Ed Harvey

Sorry, No... not in the current implementation, as this deals with small, repeating, cells. Theoretically it would be possible to create an effect like that but it would require almost a complete re-write 😔

TBH it's been soooo long since I made it (about 15 years ago😲) that I have no memory of why it is called that! Knowing me, it was probably a 'placeholder' name because I couldn't think of a descriptive one, and I never got around to thinking of something better 🙄 I just had a look at the code to refresh my memory of what it does, and it's kind of tricky to explain..... (hence why I couldn't think of a good name) very simplistically, it increases the brightness of the hue selected and darkens the color on the opposite side of the color wheel. BUT, it does this in a way that is somewhat non-linear, so the results are rather 'non-intuitive'. Basically you just play around with the controls and it does odd things to colors 🤪

At long last I managed to find time to update things and (hopefully) fix all the accrued UI issues 🙂 There is no new functionality, but some effects (mostly distorts) have been converted to use the current PdN effects UI rather than their former custom UIs. (This should reduce the need for future breakages) The latest version of the Ed Harvey Effects package is available here: EdHarveyEffects.zip

"Extract Channel" in my plugins (Ed Harvey Effects v 4.0) has quite a range of selections... ?

Aaah, lattice... One of the first effects I ever wrote, and it showed! I think I dropped it from the pack quite some time ago as it was kind-of cr@p I'm away from home for the next couple of weeks - when I return, I'll put a copy of it in the 'Archives'

sorry for the slow reply on this one, I only just noticed it... If I remember correctly, it's a relatively simple transform, the mapping is something like... for a given pixel at x,y (where x and y are relative to the center of the image) mx = (size * Math.Floor(x / size) + size / 2); my = (size * Math.Floor(y / size) + size / 2); ix = (mx - x) * shift; iy = (my - y) * shift; pixel at x-ix maps to x, and y-iy maps to y so basically it divides to image into squares, and maps shifted/scaled sections of the image into each square...hope that helps

Hmmm, you're right - sort of... Somewhere in the mists of time (I haven't done any real work on these plugins in ages) it somehow got renamed Vibrance I'll see if I can find the time in the next few days to rename it back... In the meantime, you now know where to find it Edit: On second thoughts, I think I'll leave it as is...

Since we are now at the release candidate stage, here is a version that works with paint.net V4... EdHarvey-FastFx-4-0-5120.zip [ link also placed on first post ]

Since we are now at the release candidate stage of paint.net v4, here is a version that works with paint.net V4... EdHarvey-Effects-4-0-5120.zip [ link also placed on first post ]

Obviously not using to new v4 version?

Fixed..

Here's a version of my plugins that should work with the latest paint.net 4.0 alpha... EdHarvey-Effects-4.0.5120.zip [edit] ...And the latest 'FastFx' ( I'd forgotten just how fast they were! ) EdHarvey-FastFx-4.0.5120.zip

It's very difficult to specify exact settings to duplicate standard filters as there are many variables (for example, camera response and white balance) which can effect the final picture look. Also, some of the standard filters have complex responses which are not possible to exactly model after the original light spectrum has been sampled to the 3 primary colors It's usually best to just play with the settings until you get the desired result. Having said that, the example you gave, Red 25, is actually easy as it blocks everything but Red - so Hue 0°, strength 100% If you have other specific filters for which you require settings, I probably have that info archived away somewhere at home. Just let me know which ones you are interested in, and I'll see if I can find it...

links restored

Paint.Net's Surface Blur is, in fact, a variant of the classic bilateral filter. It just uses different range functions, which (with an algorithmic sleight of hand) allow it to be calculated in roughly O(n) rather than O(n²)