For a virtual reality lab setup, where the world is displayed around the player in 360 degrees, I need to setup something to do 3D audio.
I can’t use the normal 5.1 speakermode since that relies on the player always facing the front speaker, and rear left and rear right really are more of a next-to-your-sholder position rather than at the same distance as front left and front right but then behind you. Since our player can turn around at will, that will not work to do 3D sound placement.
I want to setup something using 4 speakers and a subwoofer, and optionally use a fifth speaker as an UP speaker. Am I correct in assuming FMOD can’t do this, except by going to RAW speakermode and doing the 3D panning calculations myself per speaker?
- ndesle asked 8 years ago
It would be nice to set each ambient speaker real world position individualy yes. I’ve seen a few posts by people making a vr environment like yours.
for a 5.1 or 7.1 setup you would place your ambient speakers anywhere in your space… eg front, back, left, right, up and down in a cubical room setup.
You would need an aditional REAL WORLD position/orientation for the listener too. and FMOD would do the rest.
Supporting walking in a VR room or staying at the center, turning/tilting the head. It’s an interesting setup.
Even better would be true 3d support for stereo headset. There are methods to fool you, by modulating the sound, into thinking the sound is really ahead. Though I think doing this live is very intensive
The problem with doing headphone stereo 3D sound (which FMOD has real HRTF support planned for, as I read in another thread on this forum) has some serious issues, like the fact that no two people have the same ear-shape, so it’s hard to build a good setup that works for everybody. The second big problem with headphone 3D is that we humans use a kind of triangulation when determining where sound is coming from, and we do this by moving our head slightly, to get different readings basically. With headphones, you can’t do this. This makes it a pretty bad solution, unless you have exact orientation tracking with your headphones, but then latency can potentially become a problem again.
All in all, I prefer real speakers, they are a lot easier and they tend to work better than the much more expensive HRTF solutions imho.
Actually our new system uses personalized HRTFs, we believe it is the first of its kind. It is not a generic ‘one fits all’ HRTF that fails in so many cases (and is basically what everyone else is doing).
A press release about it is here http://www.gamesindustry.biz/articles/p … how-at-gdc
[quote="ndesle":3vlkl8cp]Ah, that’s an interesting function. Shame that it only takes 2D positions, that kinda kills my idea of using my spare (5th) speaker as an UP speaker. But at least I can get 3D (azimuth only) working quickly then.
Channel::setSpeakerMix could be an option. but then up have to take control over the 3d stuff yourself. There must be a better way.
For now it’s OK. I was going to use the UP speaker for a selected few effects (e.g. thunder, or maybe the occasional aeroplane in the sky) only anyway, so I’ll fake that by making those a custom sound where I use setSpeakerMix myself
Meanwhile I have it working with 4 speakers and the subwoofer so I’m happy!
The best solution for FMOD to support this by default in the future is by having a setSpeaker3D method that takes a 3 dimensional vector for the speaker positions, but I can imagine that my setup is such a rare occurence that it is not something that is high on the priorities list of FMOD
Please login first to submit.