View Single Post
Old 05-30-2011, 05:24 PM   #781
ares's Avatar
Join Date: Sep 2003
Location: Chapel Hill, NC
Posts: 197
Originally Posted by Moseph View Post
Hold up, what about the Sabine Equation (as referenced in my copy of Kinsler*)? That's not a linear description. I was always under the impression that the "relatively" linear behavior was because most descriptions knocked out the higher-order behaviors on purpose to give a general idea of what the room would do (i.e., the non-linear stuff is actually important).

*Fundamentals of Acoustics, 4th Edition. Kinsler, et al. Wiley: 2004. My understanding is that this is the standard intro survey text (in one form or another), and that this has been the case for roughly 50 years or so.
I'm familiar with this equation, I actually use a modified version (Eyring) to calculate the reverb times in my simulation system.

The point that I'm getting at is that in most cases, the response of a room/source/listener configuration can be described by an impulse response: convolve this IR with source audio of any amplitude and you're going to get the same result times the input audio gain coefficient. The response is linear with respect to the energy of the sound source in most cases.

There are certain non-linearities w.r.t. energy that deal with how the sound interacts with air as it travels as well the interactions with materials in the room, but these are probably so small that they wouldn't have any effect when mixing at different volumes.

Originally Posted by Moseph View Post
I'd be inclined to give him a little more credit than that. That's all stuff that gets considered in the first week of reverberation calculation. I was worrying about that stuff in my Acoustics class and we only talked about it for about 3 class days (as would be expected in a general overview of acoustic phenomenon).
Yes, my simulation system can use arbitrary geometry and every surface has material properties that specify the absorption at different frequencies. If you'd like to read about it, here is a paper I presented earlier this year at the 41st AES conference in London:
ares is offline