I’ve been trying to find a compressed 5.1 audio solution for music to be used in our game, and I’m having a difficult time. Ideally, I’d like to just drop AC3 files into FMOD, but since that format is not supported, I’ve looked at alternatives. Converting AC3 to 5.1 ogg files has had no acceptable results. I can’t afford the disk space of using multi-channel WAV files.
I got excited when I considered using WMA9 5.1 files, as they’re small and sound excellent, but even though I can import them into FMOD Designer, the output is downmixed to stereo. I have the event set as a 2D voice, and the six necessary channels are enabled in the event properties, but it still gets downmixed.
I saw in another thread that as long as the codecs are installed properly on the user’s system, the files should play back properly, so what is causing this problem? Playing the same file back in WMP sounds like it should. Is there any way to get WMA9 5.1 files to play back without having FMOD downmix the output? If the answer is no, does anyone have recommendations for formats/codecs and encoders/converters to solve this issue? I’m entertaining rendering 3 stereo pairs and stuffing them all in one sound definition to accomplish what I’m trying, but I’d rather have just one file and not have to worry about synchronization issues.
Yes, I understand. I don’t currently have the ability to render multichannel WAV files, but my programs [i:2ezjyotd]can[/i:2ezjyotd] render surround projects as 6 separate mono WAV files, which I’ve then taken and grouped them into one sound event and used the surround pan effect to send each rendered channel to the correct speaker. This works just like it needs to (although if anyone knows of an easy utility that can create multi-channel WAV files from 6 mono source files, that’d be a help to me).
The problem I was having earlier involved WMA9 5.1 files mixing down to stereo, even if turned all the feeds to the appropriate level. In that case it was ending up as quadraphonic stereo, but I was operating under the assumption that using a compressed source file would result in a smaller FSB, and thus why I was searching for a good compressed 5.1 format that FMOD would accept. However, I later understood that the format of the source files is irrelevant given that FMOD would just decompress and recompress them, so I’m now using 6 mono WAV files to get surround music and at a higher quality than I was originally expecting.
Thanks for the help though, Peter.
FMOD will convert the sound to whatever format the wavebank is specified as (when compiling the project), so the source format you are using is irrelevant really.
One thing that would be relevant though is that you are using a lossy source, AC3. You should be using multichannel PCM wav for your source data, so that it only gets lossily converted once, not twice.
At the moment only PCM or ADPCM format will allow multichannel encoding from within FMOD designer.
As for WMA9 and multichannel as a source, i havent actually tested if it works or not, so we could try it, but as i said, thats not the format that would end up in the game (ie ps2 doesnt support this sort of stuff, it only supports VAG)
multichannel wavs doesn’t bypass the downmixing at the moment. I’ve been trying with interleaved 5.1 and Quad wav files with no success. Using a monofile and panning it around ina 5.1 system through a Direct Sound driver works fine.
[quote="brett":2p209858]FMOD will convert the sound to whatever format the wavebank is specified as (when compiling the project), so the source format you are using is irrelevant really. [/quote:2p209858]
Wow, I’m glad you told me that. So if I take a 2 minute song and import a WAV copy and an OGG copy, they’ll end up the same size after compiling the project (I’m developing for PC)? I was under the (apparently false) impression that if I used a compressed format, it would make the FSB smaller. It would save me a lot of trouble and solve this problem if I was incorrect.
Edit: Sorry, I just realized that it wouldn’t take me long to actually just try this. I got the expected results.
Regarding multichannel wav’s:
The context you’re using isn’t clear, but I have made an event using a 5.1 wav file.
In 3D mode, you get a mixdown to a point source in a 3D environment, or a pair of point sources if you set the speaker spread > 0. This is the intended behaviour.
In 2D mode (which is what I assume you’re using), there should be no mixdown. However the default levels of the feeds from the source are:
L : 0dB
C : -60dB
R : 0dB
So the centre, surrounds and LFE won’t be heard unless you set them high enough.
- Anonymous answered 11 years ago
Please login first to submit.