- This topic has 6 replies, 5 voices, and was last updated 2 years, 1 month ago by
.
-
Posts
-
I have recently discovered the new deconvolution tool in PixInsight – BlurXterminator by Russell Croman
https://www.rc-astro.com/resources/BlurXTerminator/index.php
It has only been available a couple of months I beleive..
BlurXterminator uses a neural network to do deconvolution rather than the more usual Richardson-Lucy algorithms.
The results are (IMHO) staggering…
For an example,see the attached comparison between my original processing of M51 from last year (on right), and the new one with BlurXterminator (left)
I originally used the normal deconvolution tools in PixInsight and was quite pleased with the result, given the normal expectations for this kind of rig and
data set. (The original image, in full, is my personal gallery)The reworked version has much more detail and clarity. I have compared carefully with an HST image and can see no artefacts (always the risk with deconvoution and wavelet enhancement etc). Indeed the new version compares well with a slightly blurred (5×5 kernel) version of the HST image.
Unlike many tools in PI, this one is simple to use :). I would strongly recommend any PI users to give it a go… (and likewise with Russell Cromans other tool NoiseExterminator – works like a dream…)
I aim to rework a few more images to try and unearth all that latent data!
Attachments:
So, I have just purchased the commercial license for BlurXTerminator. I think it’s a very good piece of software. Much faster deconvolution than Richardson-Lucy, and the results are pretty good. Here is the first image, of M86, I have processed with it:
Maxim
Interesting.
As you say it brings out some nice detail in the dark bands and centre of the galaxy. Was surprised to see that in the top left part of the image some of the detail seemed to be blurred out. Would be interested to see what it did with something like the Rosette.
Yes, Russ Crowmans BlurXterminator does seem to do a good job. I used it for the first time together with RC’s StarXterminator to process my C/2022E3 comet image here https://britastro.org/observations/observation.php?id=20230126_165855_917730cc9af3bb9a I’m also new to Pixinsight so the processing was very time-consuming and I guess could have been done better with more practice.
Hi Peter,
If you are new to Pixinsight then – well done! It’s a bit of a learning curve but well worth it – especially for deep sky.
A good way in is with Warren Kellers book ‘Inside Pixinsight, otherwise info is scattered over the web. The early stages of calibration and stacking are well treated in videos by Adam Block (in fact all of his videos are good).
- You must be logged in to reply to this topic.