kirk23

+1. It's one of those that must be in any image soft .

So it would be working in sync on different images. For example I have an image/ layer representing height values or normal map maybe and do content-aware fill on certain area. Then I have an image or layer representing color (like a regular photo) and want to do same kind of content aware fill they way all new details would be in perfectly in sync and in same place as I previously done on an image representing my first initial height values image. I think it could be super cool to have in AP too. Surely it should be possible to record from what places the soft copying/clonning pixels from and do it in sync on different image. Like mesh deform could do on another layer for example.

Imo it's just better to render anything against black background to exr output file. With a special shadow catcher material for shadows. Usually 3d soft allow to render "layers", a special set of objects on a black background whatever actual background is. And also capable to output un-premultipled by alpha result so you wouldn't have any black halo at all. Or you could just divide 32 bit floating point exr alpha channel to premultiplied render getting same un-premultiplied result . And I even don't mention deep pixel exr and cryptomatte modern 3d soft doing. So having done it in 3d soft properly would make you never using that matte thing

There are Luminar and Alien skin one(forgot exact name) . I tried them briefly and they looked pretty much same as Lightroom classic. So the market share is quite crowded already. But who knows maybe Seriff would do wonder like they did with Affinity products. Among gazillion of existing toyish image editors they managed to do something in some respects better than Adobe products. So who knows. At the same time I would prefer they wouldn't scatter their resources among other fields and focus on to overcoming still existing gaps and possibly do even better than Adobe. There is a plenty of room for that. It's just a view from a person who pays for Lightroom as a free addon to Photoshop but not using it at all. So maybe I am just not within a target group .

1. select all same color tagged layers at once while selecting one of them in viewport with a special hotkey 2. showing only same color tagged layers once selecting one of them with another hotkey . Kind of a layer comp 3. duplicating a layer would inherit its color tag

Not only they pretty useless but the Displace one is pretty misleading too. I need a filter that would displace in iterations a pixel in a vector set by normal map a gray-scale gradient slope. like this https://docs.substance3d.com/sddoc/slope-blur-159450467.html A live one. what they name Displace is Emboss in any other image soft

It's very outdated way of channel packing in Photoshop and actually even a wrong one because with default Photoshop out of box setting if you copy RGB "grayscale" image in a buffer and paste it to a single channel it recalculates pixel values to be not exactly same . You should first set sGrey for Grey in Color Settings at least. The right way of channel packing is not doing it manually but rather with checkboxes in a layer or group "Layer style" dialog (dobleclick on a layer in layer panel) . That way you could set to what channel selected thing would add its values only. And you could make it a layercomp to quickly switch on and off to same layer. Or even a more convenient way with some 3d party packing script you could find in Internet , (sorry forgot its name) which packs your layer comps named layercomp_ALPH, layercomp_RED etc, into output tga automatically. Affinity Photo Customer Betta have so called "procedural filter" that also could pack layers to certain channel . Non-destructively, with on and off option, even swap channels a layer would add values to. No need to do something through copy/paste absolutely. I I still miss the convenience of LAyercomps although

I use DAM iMatch. It's better than anything I ever tried, better than what Lightroom Classic can offer in that regard. Doubt they could do any better. But if a fresh new DAM could sort my images without me, by some cool AI recognizing not just general stuff but every subject on every layer intelligently , and find things better than Adobe AI which to be honest is as simple minded as rock. And all this would be working on my local pc without 2gb cloud space limit I would be definitely interested. But I guess the whole AI and neural networks area is a too special discipline and would require tones of money to develop and teach. Without AI a new DAM wouldn't have any sense at all

Painting in alphas or even layer masks is not bad nor good, it's just a bit outdated approach People have been whining about Photoshop layer masks not being live, auto-updating and non-destructive for years. Now after introducing clipping groups couple years ago (since CS6) you could use smart objects as masks stacked in it's own mask stack order and blending math providing basically same flexibility as node based image editor. I don't use regular layer masks in my psd files at all anymore. You could do same in AP, well not same but close. In a word you can paint to alpha whatever you want. You just don't need to do it directly in channel dialog and could use regular layer stack for this in much more flexible and non destructive way. Procedural filter could make any layer group to be Alpha or any other of RGB , swap in between channels, etc. Just look at it closely . In my opinion AP needs to overcame still existing gap to Photoshop regarding non-destructive, 100% adjustable work style absolutely necessary for texture art first and then make all that tiny extra adornment like tga etc.

IMO Substance Designer is just one more image editor. They have a nice baker inside still missing a key feature of being able to bake from UDIM textured hi-res objects. And no it's not a shader editor, maybe partly with its pixel processor but in general it's just a node based image editor. Problem is while it's super flexible master of all trades, a kind of a swiss knife it's basically same inconvenient as a swiss knife usually is. I appreciate its node based flexibility but I prefer a simple hammer , convenience over flexibility. They are not the same things. All its basic node/tools are monstrously, unbelievably inconvenient starting from transform gizmo, svg vector tools or curve editor. I've been asking for scaling around a transform center for years. HLS node is almost unusable and you really start to appreciate AP "picker" in curve editor once you try it in SD. It lacks so many all those tiny conveniences we all have accustomed in something like Photoshop years ago already. You have to spend months to make more or less usable tools for yourself there and with a next update half of them are getting glitchy. It's beside the idea of doing everything procedural itself . Which is crazy thing IMO. Once you try to use it as a mixer with hi-res photogrammetry scans it turns slow as hell. I am looking forward to their new Alchemist but from what I saw so far I am not very impressed either. It's just unnecessary extra buildup over same Substance Designer approaches. They just don't understand what artist friendly things are. I bet it's because they mostly listen what various technical artists tell them . That folks usually have codding background experience not an art one. I believe we need just a few "specialized" things in AP and it could be actually better than what Allegorithmic is doing. For art guys at least. ps. I'v been using SD for years already and half of a time you have to care about documenting what you are doing there. Make comments, write special on screen remarks because if you don't the document turns into undecipherable Gordian knot . A week later you couldn't understand a thing in your own files among all those "exposure parameter" and functions. A totally crazy thing either about their UI It's so much a relief to be able to simply pick things of screen in AP or switch some layer on and off quickly to recall what it's doing. Weirdly impossible thing in SD while still perfectly possible in Blender node system for example.( M for bypass a node) Also a thing I requested for years, still "impossible"

I know but I want it as a "live" non-destructive effect for a pixel layer . So if I want to mirror /move/scale/distort/parly erase this layer it would color adapt and its edges would feather/fuse into new situation automatically . Even if it would take some time to calculate and not very much live and would take minutes. Might be a kind of a button to "recalculate" or something. If the code could do it with "current layer and below" it should certainly be able to do it again after I re-scale/re-position a thing. New pixel layer + "current layer and below" approach  is close but having few limitations. I can't reset scale and rotation , can't mirror a thing , can't do subtle corrections to fix edge blending issues. I believe it's all would be working much better with layers

AP "patch tool" is probably main reason I use it. It's better than Photoshop one. Would be nice it could be a live effect for a layer. So we could compose something and every layer would color match to what is beneath it automatically with a same kind of edge feathering /fusion. I bet it would be a bit slow working effect for hi res something but it would be totally understandable . I could imagine same kind of effect doing edge blending with not just feathering but proper subject matter aware blending done by AI but guess it would require a hole new team and years of development so I am afraid we will see it in Adobe first. But since the patch tool is already works in AP splendidly why wouldn't do it live or maybe not exactly live but still working for layers non destructively .

Honestly I don't even understand what do you mean exactly. Composite alpha? Game textures use one single alpha anyway and a practice to paint something into alpha by hand directly is ill one IMO. You could do a separate group of layers with it's own blending math, adjustment layers, vector shapes and hand painting and then use procedural filter to turn the whole group into composite alpha values or any other of RGBA channels, switching it off and on to preview when necessary while restricting alpha input for other layers. And pack anything to any of RGBA within the main layer stack in general. Same way you could in Photoshop where painting in alpha or even layer masks directly is absolutely not necessary too. An idea of regular layer masks in Photoshop is kind of outdated technique. Why using totally destructive layer masks while you can use non-destructive group clipping instead with it's own mask stacks with smart objects etc. Same in AP and embedded docs in mask stacks So I kind of not exactly sure where is the issue. Then you could export as png or tiff and use 3d party converter . If I am not wrong Batch Photo one for example could even do it automatically without you doing a click by looking for a new files to convert in certain input folder. I agree, would be nice to have it in AP indeed without any 3d party solution but I am lacking layercomps and chain links from Photoshop much more. A necessity to keep embedded doc somewhere in the main layer stack to keep it accessible and editable after cloning into another layers mask stack is also annoying imo So to some extent I agree AP is not ready for gamedev/textures yet , just not because of tga export or alpha channel editing. To be honest Photoshop is not very well suited for gamedev too so I believe AP could do better actually. For example they could implement a kind of edge wear mask system based on depth/heigh and normal input imported from Zbrush within "procedural filter". Like Substance Designer does. The math in use is simple like 2x2 and in fact you could do same in any image editor but with proper live filter interface it could be so much more convenient. Each time I do something in Substance Designer I feel myself like a robot drawing something with a huge long robotic arm holding a brush . AP seems so close and needs so few minor touches to be able to do all the same but with much more natural artistic feel of an instrument and a canvas behind your fingerprints.

There are Darktable and RawTherapee. Two open source free raw processing software . Perfectly non destructive. So why invent bicycle one more time. Lightroom does work a bit quicker and it's less complex, more easy probably still both of those programs could do a few things Lightroom couldn't. Like exporting 32-bit floating point un-clipped exr image from RAW file which you could edit/tonemap in APhoto directly and non destructively with all same un-clipped exposure range RAW may have. No develop persona necessary.

Super cool feature in Photoshop I lack so much in AP and Designer. Let it be named any different way. Xara has kind of similar thing "names" A special tagging system for objects that allows to select and view only layers related to certain document look/version. Snapshots is not really a replacement since once you switched back to earlier snapshot and continue to work it makes all other snapshot done after that initial one outdated. Only layer composition system allows to have a quickly switchable multiple versions of same document. Always actual . And would be super cool to be able to export those compositions as separate images of same slice