NoiseX on certain images as well though.
That is quite interesting that the performance is depending so heavily on the processor.... It might also be that NXT is not as fast on Intel chips than on ARM - on my M2 NXT is really blazing fast and needs about 5 sec or so....
Yesterday I did a short test and had the same experience, it blurs the image a bit more than XT.
In your video run, did you have your machine configured to enable GraXpert to take advantage of GPU acceleration? Note that on Windows, GraXpert doesn't use CUDA. Rather, it uses DirectML and requires a GPU that supports DirectX 12. Most reasonably modern GPUs provide this support. However, DirectML usually must be installed by the user on Windows machines.
As I use a Mac, you don't need to go any of this stuff and usually everything adapted for the new M series Macs runs blazing fast. As GraXpert presently runs extremely slow on the Mac and all Windows users write me that it is as fast as NXT on their machines, it clearly means that the Denoise routine of Graxpert is not coded appropriately for the Mac presently
Great video, but the denoising in GraXpert is far faster than NoiseX in Affinity. Maybe NoiseX is faster in PI though.
It's not about the application - it's about the processor. As it looks like GraXpert seems to have an issue with Mac M chips and hence it is so slowly on my machine. But all Intel/AMD users tell me it is very fast.
Looking forward to trying this; thank you for the update 👍