0
0

Hi!

I’m experimenting with FMOD’s obstruction/occlusion capabilities, but I can’t get one thing straight. Is FMOD really doing full geometry based obstruction/occlusion simulation?

Let’s suppose I define the geometry for a room with a doorway. There are some sound sources inside the room, and the listener is outside with the door on his left. What should be done to hear the sounds coming from the left as coming through the door?

Is it enough to define the geometry?
What directocclusion/reverbocclusion values should be used for the walls?

Thanks,
Greg

  • You must to post comments
0
0

If there is a line of ‘sight’ from the listener to the sound source there will be 100% normal sound combined with what ever reverbs. No occlusion.

[quote:11nlzfej]Is it enough to define the geometry? [/quote:11nlzfej]
You must also define reverb. Use the FMOD::Reverb object to create reverb spheres and/or use System->setReverbProperties and System->setReverbAmbientProperties.

[quote:11nlzfej]Let’s suppose I define the geometry for a room with a doorway. There are some sound sources inside the room, and the listener is outside with the door on his left. What should be done to hear the sounds coming from the left as coming through the door? [/quote:11nlzfej]
There is one layer of occlusion: the wall, so the sound is occluded by the occlusion factor of that wall. There is no calculation regarding the proximity to the door. To hear a different amount of occlusion nearer the door you would simply create polygons with different occlusion values nearer the door.
[quote:11nlzfej]
What directocclusion/reverbocclusion values should be used for the walls? [/quote:11nlzfej]
It depends on the material the walls are made of. So if it was a thick concrete wall you might have values of (1.0 0.5) so you can hear half of the reverb through the wall and none of that actual sound.

Hope this clears things up.

-Pete

  • You must to post comments
0
0

Thanks. Now it’s getting more clear, but I still feel uncomfortable about this "line of sight" thing.

Let’s assume the same room scenario again. The sound source is straight in front of the listener, but there’s a soundproof wall in between and a doorway on the left. The direct pass is obviously completely occluded, but sound should still arrive to the listener from the left.
As I see, changing the wall parameters near to the door would not solve this case, since only the direct path is accounted for.

Can this be done with FMOD?

Greg

  • You must to post comments
0
0

[quote:dbzfdxrc]but sound should still arrive to the listener from the left. [/quote:dbzfdxrc]
Correct, you would not expect the sounds to be completely occluded. So near the door you might have 90% direct occlusion and 50% reverb occlusion. Whereas behind the back wall you wouldn’t expect to hear much at all so you might have 100% direct occlusion and 90% reverb occlusion.

[quote:dbzfdxrc]As I see, changing the wall parameters near to the door would not solve this case, since only the direct path is accounted for. [/quote:dbzfdxrc]
If the listener is near the door, the direct path would pass through the part of the wall near the door. So the parameters for that part of the wall are all that are considered.

-Pete

  • You must to post comments
0
0

I think he’s really asking "Does the sounds bounce off walls" like in real life…

  • You must to post comments
0
0

[quote:3rc1stip]I think he’s really asking "Does the sounds bounce off walls" like in real life…[/quote:3rc1stip]

No it wont. It’s not feasible on current hardware to calculate the reverb/occlusion like that for many sound sources.

-Pete

  • You must to post comments
Showing 5 results
Your Answer

Please first to submit.