Thanks! That’s exactly what I was looking for! Sorry you had to see the back and forth above. I’m definitely interested in discussing the merits of this project though, but it would seem like doing something like CSP would probably be a first step, since those objects are non-detailed and we can at least get Krita to read 3D models with poser information. Then with the 3D engine set up, build the texture system on top of that?
Do you think it’d be possible to find some way to connect Krita to blender like that? Like when you open blender’s image editor, you could connect the canvas to Krita’s canvas and have the blender canvas update in real time with Krita’s and thus update the mesh in real time? I think that could be done since you’d just be copying Krita’s final canvas map to Blender’s.
No, you got it the other way around. Showing a model with a texture is “easy” (and the level of details of the model doesn’t count much, maybe for performance only); making a posable doll is much more difficult, I believe.
Ah! Yeah, I figured loading all that stuff and making it work requires building an engine to do all that stuff. But we’d still need to see the model in order to be able to “draw on it” which is what this topic is suggesting. I guess the “posable” part would be a bigger cookie but I wager if the model loading, the texture and the camera can be set up like a 3D model viewer. That would be the first step and then we can load the UV map over the canvas to draw whatever we want and see it on the 3D model?
Yes. But note that we are talking about weeks, if not months of coding It’s still a very long task. It would be good to explore the connect-to-Blender route which would help avoiding duplication of effort but frankly, I wouldn’t even know where to start there.
Hehehe, I’m trying to wrap my head around it myself as you can see above. I actually literally had a dream about this last night where I was using blender to edit my model and literally when I opened the image editing module, it opened Krita allowing me to draw like I would in Krita and see it in Blender. That’s why I jumped into this topic in the first place. I was like “Wow!”
Blender’s 2D is fairly basic but it does everything else SO well that with a powerful 2D engine built in or connected to it, you can do some tremendous things. I have a few images of something I tried to do with 2D in 3D to illustrate a use case.
The drawings were done in Krita, then exported and imported into blender. Blender does support animated textures so if I wanted to change these textures dynamically, I could, but I’d have to effectively make the animation in Krita first, then import them to blender. Seeing how I’m using the bones and weights to adjust how the character is posed though, that would affect the shape of anything I animated. I couldn’t possibly get this level of detail in blender though. This is where I agree with Zovya on his mixing the two softwares. But Yeah, the biggest brick wall is “how to best approach this” cause we’re connecting two independent softwares together.
I do have a lot of experience with web development, so the idea of the apps “Talking to each other” in real time is totally a thing. and then we just connect the dots within blender as both use Python for plugin development so translation should be easy. Extract canvas data from Krita, import that data to blender and load it as the blender image (ideally in real time). Course you’d need to run both software at the same time, but that would be amazing.
and definitely. Figuring out the end points, researching both code bases, and building the packages, definitely a huge undertaking, but any blender x Krita integration would definitely lead to combining the two best art programs into one for the ultimate art experience!
Still that does bring up how Rigging and vectoring was brought up in another thread.
Yeah, there’s no good mobile 3D modeling software. Blender did attempt to make a 3D modeling app, but it was taken down very quickly due to being too unstable. There are 3D model viewing apps and a few posing apps on mobile, but nothing to the level of editing except very simple puppet models akin to Google Sketch up or MS Paint 3D (where you make objects from primitive shapes without any real mesh editing).
On that note, I would love to see a mobile version of Krita since it’s SO easy to whip out my phone to draw something, but that would require a whole new UI for the phone. I kind of wonder if anyone’s working on that? I know Krita does have an android app, but it only really “works” on larger screens like tablets as it’s the desktop version ported over with a touch panel.
I guess at this point it really does kind of sound like “If blender can do 2D, then Krita can do 3D” and a matter of if anyone wants to start up a branch that incorporates 3D into Krita.
The 3D industry needs a freely licensed painting software thats actually good & comprehensive. Armorpaint is pretty much the only open source “painting” software that comes to mind - but it lacks far too many features to be considerable as say a substance painter competitor
If the industry needs this so urgently and wants Krita to be the tool that closes the gap, then this industry should think about supporting the development with substantial funding. Machines that are well lubricated do much more often what the person at the grease gun wants them to do.
Anyways, one of Blender’s strategic targets for 2022 is texturing. There have been a couple proof of concepts vertex painting builds lately and they seem to be progressing nicely. The intention seems to be to do texture nodes for hand painting, but my understanding is that the main focus is on the procedural side of it.
In that sense, I doubt it is worth it for Krita to venture into this too much. Especially seeing the large development fund and pace of development of Blender. I fear that if Krita were to try and pursue this, it would be overtaken by Blender and become obsolete at some point.
For one, I think Krita should explore layout making (like what Indesign is for) as part of the comic frame proposal. Alternatively, vectors are still lacking in all major apps and are also very welcome improvements.
Is there anything preventing Blender from implementing the My-paint brush engine? Or any of the other features that set Krita apart? Personally, I don’t think so, unless it’s bad decisions by the module owners and higher-ups.
Even so I doubt that. you can’t compare a hand painted texture to a procedural one. it is just two different workflows for two different results.
And for what I am still aware Blender is still awful to paint textures or handling colors for that matter to the point i use projections to paint inside Krita but the process sending back and forth is tedious despite the amazing accessibility Blender has. and my props is to Blender to actually allow that swap so easily as others lock you there.
Regardless of all the funding the way Blender handles things does not seem to ever change, because of that, it does not go further than that.
In my eyes it is artists that are unaware of Krita’s future potential in the 3D space. or I may be day dreaming as some strings to connect all dots are missing, but it still does not seem the case to my knowledge. The big problem is Qt6 itself.
I think I just did not used ArmourPaint. I have used Mari, Substance Designer/Painter, Zbrush, Blender. I am not a “texture artist” but I have done textures around. All my arguments are not to replace the others with the “especial features”, just to lay down paint with the brush and nothing else and Krita would still be the bomb. Because the brush sucks on most of them.
So how would you expect Krita to fullfill this role? Do you imagine that there will be a 3D workspace? Or do you imagine painting on a 3D model, which actually paints on a 2D UV map instead behind the scenes?
The textures has to be somehow projectable to UV space. Otherwise people could just do a render in any 3D software, export to Krita and paint on that and call it a day (which Im sure a lot of people do already)
Whats happening under the hood is probably best decided by someone who can figure out a good way tying things up with Kritas existing code base as much as possible
I would imagine it to come to be in one of two ways. A new layer that you could load a file or a tiny docker with those options and display.
Importing 3d objects moving, rotating and basic rendering them will become easy with pyqt6, project paint on objects I am not as sure.
But I think it would be a layer more because a posing 3d doll is something the developers want to do. Kinda like what CSP does. It would be the same premise but painting over a static object.
Regardless PyQt6 implementation will be a bigger project by itself. And only after that this possibility will be open, but still I imagine 3d might take a good bit to develop also despite the new modules for it that really help implementing all 3d options. And stuff like the text tool and others will come first, but the possibility will be there.
I am just the dude saying “yo have you seen this?”.
Also this can’t replace a substance painter from start to finish. Regardless it all is just painting what I am talking about but on a 3d canvas, so still inside Krita goals. And other paid 2d applications are already doing this now and they just paint too… So the idea is not niche!
Photoshop recently got an update for its 3d workspace, it’s now better than ever. Wait, no, it’s gone.
Reading about all the issues with Qt makes me wonder how eager the devs are to rely on Qt6 only for it to not work in Qt7. What a shame.
Shouldn’t posing be a phase 2 goal, I mean, it will never be a full-fledged 3d app, so why bother? Will see how Blender develops, maybe there is little need for Krita to do so in two or three years from now.
3d layers disappeared. What? It must be a graphical card thing? That option dissapears if there is some issue ussually.
I have not seen anything about Qt7 yet. That is a bit sad then.
Yeah with Blender you never really know. After I know how to paint some artwork I am going back to Blender again. I need to update myself on so much stuff.
With krita the is not to replace a 3d app but have poseable doll for reference to draw. Like the anatomy for the pose yeah? There are many phone apps that do that and they do not replace any 3d app.
But I do think Krita will replace Photoshop more and more so even if does not do the 3d aspect of things it will be a big part of the pipeline already. The newest one being storyboarding that has been a bane in the open source world. Krita is worth the investment in it.
No they straight up removed 3d because Adobe couldn’t maintain the code at a reasonably overpriced subscription
[/cynical]
Not that I care. It sucked anyways.
As a layman, I was just noting the current track record and that it would be weird to assume it to change in the near future. I know nothing of Qt apart from the two posts I read, which you referred me to.
If Krita could replace Photoshop & Illustrator in the future, no more Bugdobe, yes please!
Their software gets worse and worse (buggier) every release. Anyways, I get the picture.
No more 3d layers… omg Ps is really on a downfall.
They were pushing it way too much for what was needed in PS but they did have a cool setup with UVs and decals in case of emergency.
I imagine it is to force people into adobe painter now. I loved substance painter but after they bought it I don’t want anything to do with it anymore, they are just gonna ruin it as it becomes more and more a Adobe product.
Having read this thread Quick editing Blender Textures in Krita and the current development on Blender’s texture painting project. I am wondering: is it actually the right move to add a 3d workspace to Krita? Considering the possible performance hits and/ or additional clicks to get data across from one to the other, maybe it is more feasible to improve Blender’s brush engine. Currently, Grease Pencil pretty much only uses a pressure sensor. It has a randomiser (fuzzy dab) and you can rotate the brush tip with an angle parameter. Wouldn’t it be more feasible to bring Krita’s brush engine over to Blender?
Blender can use a better brush engine, not to mention its complete lack of brush management. Such a project is not currently scheduled, see: Paint Mode: Design Discussion & Feedback - #90 by Hologram - Blender Development - Blender Developer Talk. Importing Krita brushes/ bundles to Blender should, I think also be considered as possible interoperability improvement. And has the potential to use the new resource management system of Krita for Blender brushes. Or is this more like wishful thinking?
It’s still a very long task. It would be good to explore the connect-to-Blender route which would help avoiding duplication of effort but frankly, I wouldn’t even know where to start there.
Hello 
What a shame.
Ps is really on a downfall.