UMA - Unity Multipurpose Avatar on the Asset Store!

You might want to take a look here: http://perttuh.blogspot.fi/2013/10/unity-mecanim-and-ragdolls.html

 

Definitely looks like inverted normals.

 

I don't think it matters in this case, when I'm dragging into the scene the mesh without any textures the problem still occur, but here it is - again, the imported fbx works fine.


I don't understand how the normals could be inverted, in that case it would render the inside, but it is not. Anyway, I tried to export part of the basemesh, and created a slot from that, it worked as expected, I must have missed something during creating the meshes - I can't find what though, I have tried every single option I could come up, without luck.

 

If the actual vertex normals were flipped, then yes. However, in this case an incorrect normal map is most likely the cause. As you say, the light comes from the wrong angle which is exactly what happens if the normal map is inverted.
In order to solve the problem I would recommend looking at your workflow when creating and importing/converting the normal map.

 

Another question for you Fernando if you will indulge me:

As I mentioned earlier we are extremely impressed with UMA and the base meshes are very good. They deform very well and work well when designing clothes for them and so on.
However, for some uses the base meshes are a bit too rough. I am personally very impressed by the EVE character customization process where they end with the user taking a snapshot of their avatar which ends up being very detailed and natural looking.
So, for those uses would it be possible for you to include higher poly count versions of the base meshes? I must admit I have not checked the resolution of the included meshes in the creation kit, but I am assuming those base meshes are the same as in UMA?

Anyway, it would be extremely nice if we could find a way to swap in high detail bases meshes when needed. Would you consider making that possible?

Thanks in advance!

 

But I can't use Unity's built in ragdoll wizard or anything like that, because characters do not exist in editor.
They are generated during runtime.
That's why I need to generate ragdoll rigidbodies and colliders during runtime after UMA generates the character.

 

Well that kinda goes with the territory if you want the full powerful flexibility at run-time you loose the nice editor features.

You could still use the UMA Power Tools to create a prefab. Then it exists in the editor and you can set up all the rag doll controllers you want.

 

But what if I need to customize avatar first and THEN generate ragdoll colliders and rigidbodies?

 

Then you need a run-time ragdoll script that will set everything up when the avatar is built. There is a special event on the UMAData component that you can use to run the scripts just after the character is constructed.

 

Is there any way to do modifications of a generated UMA avatar or do you need to generate a new one with the desired changes?

UMA's hierarchy and names seems to be consistent so generating a run time ragdoll wouldn't be difficult. The simplest way would be to zip through the hierarchy after it is created and assign your colliders and rigidbodies to the transforms by name. Rather than have a separate gameobject you load in on death for ragdolling I would just store those transform in an array. Make them initially kinematic and trigger colliders and then at death zip through them again and turn off trigger and kinematic.

 

Are you saying that I can't even build a character during edit mode without buying a 30$ plugin?

 

Just include the ragdoll directly on the race prefab, and activate it when necessary.
start at ~9:00

 

Yep, I would also prefer handling the ragdoll at runtime.
As far I´ve heard, some UMA users also use http://forum.unity3d.com/threads/143609-Character-System , that provides ragdoll.

 

Hi there!
I might first work on a lower complexity LOD first. Need to find out what people need most. Higher poly version is way easier to handle though.
You may also considert dx11 tesselation for even highest quality.

edit: Forgot to say that on EVE, while playing the game, you never see the avatar with that same amount of vertices of character creation screen hehehe =)

 

Hi - and thanks for the quick answer!

Since you have the base meshes in high resolution already, I would imagine it would be relatively easy to add a base mesh at a higher quality level than currently available. Although I suspect preparing all the mesh parts (ears etc.) adds to the workload so I am not suggesting it is trivial.
Shader tessellation would be a great way to solve this, but unless something changed very recently tessellation is only supported on Windows using DX11?

I know. But being able to generate portraits of avatars that are of high quality seems to be quite appealing to players. We would definitely spend some time implementing the feature if we could arrive at a higher density UMA mesh.

The "obvious" solution to the above would of course be for us to rig our own base meshes to UMA standards but we would really like to stick to the default base mesh to leverage any content packs arriving on the asset store as well as to be able to "pay back" once we have assets worthy of releasing.

Related to this question is whether a high quality base mesh model is available for content creation? I noticed in your youtube tutorial on using Marvelous designer that you recommended using a high-detail base mesh, but as far as I can tell only normal-quality meshes are included in the content creation package? Did I miss the high resolution models or were they omitted for any specific reason?

Many thanks in advance!

 

The regular UMA Generator supports incremental builds. So if you only change DNA only the bones will change, if you only change overlays only the texture atlas will change. However sometimes the lines are fuzzy, if your overlays change size you actually need to rebuild the mesh too.

 

New Version 1.0.10 uploaded.

Changes:
Made DefaultMeshCombiner default in the demo scenes - this is useful for 3rd party contentUMACustomization now have public avatar fieldUMACrowd Males have eyebrows.Fixed script reload problems in the SlotLibrary, OverlayLibrary and RaceLibrary.Changed SlotLibrary, OverlayLibrary and RaceLibrary to use Hashes and added base classes to derive from.Refactored UMARecipeBase to work on UMAData.UMARecipe instead of UMAData. This allows for better recipe format conversion/editing.Added GetInfo(), GetBytes(), SetBytes() and GetRecipeFormats() to UMARecipeBase.

Enjoy!

 

This is a demo of the UMA system with custom support for legacy animations and blending with locomotion. The developers from TestBuilt will share a package with adjustments and instructions as soon as possible.

Stay tuned!

 

Hi there!
Using a high poly mesh requires requires including extra skinning data, this is the hardest part. Simply interpolating existing data would provide a very low quality solution.

I do provide a Higher poly version, .ZTL extension, for Zbrush users. It´s available on content creator pack.
Cheers =)

 

Yay! Great to hear from you guys! Super happy to see those results! =D

 

Announcing: UMA Recipe Tools.

Support Thread: http://forum.unity3d.com/threads/223480-UMA-Recipe-Tools-in-the-asset-store-now!
Asset store: https://www.assetstore.unity3d.com/#/content/14514

• Store recipes in 3 compact binary formats that are around 1/20 of the size of the traditional UMA text recipes!
• Allows direct Recipe editing of DNA and color values.
• Integrates with the UMA Power Tools to give you in editor previews while you edit the dna.

Early bird discount, get it for only $10.

Thanks,
Joen Joensen, UnLogick

 

Hi Fernando

Given the major appearance difference of the UMA avatar using the Skyshop shaders can you give more details of to modify the Skyshop shaders to work with specular and gloss channels of the UMA packed normalmap texture?

It seems this is the only point that need clarifying to work with this amazing skyshop IBL stuff )))

I've been adding in MuchDifferent networking into UMA and its working like a charm with the recipe dna concept.

Wicked Work !!

 

Thinking about it Fernando have you approached the Marmoset guys, they might do some UMA shaders for their product given Unity backs this and its the number one avatar creation tool now for unity?

 

I'll try contacting them tomorrow. I had to do some simple modifications, but I believe it would be better if they can provide those on their package as well.

 

Can i ask why this kinda stuff is not sold as part of your power tools? Are we going to have to purchase a different asset for each idea or you going to keep them under one asset project in the store? !!!

 

One other thought on this area, the packed normal texture your using for UMA. Do you think it would be possible (or you might already have done it) to provide an option for an unpacked specular and gloss texture which could be used just for quick tests? If its easy to modify UMA for that then provides a bit more versatility?

 

Hi there! UMA already can handle that We just pack it that way as it avoid the use of a third texture atlas. In fact, you can use as many textures as you want, this makes it possible to use displacement, full specular color and any extra data you might want. You need to properly fill the extra data at UMAGenerator and use shaders that require those.

 

One of my ambitions for the UMA Power Tools was to make it affordable so that as many as possible would benefit from it. There is a lot of power in there: the bone baking, the prefab saving, the threaded building and the editor preview. I want as many as possible to benefit from that, which is why I've added the PowerToolsIntegration class in there, so other content providers can leverage all this power into their tools as well.

If I was to continue all my UMA development under the UMA Power Tools, I would have to charge a hundred bucks, just like most other big packages out there and that goes against my vision. About the Recipe tools they're hardly necessary for everyone, they're a nicety mostly aimed at mmo and mobile developers, people that cares about network bandwidth and download sizes. So in my optics they really don't belong in the same package.

Sincerely,
Joen Joensen - UnLogick.

 

She-Hulk needs to hit the weight room.

 

So true, was editing on the fly while recording the video.

 

I tend to agree with GXMark, UnLogick. I purchased your power tools with the notion that you would continue to update that package with, well, power tools. Personally, I've had to stop using your multi-threaded generator and go back to the original as of UMA 1.0.10 because it seems to render females with a very wrinkly face (skin tone issue)? If I switch back to the regular generator, the problem goes away.

I get that you need to make a living, and I'm happy to pay for great tools, but this sort of thing just seems to indicate that we can't count on you to update your existing packages nor, in fact, the primary UMA product itself - of which you are a major contributor.

If this is going to be the approach, then I suggest you guys release a non-free UMA Pro/Premium/Plus which includes these tools built in, and that we can count on being updated instead of hunting down dozens of mini-packages which may or may not be obsolete.

 

Yes i have to agree that a major rethinking around how you do these assets needs to be carefully considered. What i suggest to help yourself financially and to keep everyone happy is to provide power tools basic and power tools pro (abit like the concept of unity). Or at least make sure that all the mini purchases are maintained in power tools pro (giving a good deal around pro will be attractive). Given the modular nature of how UMA was designed and the sheer number of mini assets which are likely to spawn from this it will stop the crazy mess of managing versions which will plague UMA later on.

 

I fully understand. Please consider it a feature request then - we would happily pay for male+female base meshes with higher detail for use in character customization / portrait modes.

Ah, do not currently have ZBrush so I did not check those files - sorry. Will be picking up Zbrush mainly for Marvelous Designer retopo so that will solve everything then.

Again, thank you for your work on UMA!

 

Code base documentation not updated

e.g. added cancelled property to latest zip but showing this as the version?

UMA version 1.0.1.0 R - Unity version 4.3

Changes
1.0.1.0R - UMAMaterialBuilder received changes and has been simplified.
1.0.1.0RC2 - Included an Example folder, moved assets and scripts specific to example

 

If you look on the asset store, most of the functionality provided by Recipe Tools can also be found in either UMA Helper or UMA Wizard. This suggests that users can choose whichever of these three they like best, or none at all. Since these other products have already established a price point for the value they provide, it seems only fair that Joen's entry should ask a similar amount. Including it with Power Tools should not really be expected, as they really are different products. Let users decide which add-on features they want, rather than turning Power Tools into expensive bloatware. Selling Power Tools and Recipe Tools as a discounted bundle, when bought together, might still be a good marketing move. In the end though, it's up to Joen how he chooses to release the products he creates.

That said, I agree that the more UMA add-ons appear, the greater the need for some kind of version control between them. There need to be clear standards to ensure that they all work together, or that users at least know which versions of which add-ons are compatible with each other. We've already seen problems in this area as a result of people upgrading Unity, UMA or it's extensions. It can only get worse if controls aren't put in place. What form these controls might take, I have no idea - perhaps a UMA certification test as part of the asset store's screening process for new and updated packs? Maybe advisories saying "This product requires at least this version of UMA, and this version of Unity", or "not yet tested with Unity 4.3.3", so users don't get any nasty surprises as the result of upgrading something. I don't know how practical this will be once there are dozens of add-ons, but it at least bears thinking about.

Perhaps the best thing is to continue giving as much feedback to the developers as possible, so that they are aware of any issues that come up, and then see how they respond. So far, I have no complaints.

 

That is unfortunate, I made a point of clearly writing which features was included. I've spent the majority of my time fixing bugs to ensure I actually delivered on my promise. I stand on my promise to deliver these and is keeping this package bug free.

My thinking was that any new tools I added on top of that would be considered treats, not an assumed weekly service.

I would have appreciated if you had filed a bug report instead of holding it against me. Until recently I held a solid 5 star review that is an indicator of good support. I just checked and is unable to reproduce the issue you describe. Perhaps you can give more information on the Power Tools support forum?

Really?

I'm afraid I don't follow your logic on this one. If you log into git hub you'll see that since the release I have been, without comparison, the most active contributor. In 11 days I've patched the power tools 4 times to deliver on my bug free tools! Since there is no way I can actually defend myself against such speculations, I'm gonna have to take the high ground here and claim that my actions speaks for my self.

This is good and constructive advice. I must admit I hadn't really considered such a thing. I had considered offering a bundle of my own packs and perhaps point to a few recommended tools. But I never really considered making it an official bundle.

I'll be sure to bring it up with the UMA Steering Group!

The reason why I have not yet created a "pro" pack is because right now I have no content for a pro pack. It's all part of the "basic" pack. I prefer to make such a pack when I actually have content for it.

Well this is actually intended. We need more hands to create good content for UMA. So rather than have a big review process we leave it open and see who actually creates good content. This means that packs will be released of various qualities. But it also means that the right pack will emerge without the steering group having to personally supervise it. It also means that it makes sense to release my tools in smaller packs so each of them can be reviewed individually.

I have intentionally steered clear of the Character editors, mostly because I'm not very good with graphical style. My idea of an editor is what you see in the UMA Recipe Tools, a cold presentation of sliders. And I mostly added that to showcase to others how to do integration with the UMA Power Tools. There is branch where people are working on stuff that will help character editors show the right list of overlays for a slot. Once that is added I'll probably add support for that, but the editor in the Recipe Tool will remain a basic editor for npcs. I'll leave it to other to make a full featured powerful editor with themes and genres.


I do have several good ideas on content creation tools, and if those amount into reality they will be part of a separate pack rather than be included in the power tools or the recipe tools. Once my tools reach a critical mass I'll most likely make a pack.

I hope I've answered all of your questions to your satisfaction, if not please don't hesitate to give me more feedback.

Sincerely,
Joen Joensen, UnLogick

 

FargleBargle ~ The way you described that perfectly illustrates that although i do understand why these assets are being kept separate (although i disagree with it) in the end this has to lead to version nightmare proportions probably more so than any other asset ever placed on this asset store. Consolidating these into packs later on will likely cause dislike with clients who have already purchased the separate items. I see this evolving nicely like always !!!

 

There is actually a screening process of UMA content. The UMA steering group have been given free reigns set up the rules that UMA content must abide by to make it into the asset store. So far we've been relaxed about the requirements. Like which folders packages install into and that they must use sub menus under the global UMA menu item.

We've done this because UMA is a great core product, but it really needs a lot of helper tools and wizards to reach it's full potential. As an open source solution we give this responsibility back to the community. Fernando and I will continue to add stuff at our own pace. If that pace isn't fast enough for you, you're free to step up and join us. Like a branch is doing right now with additional fields on the SlotData and OverlayData which will help identify which pieces match and can replace each other. We will of course review all contributions and ensure that the quality remains high!

We have intentionally kept all of the Editors and actual game integrations far away from the UMA core package. We did this because there is not ONE correct way to use UMA. I know that there are mmo frameworks integrating with UMA, I know that the ORK rpg package is looking into how to integrate with UMA. I would hope that PlayMaker steps up and integrate with UMA.

But as reality unfolds and the chaos reaches its usual proportions, rest assured the UMA steering group will begin to exercise it's Caitlyn given powers to bring order!

 

I think part of the confusion with UnLogick's assets is that other currently available extension packages are more explicitly adding on new features or providing usability improvements while power tools and recipe pack sound a bit more like upgrades or patches of the core package.

That isn't to say that UnLogick should work for free or that his extensions aren't valuable. I'm also happy to subsidize further development of UMA or compensate those who developed it, but it is a bit of an awkward situation if you end up with unofficially official extensions that other extensions end up needing to reference (a bit like the current UI situation).

 

Thanks for clarifying that. I didn't really mean my comments as criticism of the way things are being done, more as an opening to finally discuss the white elephant in the room. UMA does not exist on its own. It is an environment, within the larger Unity environment, within the larger environments of the different platforms Unity runs under. The add-on packs, extensions, and other content being developed for this environment are bound to interact with each other in unpredictable ways, especially as the environment itself evolves. Developers and users need to be aware and prepared for this. Developers who work hard to keep their products up to date and compatible with other popular add-ons will ultimately do better than those that can't be bothered.

For their part, end users need to be patient, and willing to work with developers, and give them the feedback they need to solve problems that will inevitably come up. Remember, most of these developers are just one guy, or very small teams, and their resources are limited. When I think of all the serious bugs that organizations like Autodesk and Microsoft have not only allowed into their products, but left unresolved for years, I'm amazed at how quickly UMA's developers are dealing with theirs. It's been out for less than a month at this point, but already looks very solid.

The existence of the UMA Steering Group is an encouraging sign that someone is at least trying to stay on top of an inherently chaotic situation. We've all got to do our part to make this work though, and let them do theirs.

 

For my part, I'd rather have small, inexpensive add-ons that address specific needs than some massive mega-pack, where I'd be paying for features I don't need to get the ones I do. Bundles can be a great way to package them, but won't be for everyone. Anyone who pays for cable TV should know how frustrating it is to only get channel x if you also get channel y and z, that you have no interest in. As consumers, it's up to us to decide which parts of the emerging UMA add-on market we choose to support. Money talks, after all. Still a bit of constructive feedback now and then can't hurt either.

 

I was thinking the same.. Much rather have segregated features then have to buy one giant pack of things I probably dont intend to use. Especially when it comes to optimization, not everyone wants to make use of advanced optimization features and it wouldn't necessarily make sense to bundle those features along with extra customization features, for example. It also allows for better analytics so see which features are more commonly used by watching which modules sell which will drive new features and benefit the customer in the end.

I guess everyone has different preferences, but I much prefer the modular approach that is being taken..

 

I like the modular approach better too..

I use Unity Free and I need to pick choose from assets I can afford. One can't expect an author to create a one size fits all asset package and continue to support it indefinitely with bug fixes, new features and compatibility upgrades without a seriously strong continuing revenue stream.

For example: with the Final IK product recently released...yes that's liable to be that man's full-time job now and even he is extending Final IK with a new product, Puppet Master, not indefinitely extending his original product, which is inherently poor business practice.

With UMA add-ons: maybe in the future but that doesn't change that it would be poor business to create an indefinitely upgradable one-size fits all asset package. All these cheap and free SW from Microsoft, Apple and other large corporations is enabled via huge government and business support contracts and they make these available to home consumers to try and win a leg up on bidding for these contracts. For goodness sake, not because they are desperate to have you as a customer.

One common misconception people have about Linux and GNU and all this other 'free' software is that the developers are moonlighting after work to create this software when in fact it's largely subsidized via the government, universities, and large corporations: the actually hobbyist, moonlighter contribution is miniscule when compared to the overall amount of source code in these 'free' source code movements.

That said: the amount of moonlighting time that Fernando and Joen have put into the core UMA product is substantial, comparable to when Linux was originally conceived by Torvalds so realize that. UMA isn't a 9 - 5 product paid for 100% by Unity Technology.

 

I haven't really held it against you, UnLogick, just mentioning facts. If I had, I might have given the package a poor rating - which I didn't, preferring to first notify you of the issue.

Actually I have looked at the commit history of the project on GitHub, and I did say you were a major contributor in my previous post - for which we are all tremendously grateful, UnLogick. Please don't take this as a personal attack.

Allow me to clarify my point: there is a difference between monetizing on tools which are nice-to-haves, or cater to specific use-cases which not everyone may have, and monetizing on core features which are inherent to the UMA Product itself (such as the speed of rendering and serious optimizations to existing functionality). Within your own Power Tools and Recipe Tools, you cite the following:

and even in your new Recipe Tools you note:

To me, it's pretty clear that the majority of the above are core UMA improvements, not tools which are nice to haves. And obviously, as a major developer of UMA, as well as a developer of these plugins - you've made the decision to not improve the core UMA product, in order to monetize on your additional efforts via these plugins.

And you know what? You definitely should have that opportunity. Your efforts are such that you deserve to be paid for them - hence my suggestion for a non-free UMA Pro/Premium/Plus which includes these core improvements. I'm certainly not saying you should be investing all this time and effort, and not get anything out of it. Those of us who want a pro-level UMA core should be able to pay for it and get it in one centrally maintained location, where we can have the confidence that it will be maintained in lock-step with UMA Free.

I hope that clarifies my position on the matter, and some of our frustrations with the current course of action.

 

It seems to me that UMA works just fine in its intended role, without any add-ons. In the week after it was released, when it was all alone on the asset store, developers who had the skills to dig into the code were already getting a lot done with it. Setting aside whether Power Tools and Recipe Tools are core features or not, what's the difference between a "Pro" version of UMA, that includes these options, and just adding them on to the free version? Either way, Joen only continues to get paid if he maintains and updates the packs as UMA and Unity evolves. The rest is just packaging. The developers of add-ons should probably consider labeling them like Firefox plug-ins, ie: "works with UMA versions xxx - yyy", and keeping this information up to date as new version of UMA and their add-ons are developed. This would at least partly address concerns of users wishing to upgrade. Keeping the add-ons separate has the added advantage of not introducing any licensing complications into the base UMA package either, which currently is pretty much unrestricted open source.

 

Great asset, I bought the PowerTools and am interested in this asset.

Questions:

I'm mainly working with mobile and in-game type of avatar creation and customization.
Wondering if it's possible to use PowerTools and RecipeTools in-game ? If not, could I make a request to add this feature.

Cheers.

 

Hey gang,

As promised, I am starting to get a custom character set up in Unity and would love some help on an issue. Here's an image:



On screen right is the UMA_Human_Male race prefab that I dragged into the scene.
In the middle is the custom character unified FBX that I dragged in the scene.
On screen left is the custom character race prefab that I dragged into the scene.

The custom character is a new model, modeled on top of the unified male, and shares the rig and skin data. I created a new model directory and imported the new characters unified and separated FBX files, making sure they were imported just like the UMA male. I then duplicated the male race prefab and attached everything from the new model. I duplicated the male Tpose and renamed it as well. I also duplicated the Human Male race asset file and made sure all the links pointed to my new character. So, finally, when I hook up the custom character's mesh to the Skinned Mesh Renderer, this is what I get? Any ideas or guidance would be amazing.

Thanks!

 

Go to UMAGenerator, and change the mesh combiner to the default mesh combiner if it's on legacy. If not, the scale of your fbx is probably wrong, these solved the same problem here.

 

Right. Check out this earlier post and Joen's response for more info:

http://forum.unity3d.com/threads/219175-UMA-Unity-Multipurpose-Avatar-on-the-Asset-Store!?p=1475336&viewfull=1#post1475336

 

Hi FargleBargle,

I think the crux of what you're asking is: what's the point of a UMA Pro version vs. downloading third party plugins?

Well there are several answers to that. The first is that the same people who are developing UMA Free benefit from purchases of UMA Pro. This ensures continued development and good-will towards a very high quality free product - similar to the Unity Free vs Unity Pro dynamic. On the other hand, if we rely on third party plugins for core UMA improvements - well then you run into the very real issue of conflict of interest, whereby a core UMA developer will to have to continuously choose between supporting him/herself and contributing to the free version.

The second, and perhaps more practical, reason is that of feature fragmentation. As an end-user, instead of relying on one pro package for peace of mind and a commercial-grade product, you instead end up having to either make significant modification to the core product (and therefore risk major conflicts when new version are released), or hunt down the latest and greatest set of improvements which may exist in many different packages at many different prices and compatibilies. That is not how an indie game developer wants to spend his/her time, I'll tell you that. One of the main reasons people buy instead of build is for the peace of mind of a professional-grade solution which will be continuously updated with the latest and greatest optimizations/improvements/features.

Thirdly and finally, a pro version in no way diminishes the availability of third party plugins and your ability to have a large variety to choose from. On the contrary, it may even allow third party developers to focus on more creative plugins and tools if the core stuff is already taken care of and maintained consistently. Just like in Unity Free vs Unity Pro we still have a beautifully varied asset store, but can count on the core unity stuff to be updated in one place.