Continuing the discussion from Developer Update - 16 March 2023:
Clearly you don’t understand people who regular VRChat, as has been evident for a long time. People in general don’t care about worlds as much and prefer to stay in their hang-out worlds than move. Majority of people wish to have avatars/characters and interact how they please where they please, akin to GMod, Roblox, Second Life, and many others. Worlds are extremely restrictive, as you are forced to physically GO THERE to do anything. I am quite well versed in graphic and game design even if not in the coding end of things, i don’t need an “opportunity to learn”. I know exactly what i want and why. Showing off various cool things on avatars was always a big thing since the beginning, because it’s portable and individual and can be shown anywhere to anyone. VRChat is going to fall far behind the curve if it keeps sticking guns to dated methods of inflexible content (just look at UE5 revolutionizing the face of interactive media).
Audio source count for actively playing audio is not performance intensive (and should be on it’s own thread anyway), most games use many many many audio sources, and it is common to also abuse PlayOneShot, allowing endless overlapping playback of singular audio sources. There are probably pre-made scripts or prefabbed systems for managing high counts of audio sources. Most people do not use many audio sources that play overlapping, but people like myself and many i know who are more advanced, find the limitation of audio appalling, as it severely limits USER immersion with their own avatars, and be able to play with it and with friends wherever we want, and not be restricted to a single place that has no replay value. Claiming performance issues, when much of VRChat’s performance problems are due to stuff like Udon, limited internal systems optimizations, and promoting wide use of heavy shaders is a bit silly (all of my performance problems lately are cpu bound i never had a couple years ago, most of cpu cost it outside of user ability to optimize).
I have never seen the audio buffer get maxed out in VRChat, outside of the world uoppoH UDON tests where there is sounds for collisions including multi-layer penetration, and even then, i have to extremely rapidly fire a tank round as fast as i can humanly click to penetrate and stack sounds through several impacting layers at once to make the sound buffer start cutting stuff out, so i haven’t the faintest clue what you are talking about.
Even changing the cap from 3 to 5 or 6 would make all the difference in the world. 3 simultaneous audio sources makes immersive effect audio on avatars essentially impossible due to transitions and overlap being essentially ruled out. — I’m talking about complex props and utilities and binaural immersion (think weapons, gadgets & gizmos, transforming avatars, interaction foley etc), not grandiose performances. I’ve had a canny up for this for a long time.
. . . I’m not even going to get started on how avatar audio is like half the volume of everything else (i have to have everything else half volume and then crank up my system volume to properly balance audio ingame, i can barely hear most peoples’ avatar sounds over anything else otherwise).
As a side note, having the ability to enable stuff like particle collision messages on Avatar Dynamics Contacts would also be extremely useful for some, though obviously would have performance cost, but there are plenty of mitigation strategies in the industry.