Plugins + Scripts VAMMoan 2 (teaser)

[XcOm03au]

wow, this is totally amazing. can group different "levels" of intensity together and then have them play randomly...so awesome!

 

[hazmhox]

wow, this is totally amazing. can group different "levels" of intensity together and then have them play randomly...so awesome!

I'm glad you enjoy that new feature

 

[VamEssentials]

I uploaded a Dirty Talk audio var and zip file:

Dirty Talk Audrey - Audio -

AI/TTS generated audio of a model I trained and created named Audrey. Within the VAR is an audio assetbundle version and a folder option with the individual wav audio files depending on how you prefer to use them. An additional zip file is...

hub.virtamate.com

The audio is organized into 7 Categories with 20 samples in each category for a total of 140 audio files.

0 Moans (Example: Ooh yes 1.wav)
1 Intro (Example: I'm so glad you're here 1.wav)
2A Vagina (Example: Pussy belongs to you 1.wav)
2B Vagina-Fast (Example: Pump that pussy x4 1.wav)
3 Anal (Example: Cum deep in tight ass 1.wav)
4 Orgasm (Example: Cumming again x3 can't stop 1.wav)
5 After (Example: Can't wait to fuck you again 1.wav)

What I do is load the folder based on the part in a scene I am viewing and each folder has a pool of audio files for that part. For some models it will be over 100+ audio files in that category/pool with some long descriptive titles to help separate them. I then switch between folders based on the part of the scene I am viewing or mostly just leave it on a default of Vagina-Fast or Anal which are the most commonly used.

I usually just play the audio locally with VLC Media Player without lip sync because there is not really a smooth way to load them and have it lip sync from scene to scene that I've tested yet or am aware of.

 

[VamEssentials]

Another challenge I noticed is if I open a scene with Vammoan like ascorad's Head's Will Roll for example, and the pitch is set by the creator at something like 1.05 or 1.10 for example which sounds good for moans, it auto changes the default pitch audio to that. If I set pitch audio to 1.0 so it plays the audio from the folders with something like RandomSoundFromAB on the model, it adjusts the default pitch to 1.05/1.10 on all the local dirty talk audio and sounds off. I usually just go into Vammoan and pick another voice that works with 1.0 pitch. So it seems having an option to keep moans at 1.05/1.10 in Vammoan is fine but maybe without effecting what a user manually changes the default pitch to for other main dialogue audio. Some user may not even know Vammoan is overriding pitch and think it's coming from timeline so may not even know how to fix it. Just something to consider, if that's even possible.

 

[hazmhox]

Another challenge I noticed is if I open a scene with Vammoan like ascorad's Head's Will Roll for example, and the pitch is set by the creator at something like 1.05 or 1.10 for example which sounds good for moans, it auto changes the default pitch audio to that. If I set pitch audio to 1.0 so it plays the audio from the folders with something like RandomSoundFromAB on the model, it adjusts the default pitch to 1.05/1.10 on all the local dirty talk audio and sounds off. I usually just go into Vammoan and pick another voice that works with 1.0 pitch. So it seems having an option to keep moans at 1.05/1.10 in Vammoan is fine but maybe without effecting what a user manually changes the default pitch to for other main dialogue audio. Some user may not even know Vammoan is overriding pitch and think it's coming from timeline so may not even know how to fix it. Just something to consider, if that's even possible.

No can't do sadly.

The wonderful audio setup of Unity ( this is obviously a /s ), is "AudioSource based". If you prefer, the sound itself contrary to Unreal Engine or middleware like FMOD/Wwise does not contain informations for the playback. The source is the one responsible for the final pitch. Which means that if you want a sound to shift in pitch during playthrough, you need to update the audio source.

So, I have no real way of knowing what "pitch" you'd expect for each VA file.

The only thing I could add is a "VA pitch" slider, which would default at 1 and would set the audio source to that pitch when you playback a VA file then would go back to "Voice Pitch" for vanilla VAMM2 moans.

But I won't be able to give the ability to set pitches individually for each VA files.

 

[hazmhox]

@mihuousa

but the sound is still very small when it is adjusted to the highest level. If there is music in the scene, the vocals can hardly be heard. I hope it can be improved in the new version.

You're answering to your problem in your "report" : )

The voice is mastered to be at a -18db peak. And I want to keep a proper dynamic range to have subtle, intimate voices at low intensities and normal volumes when at max. If your music overwhelms the voice... the voice is not the problem, your mixing on the music is : )

There is an important thing to understand about VAM, it requires to control all aspects of the creation. You will have to do a mix pass on your scene and make the whole audio cohesive yourself. In your case, I'm pretty sure your music is blasting... and this is where you should do the tweaks.

 

[VamEssentials]

No can't do sadly.

The wonderful audio setup of Unity ( this is obviously a /s ), is "AudioSource based". If you prefer, the sound itself contrary to Unreal Engine or middleware like FMOD/Wwise does not contain informations for the playback. The source is the one responsible for the final pitch. Which means that if you want a sound to shift in pitch during playthrough, you need to update the audio source.

So, I have no real way of knowing what "pitch" you'd expect for each VA file.

The only thing I could add is a "VA pitch" slider, which would default at 1 and would set the audio source to that pitch when you playback a VA file then would go back to "Voice Pitch" for vanilla VAMM2 moans.

But I won't be able to give the ability to set pitches individually for each VA files.

Yes that's exactly what's needed! I think most people would just want to control the overall master pitch (for loading speech audio) and moan pitch audio separately as two separate groups, as they would be from two different sets of audio. This would then give you more control to try to get two different audio sets to match better.

 

[hazmhox]

hazmhox updated VAMMoan 2 (teaser) with a new update entry:

QoL and Bug fixes

QoL

• Added a "Voice (VA) Pitch" suggested by @VamEssentials. You can now have a different pitch for VAMMoan's voice and your VA.

Bug Fixes

• Attempt at fixing other VA bugs:
  • Switched everything into a coroutine and let the loading process of the audio files go on by one to avoid fails during heavy scene loadings
  • Converted paths to absolute paths from the root of VAM instead of keeping a potential relative...

Read the rest of this update entry...

 

[BStarG2]

Amazing news, this new version looks really good and more advanced than the original (which serves its purposes well enough I'd say in most cases).

The only thing I do somewhat 'worry' about is in regards to the 'Slaps'. My hope is that somewhere in the Advanced Options section, we can place Colliders where we want them to be, and shape them more (Radius, Length, Depth, basically all three Axis customization). Because - in my opinion - the Pelvic Slap capability of VAMMoan is limited in that way (in terms of collision detection, and how 'dynamic' it is, or rather lacking if I may say).

Essentially what I'm saying is I only hope that the Slap sounds themselves will play at the corresponding speed of the animation (or rather, technically speaking, at the same speed at which the collision detection is happening... if at all possible with VAM1's engine that is, maybe it's not; I'm not sure on that myself). Currently if the animation becomes (or is set manually or in Timeline) too fast, or if the Atom characters vary too much in scaling (I.E. female too small compared to the male, or vice versa) then the collision isn't happening at all (so no Slapping sound at all happen) because the only 2 colliders we have (basically Anus and Vagina) can't register what's happening.

So what I suggest (just in case it wasn't considered, and IF of course would be possible to do), would be for VAMMoan2 to be able to use specific Atom nodes as a collision detection method (on top of having the normal colliders, or perhaps... placing new colliders on the character Atoms, whichever would work best).

Anyway, it would take more time to discuss in details but I am 100% aware that you know what you're doing way more than I know what I'm talking about. I just hope that VAMMoan 2's capabilities in that 'branch' of the plugin (the Slaps detection and precision) will be looked at as well. Because it's very neat to have super complex voices... but if slapping sounds during spicy scenes can't happen at all (or aren't happening consistently) it sort of destroys the moment so to speak. Both the voices and the sound detection (*or rather, collision detection, for slap sounds) need to work in tandem.

 

[hazmhox]

Amazing news, this new version looks really good and more advanced than the original (which serves its purposes well enough I'd say in most cases).

[...]

Thanks : )

For the slap thing, what you're suggesting is probably an "unsolvable" dilemma. The whole collision thing for slaps was just a way to ease implementation for more standard animations and users, but ideally that kind of "additional SFX" would need to be synced on the animation manually (keyframe the SFX).

The reason behind that is because, the logic of collision itself is a sub-par implementation in terms of precision and control. Due to the general "instability" of how colliders works in Unity/VAM, the faster it gets, the more complex it is to consider what is an actual slap and what is a random "touch" you should ignore.

What I could give you is a way to control through direct methods the playback of SFXs. If you prefer, a "Play Immediate SFX" and the list of available SFX in VAMM2. This way you can author your slap in your timeline animation at the speed you want without hoping that your collisions will be (or not) detected as they should.

And I think this is the way to do fast animations. I have zero hope of finding an accurate method to ensure a 170bpm slap would perfectly work through dynamic means.

 

[TBD]

Custom colliders are an option but they'll never be 100% accurate. In reality, slap sounds depend a lot on angle and shape of the surface. We'd need mesh colliders but the performance impact is probably too high.
Scripting it by hand will yield good results but is also a ton of work for mocap'ed animation. I can't imagine going through 15+ minutes of mocap to place slap sounds of varying intensity.
TL/DR: both external triggers and customizable colliders should be supported

 

[hazmhox]

Scripting it by hand will yield good results but is also a ton of work for mocap'ed animation. I can't imagine going through 15+ minutes of mocap to place slap sounds of varying intensity.

Welcome to my/our world as sound designers bru
For ex., we sync pretty much all the footsteps on lower body animations ( and we have a shit ton of them -to be polite- ). But you can consider that there is pretty much never, ever, dynamic audio triggering through collisions. You don't waste frametime on something that works exactly the same with manual keyframing.

Also, to reframe a bit: no one does 15minutes of "riding/slapping mocap". I've never came accross one scene like that... and if there are some, that's probably ultra niche. Most of the last mocap I played are not even animated on the thrust aspect, those are handled by manual animation or even old ass animation patterns.

TL/DR: both external triggers and customizable colliders should be supported

External triggers of all the SFXs is definitely a "gonna do it" one as I said earlier. I'm even thinking of creating a "mid-body" generic audio source to playback custom sounds ( it could be like fall sounds, body sounds, whatever you imagine ).

But custom colliders, not gonna happen. I'm not wasting my time on a feature that I would:

1. Need to support / debug / explain over time
2. Need to document
3. Spend time on it and would end up probably used by three persons over 5 years

The keyframe/trigger system will leverage that limitation... At some point, if a person is too lazy to keyframe it's own mocap. I'm not the one taking that to my end because said person don't want to work for the sake of their own content.

 

[TBD]

Also, to reframe a bit: no one does 15minutes of "riding/slapping mocap". I've never came accross one scene like that...

One Knotty Night - Scenes -

3-pose cowgirl scene based on X_Duke's mocap Just One Night from 2021 What's in the box? 17 minutes of mocap'ed cowgirl sex in total fine tuned SR6 integration optimized for performance in VR-POV four soundtracks to choose from including house...

hub.virtamate.com

But custom colliders, not gonna happen.

Worst case, I can hook it up to my triggers plugin. Maybe I'll write a plugin to add custom colliders to anything, at some point.

 

[hazmhox]

Well! I guess someone's is gonna discover what a SD job really is lol!

Worst case, I can hook it up to my triggers plugin. Maybe I'll write a plugin to add custom colliders to anything, at some point.

Yup! Absolutely. Having interconnected systems is a good possibility too. Since the "action" is gonna be public obviously. You could hook it up to your trigger system indeed.

 

[hazmhox]

hazmhox updated VAMMoan 2 (teaser) with a new update entry:

QoL

• Added Life 14 support
• Improved code to allow manual headAudioSource playback (which allows you to play sounds from normal triggers without VAMM2 interrupting them)

In depth explaination for manual headAudioSource playback

Before this version, VAMM2 was playing sounds on a schedule system. Which meant that, if you tried to play any sound file from, let's say... a UI Button, directly on the HeadAudioSource of the character with VAMM2, this...

Read the rest of this update entry...

 

[SlimerJSpud]

So now we can get her to say "Not like that!" instead of "Just like that" when we want to? Cool!