Broicher

BTW: I would so much appreciate it, if AP would provide a mechanism to use scripts or to interact with the Mac OS X Automator. Generating the layers is a dumb and boring work. The best would be to have a wedge filter generator, but might be asking to much ... GIMP has something like that though.

When publishing it online you scale it down dramatically, so I don't bother. If one could get "closer" to the image and zoom in, then I would simply spend a bit more time with the beauty retouching, maybe activate some lower frequency wavelets, and take better care of the edges of the affected areas. The image is 8688px x 5792px large, that is far too big to put it online. When I upload the pics I scale them to a width of 1800px and a jpg quality of 90% that is usually a good set of parameters.

Well, as you wish :-) I cut out the face and made it an uncompressed JPG in a before and after version. Note: I did of course some retouche before the filtering, but just a quick one. Before: https://www.dropbox.com/s/l98363jl8vk0uv1/2B2A0334korr_before.jpg?dl=0 After: https://www.dropbox.com/s/2qv0d043yrq8q9g/2B2A0334korr_after.jpg?dl=0 Cheers Frank

I went to the north sea coast with a friend (and model) in order to take some pictures, test my new camera, and really have a lot of new material to put Affinity Photo to the test. Here are two examples: 1. I used DxO Optics Pro for the RAW development of two separate versions of the image, one for the background and one for the foreground, that were combined using AP. Due to the distance there is not so much to see of the frequency seperation, but it was my first real result with AP. https://instagram.com/p/6cqf6Gp5Ef/ orhttps://www.facebook.com/FHBPhotography/photos/ms.c.eJwNyMERACAMArCNPBDQsv9immcItoBNNREWf4y7MzhXzANwrAaq.bps.a.897152580336984.1073741829.240183202700595/1019849258067315/?type=1 2. The next one is a lingerie picture where a lot of skin is visible. What I usually do is, I do use the frequency separation to create a wedge filter, i.e. I separate 1px as high frequency from the image, then from the remaining residual image 2px frequency, then 3px and so on. In this example until I had 1 thru 15px frequency wavelets and a residual layer. I do NOT use the entire image, I select the skin using the selection brush and copy paste the skin only into a new layer and apply the above precedure. In the resulting 15 wavelets and 1 residual layer I deselect all wavelets and then start from the small frequencies and add them again until the result looks good (here 1-7 are switched on, 8-15 are off). I add a mask layer and delete the ugly colorful edge effects with a soft brush from the mask. Here you go ... https://instagram.com/p/6kJ0sdp5GH/ orhttps://www.facebook.com/FHBPhotography/photos/ms.c.eJwNyMERACAMArCNPECguv9immcIMUQPtR108Yclwpn2zgNupQaV.bps.a.897152580336984.1073741829.240183202700595/1021510681234506/?type=1 What do you think? Cheers Frank

If it is not automator, it could be a simple scripting language (like GIMP has Scheme scripting).

No, I got you in the first place. :) I have the same issue with not having either such a picture or a description that matches a German keyboard. :( I was only pointing out, that the only thing inside Photo was the list in the help section which unfortunately isn't very helping in that respect.

I looked at the video again and I have to agree, what he is doing is actually nothing else than removing high frequencies, so basically applying a low pass filter which again is nothing else than blurring. THAT is actually not what frequency separation is capable of. So, I totally agree.

There are many ways of working with frequency separations. This is just one of them. I always start with the healing tool in order to delete the bis "issues" out of an image. Then I usually use frequency separation to separate the image into a residual layer and different wavelets. This is simply done by splitting the frequencies starting with 1px, then re-split the remaining low frequency layer again with 2px and so on and so on. In the end (due to the incremental depression) you have one residual layer at the bottom (low frequencies, so very blurred) and on top of it the single frequency wavelets (highest on top) I usually group them and then generate different wavelet layers (where I leave out different of the above wavelets) for masking. The topmost will be one where I leave the highest frequencies in and de-activate e.g. the 3px wavelet. Below it there will be a layer without the 3px and 4px wavelets and so on. If you put these on top of the original and mask the layers, you will have a very precise control over which frequency information you want to mask out in what part of your picture. You end up with something like this: http://www.frankbroicher.de/photography/gallery/showpic.php?pix=Beauty/FHB_141.jpg(sorry, it may be a bit small, but I assume you get the idea). Cheers Frank

And please inform us about the shortcuts for different keyboard layout. E.g. increasing and decreasing the brush size may be [ and ] on an US keyboard, but these character are available on a German keyboard as alt-5 and alt-6 which actually then changes the opacity, so that is not working. Maybe there is a way of customizing the shortcuts?

You can at least find the shortcuts in the Affinity Photo Help, but in the German version it does not seem to be able to handle Umlauts, so search for "tastaturkrzel". Cheers