That's one awesome equation! With all the variables I thought it would never end...if you come back to teaching procedural texture you will have to break it down for us...STUNNING work!
Dave, thanks again for excellent macros and for excellent academic explanations. I am learning more about procedural textrure filters from you. I want to create a filter that gives me 3 or 4 sliders. the first 3 would be to reduce R, G, or B in a layer and the 4 some potential adjustment to alpha. I can reduce R, G, B (e.g. R = R-r*R) where r goes from 0 to 1, but I don't want R turned to black, I want R turned to transparency so that I can use the new layer as a mask. I tried to reverse engineer your filters - but I must be missing something. Thank you in advance.
Hi Pedro. Good going. To change alpha, just select the A button at the top with the formula. You can turn on any combination of R, G, B and A all together.
@@inaffinityforaffinityphoto3220 Dave, yes, my A has been set to A-(R-r*R)-(G-g*G)-(B*b*B)-a where r, g, b, a are variables from 0,1, but I think I will change a to go from -1 to 1 - ideally I would have wanted to somehow modulate A to be weighted by the amount of R, G, or B. In other words less A for Magenta, then pure Red. Like old fashioned UNIX, the initial learning curve for this stuff is very steep. In the 70's hard to become a UNIX system programmer given the scarcity of documentation ....
It can be done, and I do remember Unix in the 70s (both wonderful). For your formula to apply to the alpha channel, you need the 'A' button selected to the right of where you input your formula. Is this ok?
@@inaffinityforaffinityphoto3220 Yes, I have it working, yes A button enabled. Well, out of graduate school I worked where Unix was invented, THE BELL TELEPHONE LABORATORIES in New Jersey. The documentation came in a tightly written book called a permuted index. I still need to figure out a weighting factor. Pure R, G, or B pixels should have more alpha, I need to play with the formula more. Thank you again for all the tutorials.
Assuming you're ok as per other comment. In late 70s I worked on OS extensions to GCOS Mod400, a Honeywell derivative of Multics. Unix was consequently a nice surprise.
Dave, in which circumstances would you use these parabolic masks, for ex midtones rather than your Luminosity Selection v2 masks? Are there advantages of using one over the other?
Fair question. They are both quite similar, so experimenting and personal preference will be a factor. Also, the parabolic ones are algorithmic and hence use less code, so should be faster in operation, especially on a slow machine. The parabolics also more closely mimic methods used in other luminosity systems, at least in the lights and darks methods.
That's one awesome equation! With all the variables I thought it would never end...if you come back to teaching procedural texture you will have to break it down for us...STUNNING work!
These look very usable. Thanks Dave.
Thank you so much. So interesting and useful.
Dave, thanks again for excellent macros and for excellent academic explanations. I am learning more about procedural textrure filters from you. I want to create a filter that gives me 3 or 4 sliders. the first 3 would be to reduce R, G, or B in a layer and the 4 some potential adjustment to alpha. I can reduce R, G, B (e.g. R = R-r*R) where r goes from 0 to 1, but I don't want R turned to black, I want R turned to transparency so that I can use the new layer as a mask. I tried to reverse engineer your filters - but I must be missing something.
Thank you in advance.
Hi Pedro. Good going. To change alpha, just select the A button at the top with the formula. You can turn on any combination of R, G, B and A all together.
@@inaffinityforaffinityphoto3220 Dave, yes, my A has been set to A-(R-r*R)-(G-g*G)-(B*b*B)-a where r, g, b, a are variables from 0,1, but I think I will change a to go from -1 to 1 - ideally I would have wanted to somehow modulate A to be weighted by the amount of R, G, or B. In other words less A for Magenta, then pure Red. Like old fashioned UNIX, the initial learning curve for this stuff is very steep. In the 70's hard to become a UNIX system programmer given the scarcity of documentation ....
It can be done, and I do remember Unix in the 70s (both wonderful). For your formula to apply to the alpha channel, you need the 'A' button selected to the right of where you input your formula. Is this ok?
@@inaffinityforaffinityphoto3220 Yes, I have it working, yes A button enabled. Well, out of graduate school I worked where Unix was invented, THE BELL TELEPHONE LABORATORIES in New Jersey. The documentation came in a tightly written book called a permuted index. I still need to figure out a weighting factor. Pure R, G, or B pixels should have more alpha, I need to play with the formula more. Thank you again for all the tutorials.
Assuming you're ok as per other comment. In late 70s I worked on OS extensions to GCOS Mod400, a Honeywell derivative of Multics. Unix was consequently a nice surprise.
That’s Great ! unbelievably! Many thanks!
Dave, in which circumstances would you use these parabolic masks, for ex midtones rather than your Luminosity Selection v2 masks? Are there advantages of using one over the other?
Fair question. They are both quite similar, so experimenting and personal preference will be a factor. Also, the parabolic ones are algorithmic and hence use less code, so should be faster in operation, especially on a slow machine. The parabolics also more closely mimic methods used in other luminosity systems, at least in the lights and darks methods.
@@inaffinityforaffinityphoto3220 I will try both out to see which suits me best. Many thanks.
I have ONE QUESTION: Do the free downloads work with my affinity photo which is on a OSx operating system (Mac)?? Thanks.
YES!!!
@@TV-tq5ed Thank again!!
THANKS