This is my first post on this forum, and really pleased to use FMOD Ex.
Me & a friend are working on a custom sound tracker (yes, like the old ones) with custom FM & PWM synthesis (already written and working), as well as PCM support (both Raw and Compressed forms).
Our sequencer granularity is 96 ticks (time divisions) by beat.
E.g. approx. 200 samples @44,1kHz for 140 BPM
My question is :
To implement for example an effect like “retrig” (which re-trigs a sound every x ticks, many times and very fast so), I guess that we are limited by FMod buffering.
Is it possible to do the following :
- Create two FMOD Systems :
- One to do the sound output (with standard buffering setup).
- One to decode the WAV, MP3, … files with a buffer granularity of, for example, 200 samples.
In this second system, could we call system::setDSPBufferSize (200, 1) ?
Would FMOD be stable with such a small value ?
In the other threads I saw that FMod developement is making really nice answers, so thank you very much in advance.
- cpierron asked 12 years ago
Brett, thank you.
I will use FMOD to read all kind of sounds, to apply DSP effects, and to do the output to multispeaker sound cards.
I’m also convinced that I must write at least a small mixer to synchronize perfectly the sounds.
FMOD is clearly a game oriented system, but because you implemented so many advanced things, it is really fine for me.
We will send you a copy of the product when finished (not before 1 year).
It won’t be commercial, but I can tell you that the quality of the integrated FM synth. is really impressive.
brett, thank you for your reply.
The numbuffers >= 2 is logic because it’s a ring buffer. Understood.
At least, we would like to be able to synchronize the following two kind of sounds :
1) Our FM synthesis (FMOD_CREATESTREAM | FMOD_OPENUSER).
2) Samples (FMOD_CREATESAMPLE* or FMOD_CREATECOMPRESSEDSAMPLE**). No file streams at this level.
- For WAVs.
** Because we would like be able to keep the MP3s compressed in memory.
So that an emulated-FM instrument (1) can be synch. perfectly with a PCM/MP3 one (2).
I guess we should create one “virtual” FMOD system for (2), to be able to get the uncompressed PCM chunks from MP3s “on the fly” (creating custom DSPs to rip the sound).
To have (1) + (2) in synch, we have to mix them ourselves and send the output to a “physical” FMOD system to the speakers.
On this base, we could apply integrated DSP effects on this system.
These effets could be a little out of sync (latency), compared to the exact timing required, but it would not be a big problem compared to the quality of the sound produced by these digital processors.
Please don’t hesitate to answer if you see a nonsens in my idea.
Long life to FMOD.
Please login first to submit.