![]() Note: DeNoise6 only can be applied to color images.Ġ. Here is what works for me using DeNoise6. But if it works for you, great! It sure is faster than PI NR. And when you miss these areas while masking, its going to mess them up. No control over which form of noise you attack and what I dislike the most: It is inevitably going to sharpen some areas, as well. ![]() That combined with way too aggressive settings (when I use it, I use a factor between 1 and 5). But I also believe that 80%+ of the users don't mask (or entirely remove) their stars and high SNR areas. It proves that, especially when used with care, Topaz can be a fine choice. I had to scale way back on the Denoise AI otherwise it did destroy the detail I had gained with deconvoluion. Here is a reprocess of my 2015 M45 that I used Topaz in place of TGV Denoise after my LRGB combine as non-linear image and before any other post linear processing. Each image is different of course and perhaps Topaz wont work for all images. But if done carefully I think Topaz can be effectively used. Most stars (except for the larger ones) must be removed before Topaz denoised and then added back later in PI. The only way I have found to control it is to run multiple instances of it like using the Clear AI method for the low signal areas and a conservative Denoise AI for the high signal areas and then use pixel math and a mask to blend the two back in PI. Granted I have found that Topaz Denoise AI when applied aggressively easily destroys data. It's an okay choice when starting out but once you really research noise and noise behaviour you'll soon come to realize that it's suboptimal for Astrophotography. But in those cases it really meets its limit) and in addition to that will often create weird patterns. If you have strongly stretched data, in more cases than not, topaz is going to obliterate your data(note: many tools will, its not like topaz is worse than many. The fact that you can't affect all the noise in the same way or set multiple samples like in Dfine2 really makes it a bad choice for Astro images. Having tried many noise reduction techniques including all of Pixinsights tools, Photoshop actions, topaz, NIK dfine and more I think it's safe to say that topaz is the least consistent piece of software.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |