![]() Just as in FastStone Viewer: although setting USM radius 3.3px, amount 9 does up the acutance noticeably, applying USM radius 0.5px, amount 23, on top of that ** makes a big improvement. So it is that, in my referenced moon image - while RT does very well - a light application of Piccure does add further improvement. In a broader context, I find that multiple steps give better acutance than one giant stride. It mentions 'Piccure' which claims to add contrast based on automatically accounting for object sizes in an image - in other words, it sets various "radii" without you having to estimate objects' pixel sizes. You might find my triptych image here of interest. JohnGood illustrations and technical info re: method. Certainly they were if I gave it more 'Very Small' wavelet processing.Īnother (expected) consequence is that the file size is increased, especially with wavelets, for a fixed jpeg compression. These could possibly have been over-sharpened. Note that the mottling on the petals is probably raised extrusions on the surface. #3 With de-blur deconvolution, using Gaussian and Lucy Richardson algorithm, 10 iterations:īoth methods show a noticeable increase in sharpness, but the wavelets are better. #2 With wavelet sharpening, a small amount at 'very small' a bit more at 'Small' and a bit more still at 'Medium': #1 Original image with minimal post-processing: This was particularly noticeable in the hairs at the base of the stamens. The original raw image was pretty sharp, but it still showed improvements when subject to wavelet processing, and to Lucy Richardson deconvolution in Astra Image. This did not cause any problems getting down to it, just getting up again. This Scarlet Pimpernel was growing in short grassland, so I was able to lie down in front of it to photograph it. These have the advantage of staying still. Still on Alderney, I took photos of flowers. That Astra Image is pretty good, IMHO.Thanks Ted. Here's the output PP3 so you see what I did (a lot): ![]() I think you can use it in RT either by loading it or by changing it's name to your original crop. Had I not done that, RT might have about equaled the Astro Image. ![]() You'll see in the PP3 that I used some highlight compression and I might have cut back on Lab Lightness a bit. Here's the output PP3 so you can see what I did (a lot): Here's my result at right with Astra in the middle: ![]() I gave it try, needs a lot more than I said above to even come close to Astra Image! Then in the same tab I go to sharpening and select the deconvolution method - maybe just the defaults will do for your image. In the second tab where the sharpening/NR functions are, there is a function called Contrast By Detail Levels - this is simple wavelet processing for contrast only and has two handy buttons at the top to increase or decrease all the sliders - I find that one click on the Contrast+ button is sufficient - perhaps two for your butterfly. I had read the wavelet documentation, but I could not work out how to actually apply it to an image. I am still trying to find my way around Raw Therapee. ![]()
0 Comments
Leave a Reply. |