Fonebone

+1 Any comments on this one from Serif since it was first submitted?

Wgphoto you are on fire! I was going to make this exact suggestion, too.

I second this request, with some additional suggestions. In order to quickly customize the frequency separation process, adding more options to the FS filter itself would be great. For example, being able to choose a different blur type besides Gaussian blur as the basis of the FS. But for full FS customization, as well as performing many other non-FS effects, the Apply Image functionality needs several things added: 1. Most importantly, it needs the ability to apply one layer to another layer (or layers) within the same file, without having to use the workaround of exporting the layer to be applied as a separate file. As it is, within the same file you can only apply a layer to itself, which is very limited in its utility. Perhaps changing the option selectors to “Source Layer” and “Destination/Modified Layer” with options for each such as layer above, current layer, layer below, or something like that. You could also retain the “Image” (outside file) option for the Source Layer selector. So, for example, you could select the current layer as the source layer and apply it to (and modify) the layer immediately below, using equations to modify the individual channels. Even better than that would be a way to perform algebraic combinations of layers (as a whole or by individual channels) with a new layer as the output. This would also be non-destructive to the original layers and wouldn’t necessitate copying the destination layer before applying another layer to it. For example, the 8-bit color depth frequency separation formulas for each of the R, G, and B channels would be something like: 0.5*Layer1(“Original”) - 0.5*Layer2(“GaussianBlur/LowFreq”) + 0.5 = NewLayer(“High Freq”). In theory, it wouldn’t have to be limited to combining just 2 layers. You could almost call that a whole new feature (call it “Layer Algebra” or something) which could do what Apply Image can but a lot more as well. 2. It needs to be made more obvious what layer is being applied to what other layer, and what layer is modified in the process. My suggestions above would be a couple of ways of addressing this. 3. It needs to be much better documented. As an example, I still can’t figure out how to incorporate the Parameter A, B, and C options into the equations and haven’t seen any documentation on it at all. If you’re supposed to be able to put those parameters into the equations and then use the corresponding sliders to see the effect of changing them “live” that would undoubtedly be useful. Maybe this could be expanded as well, to include more than 3 Parameter sliders, and to allow arbitrary end range numbers besides the default range of 0 to 1 (though you could probably achieve this in the equation itself, albeit with a bit more difficulty). 4. An “Invert Layer” check box like PS has would be nice, though again you could achieve this as part of the equations or, more clunkily, by inverting layers before applying them. Anyway, those have been my ongoing thoughts on the topic. It just took seeing Wgphoto’s post to motivate me to take the time to put it all down.