happy pride month everyone. give money to your local trans lesbians (this does include us. give us money)

anyway it's uhhh devlog time. i'm a bit sick and my mind is a bit empty but i think i can do this let's fucking go

gonna start off with some fun news: the veadotube Stream Deck plugin is in their official store, finally! you no longer have to download it from itch.io. this happened like this week so we haven't had time to update the website about it, but it'll be done soon :] they did unfortunately capitalise "veadotube" in the title but it's how it is i guess

what's going on in the lab/ko-fi right now?

actually nothing <3 we're working on the very first pre-alpha build of the full version post-patreon, which of course takes time, especially with everything else going on (still going with contract jobs, medical issues, surgery stuff) so we're moving at a slow pace. we're nearing completion however, just need to polish it a bit more so that people can actually use it.

what's gonna be in "the very first pre-alpha build of the full version post-patreon"?

earlier we mentioned veadotube live, the one half of the full version that lets you import multiple avatars and put them on screen and whatnot. this is where we are rn! this will be a very early look at live, which should feature:

• importing avatars from veadotube mini, as well as image files
• positioning them on screen, sorta like editing sources on OBS
• exposing the underlying node system, letting the user edit how avatar inputs are handled

the original draft for the very first build also included anchoring objects between each other (enabling prop support), a scene state system (to move avatars around the screen through shortcuts and whatnot), discord voice call integration (yea), but we realised we already had a lot to introduce in this build. mainly due to the node system.

here's a wip screenshot that absolutely does not represent the final design of everything:

ok but, what is the node system?

in order for mini avatars to work, they need two different inputs: the state that the avatar is in, and whether the avatar should be speaking or not. in veado mini you don't have to worry too much about this -- the UI automatically connects things together. in veado live however, you get the full experience of playing with these inputs:

let's start with the right-side nodes, which are doing the voice activity part -- it starts with a microphone node which outputs both the audio data and the current volume. a volume filter node takes the volume from the microphone and lets you adjust it just like mini does, and then outputs true or false, depending on whether the avatar should be speaking. then the avatar takes that value and makes the avatar talk :]

the left-side nodes do the state machine part! the way mini does it behind the scenes is to have one state storage node and multiple state event nodes. state event nodes are the ones that report stuff like "set the current state to state #1" when a hotkey is pressed, or "push state #2" and "pop state #2" for the press-release hotkey mode. the state storage node takes these events and applies them to its internal stack so it can keep track of pushes and pops.

there are different ways to do state events! there are two specialised nodes for this: the signal state event node, which assigns multiple shortcuts to different states (which mini uses when setting shortcuts to states), and the interprocess state event node, which takes events from outside sources (which mini uses to listen to stream decks and whatnot). i'm not showing those in detail bc i haven't set on a design for them but i hope you get the idea.

let's talk a bit more about microphones then! mini has options for noise reduction, and guess what, those are nodes as well >:] here's what mini does behind the scenes, for each noise mode:

first one is the normal mode; second one is the Silero mode which uses the Silero VAD node, which outputs the scores for silence and voice (and when connected directly to the volume filter, it uses an average of both); third one is the Oculus mode which uses the Oculus Lipsync node, which actually outputs different visemes (if you use HÖNK you should be familiar with this) but when connected directly it only uses whether the speaker is silent or not.

that's right veado mini almost comes with lipsync support but we don't use it <3 but in veado live you could maybe take a mini avatar where each state is a viseme and then connect the Oculus Lipsync node to the state input of the avatar? use your imagination!

you could also do more tricks with state events: another state event node is the boolean state event node. it takes an input that can be true or false, and turns it into state events for when it goes from false to true and vice versa. so you could, for example, connect the output of the volume filter to it, and toggle a special state for when it detects a loud volume (e.g. screaming)!

there are lots of different ways you could build these nodes, which is why the very first version will be focused on testing them and making sure they're working fine and all that.

will i be able to add more inputs to mini avatars?

no, unfortunately! they're built to have two inputs, as they're mini avatars, so they're simple as hell. the idea is to make live stable enough with all the features i mentioned above (anchoring, stating, discording), and then move on to veadotube editor, where you can actually build more complex avatars and then you can expose whatever inputs you want. multiple state inputs, positioning inputs (for face tracking), whatnot!

and when is all this coming?

soon! we have delayed the first full version build way too much, this was supposed to come out to lab testers and ko-fi supporters in april, and then we pushed it to may, and now it's june and if we don't release this in a week or so we'll die, so that's how it's gonna be.

if you're asking about the general public... no idea, honestly! a public release of the full version will happen when everything is done, and it's done when it's done, so yeah. it's a mystery for now.

and what else?

nothing much! things are going at their own pace and we're happy about it. we're only gonna have cool things to announce from now on! yay development!

will add a reminder in here for the other things we do, like BELLA's monday art commission streams, and the fact that i talk about development in the Sindigatas server which is available even if you're not subscribed to the ko-fi :]

that's it bye mwa