Salsa and UMA work fine together. The HD body shouldn’t matter
Hi, hey can you tell me please if Salsa One Click for CC4 avatars, for the emotes that are automatically setup, are they supposed to use bones+blendshapes for the expressions of the avatar, or just blendshapes? I’ve used the One Click and the EmoteR is using only blendshapes for the expressions, but I read somewhere that CC4 avatars doesn’t import the bones properly into Unity so I thinking I’m maybe missing out on some expressiveness? It’s possible I’m way off on this.
By the way Salsa is so cool, love how it makes what it does so relatively easy. I’m using it with Inworld.AI (I replaced their default RPM avatars with a CC4 avatar), I had to do something funky to be able to use the Inworld audio (wish I’d seen your comment a few pages back about custom filter chain insert first, would have saved me some headache). My much more primitive solution was to simply save the incoming Inworld audio as a wav then play that, latency isn’t as bad as one might think. But your solution will be better, will try it very soon.
Edit : have implemented a custom filter chain insert, it’s much better, great stuff 
1 Like
Hello,
The CC4 OneClick does use the jaw bone for lipsync and eye bones for Eyes setup. EmoteR only uses blendshapes and the configured emotes are designed to be very subtle, expressing while performing lipsync. You are obviously free to add other emotes to your configuration – we just always felt like less was more for emphasizer emotes.
Glad the filter chain worked well for you!!
D.
1 Like
Thankyou for your reply @Crazy-Minnow-Studio , I agree that subtle expressions are better, it’s more realistic.
May I ask you about 1 more aspect of Salsa please, namely the Eyes feature? I’ve enabled the Look Target to be the users head and now the avatar head and eyes realistically track the users head which is great. However when I start specific animations for the avatar, the rotation clamp doesn’t restrict the angle that the avatars head goes to (I assume the rotation clamp value does this?), the avatar stays looking at the user even when its body from the neck down is facing away from the user.
I’ve noticed that this does this only with animations I have setup like in the image (the Arms Hip Hop Dance animation image), whereas with other animations like in the image (the Idle animation image) the head and eyes track normally and rotation clamp applies, the eyes and head have a maximum rotation. The dance animations are much better setup like they are, rather than setup like the Idle animation, but I’d like the head and eyes to track as they should when the avatar is using the dance animation. Is this something I can fix?
The dance animation that head/eye tracking doesn’t adhere to the rotation clamp:
The Idle animation that head/eye tracking adheres to the rotation clamp:
Hi Tyke18,
I don’t have an immediate answer to why this is happening. Can you setup a very simple scene with your character and the two animations, and make it available to me so that I can do some testing? If you use a shared drive or something like, you can send it to Mike at crazyminnow com.
Thanks,
Michael
Hi Michael, yes no problem I will do that. The app is a VR app so the scene will be for VR but it will play ok in the Unity Editor. Allow me a little time to create a scene and I will email you a google drive link to where you can download it. Thanks. Cris.
1 Like
Hello, I am using your SALSA LipSync Suite assets.
I have a few questions.
- Is version 2.5.5 currently available on the Asset Store the latest?
- Then, isn’t the “OneClick” function attached to version 2.5.5 the latest version?
- Using the “Ready Player Me” service, a customizing system feature, we would like to implement a lip sync function by adding your company’s SALSA LipSync Suite to the character. Is it possible to apply it to dynamically created 3D models?
I would appreciate your reply.
2.5.5 is the current base package. OneClickBase is included in the core package, but has been updated since that release and is version 2.7.2+ as are the OneClicks themselves. We maintain a downloads portal with add-ons so that we can update them more quickly and individually. Sometimes add-ons are updated to address fringe issues for customers and are made available for testing with the general customer base. Once they have been sufficiently tested, they will likely be included in the core product where it makes sense. The downloads portal includes notes and requirements in the description of each file. The next time we update the core package, the current OneClickBase updates will be included.
Yes, you can apply OneClicks at runtime, or create your own runtime configuration. It is necessary to update/bake the configuration data into the serialized controller data, so be mindful of this step. https://crazyminnowstudio.com/docs/salsa-lip-sync/modules/further-reading/runtime-setup/
D.
Hello Crazy-Minnow-Studio, it will be good to see a 2d and 3d fully beginner friendly to advanced tutorials.
I have made the eyes and mouth movement with sprite swaps but I am facing a problem of mouth movement being jittery or too fast with 2d rollback and advanced dynamics turned on.
The eyes are working okay in the beginning with eyelids blinking in play mode and after a few minutes the eye seems to be closed most of the time and blinking time becomes eyes open time.
I think I am doing something wrong. Let me know what I should show to get information to resolve it.
Hi,
Please send an email to assetsupport@crazyminnow.com with your invoice number, and screenshots of your entire expanded SALSA Suite inspectors.
Thanks,
Michael
Hello!
We are creating a Point and Click adventure using Unity, Adventure Creator and Salsa to do the lipsync.
The problem is the lipsync doesn’t work with FMOD. The programer and I visited a lot of forums looking for a solution but we are still having this problem.
If we don’t use FMOD the lipsync is working perfectly, but we need some adaptative music and audio in our game
We hope you can help us.
Best wishes,
Carlos
1 Like
Hi!
I encountered the same issue when I was using another middleware program called Wwise. AFAIK there’s no way to keep using SALSA as a the real time lipsync solution combined with these middleware programs.
However, what I did was to create a little animation clip recorder that records the blendshapes (and maybe also the transforms…?) of the model. You can use this to pre-record the lipsync as an animation. It then saves the new audioclip, the middleware trigger, and some other useful data such as duration into a scriptable object, from which the Player can do it’s magic.
To record you need to (temporarily) import the audio as an AudioClip in Unity. Then when you’re playing the game, you disable SALSA, and the Player is starting the new animation as a RuntimeAnimatorController override, and it sends a trigger to the middleware to start playing the audio file.
Keep EmoteR and Eyes running live for best results.
The system worked quite well, but had some obvious downsides. No more real time lipsync meant that you can’t easily swap out audio clips. The recording pipeline was… fine… but it was always another bunch of steps to do to get everything up and running. Make sure you’re really needing Wwise / FMOD if you’re going this route. An alternative is to run some audio via FMOD, and all voice / lipsynced audio via Unity + SALSA.
I have the SALSA + Wwise implementation here on my github. You’re going to need to modify it to suit the FMOD workflow, I don’t know how to do that. If you have any questions let me know, even though it’s been a couple of years since I last touched that project.
Post your FMOD implementation here too, I’m sure we haven’t been the only SALSAers who wanted to work with middleware
Best,
Maarten
1 Like
check that answer:
1 Like
Thanks @SergS_Workshop I was just getting ready to post a link back to that post. @Trunksome posts some code a little higher up in that thread, beginning Feb 9 (SALSA Lipsync Suite - lip-sync, emote, head, eye, and eyelid control system. - #2527 by Trunksome).
And the SALSA documentation he refers to is here: Custom Filter Chain - SALSA LipSync Suite v2
D.
Can you take a look at timeflow Introduction | Timeflow , which is something like unity timeline and Cinematic Sequencer SLATE；Consider whether you will make a Integration of Timeflow Animation System with Lipsync;
According to the Timeflow documentation you linked, it appears that Timeflow supports integration with Unity’s Timeline system and we already have an integration package for Timeline in our downloads portal. Timeline for SALSA - SALSA LipSync Suite v2
Hope that helps,
D.
Can it support visionos?
Apparently, Unity’s visionOS support does not currently support blend shapes. Most characters on the market today use blend shapes for visemes. SALSA would likely work with a bone-based character, but we don’t have the hardware to test and confirm that.
Hi, visionos 2.0 and latest unity can support blendshape, please check. And you can test in VisionPro Simulator if you have time
OneClickUMA update available: version 2.7.3
For anyone using the UMA OneClick, with the UMA v2.13 update a new parameter is now required for UMASkeleton which broke the OneClick. This has been repaired in OneClickUMA v2.7.3.
You can apply this update over your existing OneClickUMA add-on install.
Thanks,
D.
1 Like