kirk23

There were rumors or perhaps just a kind of users wishing that Serif wold do materoal/texture painting veersion of their soft. Would be nice to just have a version of 16bit per pixel RGBA+ XY normals (Z can be reconstructed)+ roughness + height . Basically RGBAXYRH 8 channels 16bit mode and brush dabs supporting same 8 channels . Would in-painting and patch brush work across all those channels to hide seams I would be super happy. All current material creation tools are too complicated and lack the convenience of regular image editing software. All of them lack simple pixel snap so you have to work in doubled resolution at least to keep crisp details. Supporting this 8 channels mode would bring new crowd from gamedev . Don't even need any 3d preview/ window really . It sort of works even now but require to many extra troubles keeping all the channels in sync.

Needs to a part of a brush design with ability to detach from a brush stroke and turn into symbols. With option to use many diffrent of them including bitmap objects , alternating and offsetting both along and perpendicular to the curve and be deformed by the cure.

And flip x , flip Y . would be so much helpful allowing to match the patch to underlying image more precisely. I believe it should be an easy to do feature. Another improvements could be ability to have the patch source from other document and ability to patch alpha /transparency same way it works on RGB . I mean not adding to existing transparency but rather replacing it. As an option.

Paste as a regular layer and add procedural filter where select channel the layer would input only .

it's a critical issue for gamedev. They could win a whole new market by fixing it .

Currently it looks like there is a threshold value in alpha . if it's less than 1 in 255 scale it multiplies over RGB values when you open a file . It's super annoying issue preventing Affinity to be used in gamedev for example. I can save tga ok if I put something with 0.001 value in Alpha so no black holes when you save tgas but once you open such tga back you see those black holes again.

With state name coming as a suffix after a slice name

In a word working same way as it does on RGB with adaptation to surrounding pixels. Currently I see no difference at all from that transparent checkbox. Please, make an option for alpha to be same as other channels. Lets say I have hairs with alpha and would like to patch them from another place or another photo. Currently the alpha ads to existing one doing like thicker hairs and I would like the patch to just replace the silhouette completely with smooth adaptation around patching area. Also would be super helpful to see like 2x2 mesh before applying so we could deform/ bend patching piece a bit . Also flipX and flipX would be very helpful either. Lets do it better than Photoshop content aware move. It needs so little to be actually useful .

I love affinity in-painting but it's often does repeating details, copies things too close off selection . Would be nice to have flip and scale options with adjustment sliders and source area masking . Or perhaps just distance slider the thing would search pixels to inpaint over. Very subtle rotation would be nice option too. Just a tad bit like 1-10 degree

I would like just an easy button to save all customer created things : macros , batches, imagee presets, filter presets etc in one file or make it the way Blender does keeping it all available to new updates. I am also tired of setting everything up again and again. As of gamedev in general the software lacks a lot. Starting from creating black holes at zero alpha . Their target market is starting to get affected by AI diminishing human input demand so why not turn slightly to an another market which is hardly touched, at least where AI does something totally rubbish yet. Height to normal map. 2d displacement that would be really useful , the current one is still not working properly. UV image to image deformation/projection. All it needs is just a slightly more advanced procedural filter with multi layer access . Something we could ask chatGPT to script/code maybe. A few simple conveniences here and there , like multi texture export. Cryptomatte to masks for compositing in CG content . The fact, as a material artist for games I often need a simple tool for compositing materials , height blending something with a gradient . Something where Substance Designer or Substance Painter are both huge pain in your a... . Maybe something to quickly fix bad photogrammetry. Like depth and color layer in-painting in sync without inventing crazy workarounds. Sadly image software developers are in total ignorance of what's necessary for gamedev or at least it's very much looks like so. Including Adobe . What they do as Sampler is totally useless . Their Ai too doing weird blurry stuff repeating like crazy. I use substance Designer mostly and oh boy, it took me almost two decades to make a set of my own noises and compositing nodes I could consider just close to be convenient and most of them doesn't work in Painter. Still hope for a simple software where I wouldn't have to tweak gazillion sliders in "house of cards" style of node construct or recollect why I did all those anchors yesterday. Once thought Affinity photo/ Designer pair could be one but now not especially any more.

Chat GPT writes best on Python for some reason. Sometimes adding convenient features I even forget to ask for. So why not Python? I just wonder. It writes javascripts for Photoshop too but it sometimes takes days to re-do till they start to really work.

I would love Affinity would allow to have depth as a channel . Not spare channel but like RGBD model since many mobile cameras starts to include depth channel from double cameras or time of fly sensors /cameras. It's a long overdue image model.

Subscriptions are fine. They are only annoying when you pay for something you absolutely don't need and don't get features you actually need. Like the whole Photoshop one. I wish I 'd buy cs6 . After it I see nothing worth paying for. Had a huge hope for Ai selection and ended up never using it . It just never recognize subjects I need to select. Some materials features for example like rusted spots or shadows I need to de-light. So gave up on this subject I am not even bothering to try any more. Same for Ai fill . Like filling the shadows with same unshaded material. It's like explaining something to rock . Always does something blurry and weird. Ai in Sampler is a bit better but you also end up never using it.

IMAtch is in fact general purpose DAM. You can add any kid of files to its database. I have just added Blender files to IMatch and asked chat GPT to write an addon for Blender that would write hierarchical keywords, credit and description input fields to sidecar XMP. IMAtch reads this sidecar file perfectly . If only chatGPT could write such add-ons for Affinity photo with Python support or something. Took chat GPT 20 seconds. ps. attached the blender add-on if anyone wants to try with other DAMs . Now I wonder if I could ask the chat to turn Blender into full scale 2d bitmap editor with all features I need before Serif or Adobe would do it :). It sure has lots of it's 2d animation draw tools of its own. And BTW i created the request for exporting XMP sidecar files for Affinity photo . xmp_metadata_hierarchical_addon.py

XMP file with a same name . Should be somewhere in preferences .As well as read option if XMP file is present when opening a file. Could be an easy option for any 3d party DAM software to read aphoto metadata.