Free character creation in Unity

Umm, you expect that level of detail for free models? Why not use Daz then or Morph3d? I am always confused as to why people bash UMA who do not want to use it and when there is something there that is exactly what they describe already.

BTW, you can put new bones in UMA. I have done it with aliens, but have not yet tried to do it with a human. However, I bet with some experimentation you can figure it out. It is not hard and there are people who can help you on the slack channel.

Try putting some extra bones in the face. Or better yet, there is an experimental blend shape demo in the newest UMA. Try that and see about putting more blendshapes on the face and maybe in the stomach area. Those, imho are the places that need the most work.

Come play in the UMA playground. Like anything else, it takes time and the ability to be patient when learning something new. But you can see good results.

And do not forget to check out the link above for Will's new models. Those should be ready very soon.

 

There is something that UMA can do.

Give you the ability to make your own textures and your own materials. Give you the ability to add your own bones and blendshapes. Give you the ability to convert your own models to UMAs. Give you the ability to make your own clothing and easily convert to just work.

They already do that.

And they do it for free....all of it. Lots and lots of work.

 

You keep bringing this up. Why do you need to rebuild the UMA rig? It has nothing to do with creating a photorealistic character. The rig that currently exists works with mecanim humanoid, which I assume is the main point, and then it also has bones you can optionally use for resizing and expressions.

If you want to change some of the bones or make a completely new rig you can, but you certainly don't have to. Is there something else you're trying to do that you need to build a new rig for?

And if you do want to build a new rig, it's not that big of a deal. It can be done in Blender. Creating new bones is done with a click of the mouse, and then you assign weights to the mesh so that the bone influences the mesh. It's not an endless task, especially since you'll use mirroring to do half the bones. You might spend several hours on it one day, and then tweak it a little the next day ...?

This is in the art department. You've got to either create the meshes and textures you need for your characters or buy them. It has almost no relation to UMA.

When you have your photorealistic skin texture and shader, you can implement them in UMA. There are options out there for shaders, like from Alloy or Uber or maybe a specific special purpose skin shader. Or you can make one using Shader Forge or Amplify.

Bottom line: you can use any shader and skin texture you wish to with UMA. But you can't reasonably expect UMA to develop them for you, as this falls under the category of "each individual needs to make their own game." Until the day that AAA art is commonplace and freely had, that is.

It isn't trivial when you don't understand how UMA works or how to use Blender. But once you invest in learning both, it's not that bad. Making a game requires actual work: programming, art, and game design. It will look less imposing and less impossible once you start familiarizing yourself with the tools and the process.

This is the art department again, not UMA. Art in the asset store right now is predominantly European ... and maybe zombies. That is probably a reflection of where the whole industry is at right now. Many of us would like better off the shelf art options. I think the best likely source for high quality body textures is MCS, which derives their art from DAZ. If you want this, go to their thread and be the umpteenth person to ask for more ethnic options. Keep pushing and eventually it will happen. It's just a matter of getting them to prioritize it.

As for obese characters, again, if you have an art need for your game that is not already fulfilled in UMA you can always create it. There are no technical hurdles preventing anyone from making and using obese character meshes. But you will run into the problem of having to make completely different clothing sets for the obese models, because otherwise the normals and possibly the albedos are going to get stretched. Again, this is far from impossible - you can totally do it. You do have to make different art for each new race, generally speaking. This is not a deficiency of UMA, but a reality of modeling where you use textures in a shader.

So you can use UMA to make modestly obese characters with the basic character models and stock art. But if you move the sliders too far you get texture distortion, and if you make a new obese race, you need new slots and textures for it anyway. Those are art problems or art technical limitations, not UMA problems.

The default art is there simply to get people started. You can prototype for your game with it. It is not intended to be the final art for anybody's game. An effort was made to create a pair of humanoid models compatible with mecanim and of sufficient quality that it could form a standard for artists in the asset store. Since most games use moderately fit looking human characters, that's what the demo models reflect. If you can't find the type or quality of art you want in the store, you'll either have to make your own or bring your standards down to what's available.

I'm not defending UMA, I'm trying to explain it. It's not what you (and others) seem to think it is, and it's not what you seem to think it should be. It's not art. It's a system to take your art and make it procedural, that's all.

 

For gazillion times, I do *love* using UMA, and I'm really *grateful* for all the work they've been putting into the project for free. All I'm saying is its default models and textures in their current form could be insufficient for creating photorealistic models with sufficient diversity, and that can potentially be improved in future.

It's only because you said multiple times that I can just throw in any external models, like one from Daz or Fuse if I wanted.

The main point of using UMA is its ability to customize its characters, changing facial features, body shapes and so on, and that requires pose bones, and LOTS of them. Even with the relatively small number of morph options its default races provide, there are quite many pose bones already.

Replicating those posebones are quite time consuming task, and its in only part of the work if I'm to create a custom race out of external model.

And I actually don't want to do all these thing myself. I just wish UMA could provide better default meshes and textures with more morph options in future, so I don't have to do that which has been my main point from the start.[/QUOTE]

Of course, I don't want to actually do that. But with the default UMA model and textures, the best I can do is something like this:

I can't make those insect like legs any more realistic with what DNAs it provides, I can't really make an authentic Asian character face with those handful facial morphs, and there's no hope of achieving photorealistic realism with that default texture.

Again, I'm *not* saying UMA is useless or anything like that. All I'm saying is it's not likely that everyone can create photorealistic characters with UMA with ease now so there's nothing need to be improved further.

I'll believe it when I see it. I'd suggest you to try creating a photorealistic UMA race with comparable level of customizability with one or two hairs and clothing options.

There's no such options available in Asset Store for whatever reason, so if you can create one in a couple of days, I bet you can get a lot of money by doing that - just see how MCS is doing with their less than photorealistic (but still better looking than UMA) models themselves.

UMA already provides default textures which could be good enough for many types of games. What makes it so unreasonable if someone wanted a better version might be included in future, for other types of games too?

I'm planning to try creating a more realistic skin texture myself in future, since I have no hope of improving UMA character's look without it. And if manage to come up with someting usable, I'll definitely try to contribute it to the project for free so other people might use it.

But according to you, that would be quite pointless since nobody will need my work because each individual have their own specific needs so they better do their own art.

The whole point of using UMA is that they provide a framework and common model which enable people to create and reuse items like hairs, clothes, textures, and etc. which are specifically designed to be compatible with it.

If I have to do all the artwork myself, and including putting tons of posebones and creating hairs or clothings, then all that remains in what UMA provides is some scripts. And as programming is the least problematic part for me, I can probably create something comparable for my specific needs in a few days, then I wouldn't really need UMA at all in that case.

Actually, there's some work is going on right now to make much natural looking obese characters with UMA. I was quite impressed by the result, but probably he needn't to put so much work in that department because there was absolutely nothing UMA was lacking for creating such characters already.

No, default mesh of such a framework is there to work as a standard for which many third party vendors create variety of items like hairs or clothings, just like how it works with Genesis characters in Daz.

While not eliminating possibility to create a custom model if you have enough resources, I believe it's always been quite an important goal for UMA to provide a quality base mesh, so people can create and reuse compatible clothings or hair items.

I know that many of us including myself love UMA, and feel grateful for the hardwork they've been doing for creating such a wonderful project.

But insisting that it already has everything people needs so we shouldn't talk about how it can be further improved in certain areas where it's not suitable for now feels quite unreasonable to me.

 

Again, I'm *not* bashing UMA and I do *love* working with it. And I'm really grateful for them for creating such a great work for free.

The reason I don't use Morph3D is because it's not open source. I want open source projects like UMA to succeed, and I'm even willing to contribute to the project, if I ever come up with something useful for their use.

That being said, it's clear to me creating photorealistic characters is not something UMA is suitable for in its current form, and I believe it can be improved further to achieve that.

But to do so, we should be able to at least discuss what's missing currently and how to improve the situation without arguing endlessly about how we should be grateful for what it already provides, or how it has everything already so anything must be trivial for everyone when it's obviously not the case.

Even if we won't see better default mesh/textures anytime soon, I believe it's worth discussing about it now as it can help people to form more realistic idea as to what additional works might be needed if they wanted to use UMA to create photorealistic characters today.

Obviously, saying either that "it already has everything you need" or "just use Morph3D" won't encourage people from using UMA for such kind of works.

 

Hey, man. Just trying to help you out. But I've done my part. Good luck.

 

Don't worry, I know that you tried to help me. And I'm sorry if I sounded too controntational to you.

 

I visited the Slack channel today, and it seems that several people are working on more realistic meshes currently, and from what I heard some of them might be released for free of charge too.

So, while it's not easy to create realistic looking models today it seems we can expect it to improve soon enough, at least regarding the meshes.

 

You did sound confrontational. I would like to correct a few things you attributed to me which are not right.

You've said that I said that UMA already contains everything a person needs. I did not say that; that's like the opposite of what I've said. What I've said is that you need to supply your own art, because UMA is just a system for procedural character generation. It is not the art. Yes, there is some demo art, but that's just to get you started, so you can see how it works and prototype your game using UMA. UMA is only a framework.

I've already used the analogy of a prototype (handmade models) versus a factory (procedural generation), so how about a new one where we have a paper bag lunch. UMA is the empty paper bag - the container and delivery system - but you need to supply the actual food (the art).

You've complained that if that's true, then all UMA is, is a bunch of scripts. I'd say that's about right. UMA is a bunch of scripts that describe a system for procedural generation of characters, where you input a base character and a library of meshes and textures designed to provide alternate versions of that base model. This bunch of scripts helps for certain types of games (not all), and it is network-friendly and asset bundle-friendly. That's pretty much it. It's not rocket science. You could possibly do up a similar system on your own, if you're familiar with coding and can do things like smooth normals along the seams of meshes, adjust bone lengths, and make texture atlases.

So if a game dev is looking to make a photo realistic game with photo realistic characters, they can do that insofar as they can achieve that goal in Unity. It will come down to mostly lighting, shadows, and materials. And if you choose to use UMA, there is nothing in UMA - which just assembles the meshes and materials you give it - that will stop you from making photo realistic characters.

All you need from UMA is linear color space on the atlases it creates, plus smooth seams on joining meshes, and it already does that.

 

I thought you wanted to step out of the discussion? Sorry, but you are reiterating what you've already claimed for which I already made my points.

UMA is *not* just about scripts. Ask anyone who are actually working on the project, or for commercial assets that include clothings or hair options for the base meshes. Claiming otherwise would be the same as saying Daz3D or Poser application itself is all that matters while millions of third party contents we have for their base models haven't contributed anything to their success.

There's a reason why they are hesitant to replace their less than ideal base meshes with something better, unless it's perfectly compatible with the existing one.

And if I can apply your logic to Unity, then I can possibly rebuke any feature requests whatever by repeating that Unity is just a generic framework so it doesn't stop you from implementing whatever you want on your own.

Needless to say, that's neither going to help those people nor Unity project itself, if that makes people stop discussing about how it can be further improved.

 

Question: Is there any interest in an unofficial BoF session at Unite Austin?

 

I did not see anyone saying UMA has everything that YOU need. It has everything you need to start and add your own stuff.

See...the comment about improving the situation bothers me. Everyone who uses UMA has different needs. So it is really unfair to assume that we all want photo realism or that anyone else but you considers it a NEED rather than a want. lol Now..I am guessing you are not the only one. However, if that is something you want, they YOU need to join us in the Slack channel and do your part. You have to create those textures and texture maps to make it happen.

You cannot expect the hard working people who are currently working on what they need from UMA and are donating their time and efforts free of charge, should turn around and make what YOU need. You have to do it yourself.

That is what open source projects are all about.

I do not really think you are bashing UMA and I am glad that you want to support it. But the best way to support it is to look at your needs for UMA and start learning how to make those come true. Then others who have the same needs can make use of what you make.

You cannot expect anyone else to do it.

They are not hesitant other than the fact that any new mesh should be compatible with existing clothing. Those of us who have bought Will's packs want a model that works with those clothes. Will is making some as I have mentioned but rightfully so, those will be for sale.

So...again, if you want better models, you could contribute by making them because honestly, there are very few character artists of the quality you want who are willing to work for free. I do not blame them. They can get $500 and up for a good model. Why would they do it for free?

All I see honestly is that you see UMA equal to a commercial product and you cannot see why the people working on it right now do not make better models. But they are all programmers, not artists. Fernando, who made the original models is not an artist either. He most like believed that folks would make their own models.

I am saying the same things over and over and I am not sure you understand yet. lol Without a character artist, you will not see new models. BUT you can buy Will's. And you can use any character model from the asset store or DAZ or whatever.

But right now, you are not going to get anyone involved with UMA to make them for you because we have no one to do that. I would like better models myself, but I cannot make them. So I will buy Will's.

 

I never said UMA needs to shift their focus to photorealistic characters because everyone wants them. I merely said that I wish they could also provide an option that can be used for such a task. It wouldn't really prevent anyone from using UMA for their usual, less than photorealistic characters.

By the way, I already joined the channel and if I remember correctly, it could be you who invited me some months ago

I believe I know a bit about how open source works, because I've created and worked on quite many open source projects myself since I started working as a software developer and using Linux as my main desktop from the early 2000s. I still maintain a few of them, some of which have been used by other people in commercial projects too.

And from what I've learned from my experience of working on such projects, and also of talking with people on UMA's Slack channel, discussing about missing features are never something frowned upon in such projects.

Actually, that's also how open source works - if you have time and ability, do it yourself. Otherwise, ask who does politely. Open source is all about participation, and that doesn't only include programmers since making a feature request, or issue report are also important ways of contributing to an open source project.

I didn't bash them for missing features, or demand them to shift their priorities to do what I need first. So I don't think I said anything inappropriate in this thread, regarding the open source nature of the project.

That was exactly my point I didn't say that to find fault with UMA's policy, but to point out compatibility with existing contents is quite important to the one I was arguing with, who claimed that the default mesh is not important because UMA is just bunch of scripts.

Because of the same reason someone is already in the process of making a better mesh for UMA for free, and also because I'll definitely do the same for textures or scripts, if I ever come up with anything useful for the project in future?

Actually, I heard Will's also working on a better mesh which is compatible with the current one. And there's also different one in the process by another member which will be released for free.

I heard of those things only after I asked my initial question, so now I know that I have a better chance of making a realistic character with UMA, since creating a compatible mesh with morphs could be the hardest part of the process.

All I wanted to get from my initial question was such information. And what it can do or cannot do today, so I can have better idea of its capabilities and limitations in creating more realistic characters.

I never bashed the project or demanded they must do what I want right now, and I believe I have a fairly good idea how an open source project works too. I even talked about the same things on UMA's slack channel before, and I got only very informative and friendly responses.

So, I'm quite baffled to understand why so many people seem to think that we shouldn't even discuss such matters here, when those people who actually work on the project are open to feature requests, like any other good open source developers do.

 

I am quite baffled then that you said all these things on this thread when you already know everything I said! lol

You know Will is making new models and that another person is releasing some for free...although at the moment, I do not think they work with existing clothing.

You know that you can ask on the slack channel but you also know that there are no other artists other than the two above and those are already making models.

You know that there are others who want things but they are either already in process or there is no one volunteering to do them.

You know that open source means everyone pitches in.

It seems to me you have your answers. I know you are one of the good guys so not sure why you did not tell us that you knew it all before posting what seemed like a rather demanding thread for an open source free asset.

I will rest easy now that I know you are well informed. Really, you should have said all that the first time. lol

Discussion is great, but it should happen in the UMA Slack channel. UMA is finished for most of us and now it is our turn to take UMA and make it what we want it to be. Comments in this thread completely misinformed potential users on the actual purpose of UMA. It made it seem like UMA was not worth the time because the art was not fully fleshed when in reality, it is only a tool to use to make your OWN characters.

I felt the need to set the record straight. So it is done.

 

As I already said above, at the time of my first post I didn't know that there were two different high quality meshes are in the work.

And honestly, I feel it to be a bit unfair if I was to be accused of starting an arguement or providing misinformation to new users, since I was only replyinig to others who claimed that UMA was nothing but some bunch of scripts, and people needed to create their own arts if they wanted to get any realistic result with UMA, because models it provides are just placeholders or examples.

I know you have a correct idea of what UMA actually is, so I don't know why you thought it was my posts, rather than the ones from a person I was arguing with that need to be corrected, or could give people negative impression about the project.

Anyway, I don't want to prolong this discussion any longer as now I feel better about prospect of creating better looking characters with UMA.

 

but...but...I mentioned it several times in different posts! lol

You did not start an argument, but yeah, I do think that you confused the purpose of UMA. That is my opinion so if you do not agree, that is fine.

I am glad you feel better about it. Glad you checked out the new stuff that is coming. I am very excited about the prospects of UMA and I love what we can do with them. We do plan to get out some asset packs of UMA part/accessories/textures in the near future, after Unite most likely.

I will start a thread in WIP after Unite so my artists can post some of their ideas and get feedback.

 

I don't know if this will help anyone who is looking for free character solutions, but for people maybe who are having issues with skinning characters, maybe something like this plugin would help.

SKINN

 

If you can you think you could try to find a way to actually do this... I`m having a hard time trying to make the characters for a game I`m making

 

Check out the latest version and info for UMA.

 

If you go to Youtube and look for Secret Anorak, you will find a bunch of videos that can help you use UMA, including adding models to make your own UMAs. Remember though that custom models will probably not work with Will's clothing.