-
Hi Guest!
We have recently posted a VaM2 2024 Progress Update on Patreon!
Check out the details here: https://www.patreon.com/posts/vam2-august-2024-111164439 -
Hi Guest!
We are excited to announce a new feature on the Hub: Favorites!
You can now add resources to your favorites, and organize your favorites into collections! You can check out the details in our official announcement!
- Thread starter hazmhox
- Start date
Directly play a certain moan level
- Messages
- 1,507
- Reactions
- Points
- 173
I'm exploring the features of this plugin and it's great so far. I found what seems like a bug, though, where I've sync'd a folder of SFX voice line mp3 files. Everything works fine initially, and I can trigger them via Play VA Immediate from timeline trigger keyframes. However, after I save the scene and reload it later, it raises errors like these for all the voice lines that worked before I reloaded the scene:
!> VAMMOAN2 Teaser - Exception caught: Voice line not found (What do you think)
If I open up the plugin's UI, it still shows the correct folder in the SFX options. If I press the Sync folder button, it indicates it found all the files for the Total VA files, with 0 new files added. However, I see the red hourglass icon spinning for a while as it did during the initial importing. After that, it works again for that session of the scene. So, it seems like it may be storing the list of files, but just not hitting the loading step in this case. This issue is 100% repeatable over about 5 scene loads and 2 VAM sessions so far.
!> VAMMOAN2 Teaser - Exception caught: Voice line not found (What do you think)
If I open up the plugin's UI, it still shows the correct folder in the SFX options. If I press the Sync folder button, it indicates it found all the files for the Total VA files, with 0 new files added. However, I see the red hourglass icon spinning for a while as it did during the initial importing. After that, it works again for that session of the scene. So, it seems like it may be storing the list of files, but just not hitting the loading step in this case. This issue is 100% repeatable over about 5 scene loads and 2 VAM sessions so far.
- Messages
- 1,507
- Reactions
- Points
- 173
Damn! I tested that part thoroughly prior to releasing... this is a bummer.
Could you pack the scene as a local version (ie: zip file not var) and send it to me so that I can try to debug?
This is indeed pretty annoying : )
OMG OMG OMG OMG OMG....lel@XcOm03au
Yo,
IT IS, fully released. I mention it is an early access at the beginning to be sure that no one gets mad about potential bugs I missed. The plugin has been thoroughly tested and should not have much, but I prefer to avoid any surprise
It is production ready, you can use it in your scenes.
And if it was not about that mention but more about the "teaser" in the title, this is just meant to imply it's a teaser of VAM2. It's highly unlikely new voices will be released before VAM2, but it's a fully usable and enjoyable plugin for VAM1, so don't limit yourself and use it, that's why it is available
Finally, yes the VA system is per character.
- Messages
- 1,507
- Reactions
- Points
- 173
Schedule or Play VA in the actions/triggers : D
I still need to do the full documentation on that, but this is already mentionned in the documentation
Thanks, I'll look into it.Schedule or Play VA in the actions/triggers : D
I still need to do the full documentation on that, but this is already mentionned in the documentation
- Messages
- 1,507
- Reactions
- Points
- 173
I might not be understanding, but how is that relevant?
Could you give me a use case?
It's a VA system (voice acting), voices are generally triggered on purpose for a specific situation... not randomly : )
I have 1000 voice lines of dialog. I want to press "play" and have a scene (voice dialog between two people) play out randomly.
Currently I am using lip sync for the random part, but your software is able to select per folder per character....just wish I was able to play randomly per folder.
Currently I am using lip sync for the random part, but your software is able to select per folder per character....just wish I was able to play randomly per folder.
- Messages
- 1,507
- Reactions
- Points
- 173
This is something I could add, as the action, like I don't know... "Schedule/PlayImmediate Random VA".
But to be crystal clear:
If your folder contains:
- "Hey How are you"
- "What do you wanna eat"
- "Look out there's a dragon behind you"
That action will pick ANY of those randomly, without context or anything. Is that what you want?
- Messages
- 194
- Reactions
- Points
- 28
For only random voices implementation you can also use audiomate, for playing random voices you choose from the folder or files and the probability of how frequently voice is played so it may play with slight pauses instead of repeatedly speaking after one voice is played or without delay meaning next voice is played instantly randomely.
To be specific;
every (x) seconds, randomly play audio file in pre-selected folder
Then I normally hit record and get a randomly generated conversation and leave it recording
- Messages
- 1,507
- Reactions
- Points
- 173
To be specific;
every (x) seconds, randomly play audio file in pre-selected folder
Then I normally hit record and get a randomly generated conversation and leave it recording
It will obviously only play from the synced folder. This is how the feature works
You'll have to handle the "every second" yourself (which is not hard to do with a timeline).
I honestly will add it maybe because it could help you or like a couple of other persons... but I honestly don't see much use case in reality. So adding a forced playback overtime is not meaningful in the plugin, and it can be done very easily with timeline or an animation pattern.
Am grateful that you will add it <3It will obviously only play from the synced folder. This is how the feature works
You'll have to handle the "every second" yourself (which is not hard to do with a timeline).
I honestly will add it maybe because it could help you or like a couple of other persons... but I honestly don't see much use case in reality. So adding a forced playback overtime is not meaningful in the plugin, and it can be done very easily with timeline or an animation pattern.
I will be your thrull forever!
For the "Play Immediate Random VA" feature, I could also really use a random pooling feature. My use case is that I like to record at least a few variations on any voice line (or types of reaction sounds) and ideally VAMMoan2 could provide a feature for that, where the scene creator would provide a set of voice lines from which the plugin would randomly pull at runtime, either by manual selection of lines, a filename pattern match, or a folder (see MacGruber's Logic Brick "Random Sound From AB" for some examples of such pooling).
I've been using "Random Sound From AB" plugin for this, but it's messy and inefficient overall to use like this, as it requires temporarily toggling off VAMMoan, to avoid conflicts with the person's audio source (and a second audio source doesn't work well for lip sync, etc.), then using a "Random Sound From AB" plugin per pool of lines. That quickly adds up to a lot of plugins for a long conversation, with one per pool. It's also quite a chore to manage all the togging between VAMMoan and other systems like that, all competing for the same audio source.
Another approach, rather than adding this random voice line pool within VAMMoan, would be some friendly way to handshake between it and other systems, for tight control of sharing the audio source that's easy and robust to implement.
Example use case:
I've been using "Random Sound From AB" plugin for this, but it's messy and inefficient overall to use like this, as it requires temporarily toggling off VAMMoan, to avoid conflicts with the person's audio source (and a second audio source doesn't work well for lip sync, etc.), then using a "Random Sound From AB" plugin per pool of lines. That quickly adds up to a lot of plugins for a long conversation, with one per pool. It's also quite a chore to manage all the togging between VAMMoan and other systems like that, all competing for the same audio source.
Another approach, rather than adding this random voice line pool within VAMMoan, would be some friendly way to handshake between it and other systems, for tight control of sharing the audio source that's easy and robust to implement.
Example use case:
- Person is moaning...
- Timeline triggers "Play Immediate Random VA" from "Don't stop!" pool of lines (3 variations).
- After line delivery completes, person goes pack to moaning...
- After some condition, the Timeline triggers "Play Immediate Random VA" from "Yes, just like that" pool of lines (4 variations)
- After line delivery completes, person goes pack to moaning...
The random pooling feature sounds awesome! I too make variations of voice lines
- Messages
- 1,507
- Reactions
- Points
- 173
Yeah, you could also answer to my request to your prior question before diving into another one : )
But to answer your suggestion, for now, I would like to see people actually use the plugin and share scenes with it... AND use the standard VA system before going into complex developments. We were discussing that recently: Voice cloning has been around at a faily reasonable quality, everyone was excited about it here, and still... 3 years later, the number of scenes completely narrated/voice acted is abysmal. So all and all, it was initially a "cherry on top" thing that VA system, not the primary focus.
There are far more important features I would like to implement before diving into granular systems like the one you're suggesting.
That's like having two kids and only asking one of the two to be nice with the other one
I could improve VAMM way of handling the audio, it would NOT improve the way other plugin are doing their job. VAMM2 is on the "good side" of the fence: it waits for ANY sound playing back to finish before trying to playback new sounds.
On the other hand, since I'm not using VAM standard audiosource and creating my own pool, for other plugins it would require them to fit into my pool. The reason behind that choice ('coz I thought about it), is that the normal AudioSourceControl from VAM, requires standard "scene audio" to handle the queue normally. Which is a lot of fiddling just to playback a clip. Knowing exactly how much effort is put into audio "in general", I know this is not a priority at all, and we will try to fix that limitation for VAM 2 by creating a more robust and "normalized" audio system (I'm not even talking about plugins here, just audio sources).
I've tested this a bit more. It happens about every time in one rather complex scene, but only on one of the two people with a folder sync'd for VAMMoan 2 within that scene. The other person with only a couple dozen files always works, but the one with 56 never works.
I tried to pare it down to a more reasonable repro case, by starting with a blank scene, adding just one person, and sync'ing to the folder with 56 voice lines. That repro'd the problem once, the first time I restarted VAM and loaded that saved scene, but then that worked fine in all subsequent attempts to repro it there.
It seems it may be related to lengthy loading times or a race condition of some sort.
If I figure out better repro steps or narrow it down further I'll be sure to let you know.
I can understand that point of view. I'm not meaning to be pushy as to what you end up doing with VAMMoan 2.
I'm just trying to share with you the overall situation as I've experienced it while creating scenes, and provide some suggestions for potential improvements. I think many more people would create more conversational scenes if the creation process was simpler and more intuitive overall.
From my perspective, it's currently quite cumbersome to try and stitch together quality scenes with back and forth conversations, largely due to lack of easy to use features such as I was suggesting.
When I've tried to do this kind of thing in VAM previously, while using the older version of VAMMoan, I ran into quite a few headaches. When I set up triggers to play the voice lines, they were often cut short by moans. I tried routing the voice lines to another audio source in parallel to avoid that, but then the lip sync didn't work for the voice lines. That time around I ended up with a solution that was overly complex to manage by hand, where I was sharing the same audio source, setting up triggers to disable VAMMoan, playing the voice line, then re-enabling VAMMoan. It was a lot of careful management of triggers, and it still resulted in some janky playback in practice.
The voice line getting cut off by moaning was just one issue, as the flip side can be another problem. If the voice line is waiting for the moaning to end, such as by just queuing up a voice line, that injects a lot of variability in when the voice line plays, which greatly complicates the conversation creation process, where you have to build in lots of empty spaces that might be consumed by something in the queue, leading to awkward long pauses if nothing happens to be playing then for a particular run through.
Things like the logic brick approach for pooling options requiring a plugin component per pool of variations also crank up the complexity. That's all on top of writing the dialog, turning it into audio files, editing that into individual sound clips, compiling those into an asset bundle, etc.
I don't have all the answers yet either, and conversation systems and interactive gameplay/cutscenes are always a hard nut to crack, as I've experienced even in AAA game making. A more holistic approach for VAM 2 would likely be very helpful. In the meantime, having a few more tools in our toolbox for VAM 1 could improve things at least somewhat, perhaps even reaching a tipping point where more creators generate more conversations.
Regardless of where you end up taking this plugin, I greatly appreciate your consideration.
Last edited:
- Messages
- 1,507
- Reactions
- Points
- 173
Thank you! I'll check if I can try to see if there something possible to catch that would "fail" in a long loading time circumstance in the meantime.
- Messages
- 1,507
- Reactions
- Points
- 173
I understand your point of view. But your suggestion is not gonna fix what you are talking about here. VAMM2 provides either the pooling system or an immediate way to playback without waiting for the next sample.
In a standard situation, it works flawlessly ( ie: breathing idle and a discussion ). In a situation where you'd want a moan (sex) performance + synced dialog, this is gonna be more complex to keep it in sync if you want a proper performance without cutting the samples in the middle.
You are expecting a procedural/random plugin to fit into the "very intentional dialog/script you have in mind". Which is quite an impossible task on the automated mode. Nothing prevents you from controlling and pooling every single sample yourself on the other hand, but it's a huge task depending on the length of your script.
Knowing how AA/AAA cutscenes are made, you can't expect a plugin to remove all hurdles and tedious aspect of what "cutscene creation" is, because it's inherently complex and very "arbitrary" content. In a real situation, you have to record the mocap, the voice performance, clean up all that, import them into the engine, set them up, script them, configure which line plays where during what animation, ensure audio is properly played back... is it spatialized or is it on the center speaker? etc etc etc... I'm even skipping a lot of other details involved in that process because the list goes on and on.
Cutscenes and dialog management is a huge time consumer. There is no magic function or feature that would reduce the time needed to achieve a proper scene at a granular level. You will have to check every line, verify the sync and so on. You cannot expect a tool to produce a good dialog sequence by throwing wave files at it and hoping it sticks. It's a complete handcrafted process and is the reason why it's one of the biggest budget sink in AA/AAA games.
And if toolset/software claims it could do that, it's just marketing bs to ensure they get your subscription : )
To give you an example: we're working with Juno and Spacedog (so three people, knowing VAM like the back of their hands) for an halloween scene, and it's heavily scripted. We knew we could wrap it in a month but it was still a big challenge. We are almost done, but without a single line of dialog, just handling all the scripting process to have a perfect smooth experience from main menu to ending, we're barely on time.
As soon as you choose to make a very "custom", scripted, narrated experience... there is a shit ton of work involved. I'm not saying that tools can't evolve and make things a bit easier. But you will always have to spend hours if not weeks of work, if you want a very polished scene.
No tool is gonna fix that : )
It's the case in engines like UE5... how the hell you'd want that to be better in a free, "homemade/indie" plugin or game like VAM?
I agree there is no silver bullet for this area. I am not expecting any solution that makes it all super simple and easy.
In my experience there are still usually many ways we can mitigate the biggest stumbling blocks and improve the situation to make various aspects more intuitive and friendlier to use, ever working towards a smoother and more enjoyable user experience.
Perhaps I have not yet tried exactly what it is you are suggesting, with respect to using "the standard VA system" alongside VAMMoan 2. Would you please briefly describe what you mean by that?
In my experience there are still usually many ways we can mitigate the biggest stumbling blocks and improve the situation to make various aspects more intuitive and friendlier to use, ever working towards a smoother and more enjoyable user experience.
Perhaps I have not yet tried exactly what it is you are suggesting, with respect to using "the standard VA system" alongside VAMMoan 2. Would you please briefly describe what you mean by that?
