ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: Meaningful Way to Characterize Reverb?  (Read 4547 times)

Dan Mortensen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 835
Meaningful Way to Characterize Reverb?
« on: May 03, 2012, 03:09:25 pm »

Hi Everyone,

Is there a meaningful way to visually represent the characteristics of one reverb vs. another?

I'm working on a project that will examine CBS' 30th St. Studio (torn down for many years), and the excellence of its reverb chamber (emulated contemporaneously at Abbey Road and many other places).

The thought arose to compare a live chamber to a digital one (we have access to a small chamber that a guy built in his basement recording studio, which he has tried to optimize to sound like the 30th St. one), and then I thought of all the hours I've spent trying to get a believable sound out of digital reverbs and am wondering if there's a way to create some kind of picture that I can see what I like, so then there's a goal to shoot for, and thus a way to try to replicate it.

A waterfall plot jumped to mind, but I have a SIM which does not have that feature, and am at a loss to know where to go next.

So I guess that's several questions:

1) Is there a particular measurement system that will accurately characterize the parameters of interest in a reverb device which you have access to?

2) Would matching two visual representations from two devices result in matching reverb sounds?

3) If all you have is previously recorded work, and thus have no way to put a specific signal through the chamber, is there a way to characterize the reverb (as discussed above) contained in the recording?

Thanks for the help,
Dan

Logged

Thomas Harkin

  • Full Member
  • ***
  • Offline Offline
  • Posts: 103
  • Oklahoma City, Oklahoma, USA
Re: Meaningful Way to Characterize Reverb?
« Reply #1 on: May 03, 2012, 03:33:25 pm »

Hi Everyone,

Is there a meaningful way to visually represent the characteristics of one reverb vs. another?

So I guess that's several questions:

1) Is there a particular measurement system that will accurately characterize the parameters of interest in a reverb device which you have access to?

2) Would matching two visual representations from two devices result in matching reverb sounds?

3) If all you have is previously recorded work, and thus have no way to put a specific signal through the chamber, is there a way to characterize the reverb (as discussed above) contained in the recording?

Thanks for the help,
Dan

Is this what you are talking about?

http://en.wikipedia.org/wiki/Convolution_reverb

There are a number of programs out there.  I'll try to find out which one a friend of mine used to recreate an outdoor concert space that was video'd.  He didn't have enough inputs for ambient mics, so he recreated the space in the computer.
Logged

Dan Mortensen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 835
Re: Meaningful Way to Characterize Reverb?
« Reply #2 on: May 03, 2012, 04:31:28 pm »

Is this what you are talking about?

http://en.wikipedia.org/wiki/Convolution_reverb

There are a number of programs out there.  I'll try to find out which one a friend of mine used to recreate an outdoor concert space that was video'd.  He didn't have enough inputs for ambient mics, so he recreated the space in the computer.

Hi Thomas,

Thanks for the response.

That isn't quite what I'm looking for, or at least isn't what I had in mind when asking the questions. And in thinking about it for a minute, I think it actually does answer the questions, but in a way that will be less directly useful to me.

What I was thinking with question #2 was taking a reverb device, like an SPX-2000 or something, which has a reverb that I like the qualities of, and generating some sort of visual representation of its performance in the parameters that influence what I hear and like in a reverb, and then using that representation as a goal to adjust a TC M-OneXL (or whatever) through the measurement system to get a sound from the M-OneXL that matches.

With the above scenario, you can use whatever test signal is appropriate. Question #3 wonders if there is a way to characterize a reverb within a recording (used as the test signal), to then be able to approximate it using the methodology above.

The convolution reverb device would be most capable of taking data from thousands or millions of points to match, but then that device becomes your reverb device. I'm wondering if there is something similar to a waterfall plot that would be useful for this purpose, so that the random multi-effect processor can be intelligently adjusted to approximate a desired spectral decay characteristic.

Thanks,
Dan



Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9077
  • Atlanta GA
Re: Meaningful Way to Characterize Reverb?
« Reply #3 on: May 03, 2012, 04:37:16 pm »

Hi Everyone,

Is there a meaningful way to visually represent the characteristics of one reverb vs. another?

I'm working on a project that will examine CBS' 30th St. Studio (torn down for many years), and the excellence of its reverb chamber (emulated contemporaneously at Abbey Road and many other places).

The thought arose to compare a live chamber to a digital one (we have access to a small chamber that a guy built in his basement recording studio, which he has tried to optimize to sound like the 30th St. one), and then I thought of all the hours I've spent trying to get a believable sound out of digital reverbs and am wondering if there's a way to create some kind of picture that I can see what I like, so then there's a goal to shoot for, and thus a way to try to replicate it.

A waterfall plot jumped to mind, but I have a SIM which does not have that feature, and am at a loss to know where to go next.

So I guess that's several questions:

1) Is there a particular measurement system that will accurately characterize the parameters of interest in a reverb device which you have access to?

2) Would matching two visual representations from two devices result in matching reverb sounds?

3) If all you have is previously recorded work, and thus have no way to put a specific signal through the chamber, is there a way to characterize the reverb (as discussed above) contained in the recording?

Thanks for the help,
Dan
An impulse response would be the best way.

But there is a lot more to the "sound" of reverb than just time.  The initial time gap is a big thing that makes you think you are in a small or large room.  Yes you can have a long reverb time in a small(ish) room, and it not sound like a large room.  The HF rolloff is another big clue that our ears give us as to the size of a particular room. 
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 16643
  • Hickory, Mississippi, USA
    • Resotune
Re: Meaningful Way to Characterize Reverb?
« Reply #4 on: May 03, 2012, 04:41:28 pm »

Perhaps an RT60 for different frequency ranges would give a crude characterization.

JR
Logged
When in doubt do what's right.

Thomas Harkin

  • Full Member
  • ***
  • Offline Offline
  • Posts: 103
  • Oklahoma City, Oklahoma, USA
Re: Meaningful Way to Characterize Reverb?
« Reply #5 on: May 03, 2012, 05:14:57 pm »

Hi Thomas,

Thanks for the response.

That isn't quite what I'm looking for, or at least isn't what I had in mind when asking the questions. And in thinking about it for a minute, I think it actually does answer the questions, but in a way that will be less directly useful to me.

What I was thinking with question #2 was taking a reverb device, like an SPX-2000 or something, which has a reverb that I like the qualities of, and generating some sort of visual representation of its performance in the parameters that influence what I hear and like in a reverb, and then using that representation as a goal to adjust a TC M-OneXL (or whatever) through the measurement system to get a sound from the M-OneXL that matches.

With the above scenario, you can use whatever test signal is appropriate. Question #3 wonders if there is a way to characterize a reverb within a recording (used as the test signal), to then be able to approximate it using the methodology above.

The convolution reverb device would be most capable of taking data from thousands or millions of points to match, but then that device becomes your reverb device. I'm wondering if there is something similar to a waterfall plot that would be useful for this purpose, so that the random multi-effect processor can be intelligently adjusted to approximate a desired spectral decay characteristic.

Thanks,
Dan

Dan,

Copied from that wiki:
Machine simulation

It is also possible to sample the impulse response of a reverberation unit, instead of sampling a real space. Thus, it is possible to use a convolution reverb in place of a hardware machine. The techniques used to sample a reverberation unit are the same as the ones used to sample real spaces.

These people might be able to help.  http://www.audioease.com/Pages/Altiverb/
Logged

Dan Mortensen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 835
Re: Meaningful Way to Characterize Reverb?
« Reply #6 on: May 03, 2012, 11:05:49 pm »

Dan,

Copied from that wiki:
Machine simulation

It is also possible to sample the impulse response of a reverberation unit, instead of sampling a real space. Thus, it is possible to use a convolution reverb in place of a hardware machine. The techniques used to sample a reverberation unit are the same as the ones used to sample real spaces.

These people might be able to help.  http://www.audioease.com/Pages/Altiverb/

Hi Thomas,

This thing is quite a gizmo! The videos are quite impressive, and the modeling that it does is obviously superior to what a human could do with eye/ear. The video showing how you can take samples of a space to later recreate within the gizmo is pretty cool:

http://www.audioease.com/Pages/Altiverb/sampling.php

At 5:01 this waterfall appears, with what I assume are frequency indicators hidden by the demonstrator's picture (below).

I'm assuming that Time Zero is the rearmost plot, with later frequency plots appearing closer to the viewer until complete decay. Are the white dividers indicators of seconds?

This is a software product; the best I can figure is that it is a ProTools plug-in, but they say somewhere that it works in other platforms as well. I am not knowledgeable enough to know how one would use this in a live sound application with a non-digital console (can it be used with digital consoles?).

I have a pending purchase of another product with a local Altiverb dealer, and will ask them these kinds of questions. I'm sure they'd be happy to sell me something else.

Thanks again for your help,
Dan

Logged

Iain.Macdonald

  • Full Member
  • ***
  • Offline Offline
  • Posts: 213
Re: Meaningful Way to Characterize Reverb?
« Reply #7 on: May 04, 2012, 02:06:19 pm »

 8)
« Last Edit: May 05, 2012, 12:04:18 pm by Lost »
Logged

ProSoundWeb Community

Re: Meaningful Way to Characterize Reverb?
« Reply #7 on: May 04, 2012, 02:06:19 pm »


Pages: [1]   Go Up
 



Page created in 0.039 seconds with 23 queries.