to make a calculation of time reverb?
hi comunity! i would like to calculate the reverberation time of my house with max msp. is it possible with max msp and a microphone for example????
If someone know a patch or tutorial please share!! (^_^)
thanxxxx!!!
Hi alersito
When you say 'house' do you mean the realworld reverberation times of each room/one room, or the theoretical total reverb time of your entire house? The answers also depend on whether you need a mathematical algorithm for recreating the realworld room, or a system for capturing the rooms' reverb. If it is the former, you first need a measuring tape. Then you should search for 'Sabine's formula', which states that:
Tr = (Kv)/Se
where 'Tr' is total reverb time, 'v' is the volume of the room in meters cubed, 'Se' is the effective absorbtion area (m2) and 'K' is a constant (0.16). 'Se' itself is dependent on a number of factors, which can easily be determined if you search for 'absorbtion coefficient'.
HTH
Brendan
thankss n00b_meister i want to say the realworld reverberation time. playing a pink noise and then calculating the reverberation time.
OK, I think I understand; by 'calculate' you mean 'capture', like an impulse response? Sending a short burst of noise into the room via loudspeaker and then capturing the reverberations? Most sources recommend using a 0db burst of white noise for the duration of 1 sample.
The practical application of this is beyond me I'm afraid, but here's an article I found via google, on convolution reverb, which may or may not help:
Brendan
thanks brendan.
i think that could be possible calculate the Tr using a source with pink noise and 100 dB. when the source stop send a bang to a clock and when the level decrease 60dB send another bang and stimate the time between the 2 bangs.
i dont know if this is ok or not. (^_^)
Here is a patch which does it.
It calculates the RT30 and multiplies the duration because in most situations, you may not have a 60dB. range between a hand clap and the silence. In my case hand clap is necessary.
It has a few issues such as getting stuck sometimes, so if you find a better way, please share it.
RT60 is not dbFS, it is always relative to the source event. so just record anything and measure both, the source and the tail.
pink noise seems appropiate, but it should work likewise with a cello or a clap.
and if you are paranoid:
1.) dont forget to add what has been lost due to the distance between the microphone and the cello
2.) use at least 4 capsules (neumann sm or a tetrahedral ambisonics mic)
3.) use different positions in the room and approximate the values and/or collect the data in a list