A

Anonymous

Answered

Removing slight echo with audtion - recorded in room with hard walls

Forum|Forum|16 years ago
May 28, 2009
8 replies
166852 views

Hi

I just downloaded a trial of Audition CS4 for windows.

I have already used it to remove a few clicks which are evident in this short (attached) PRE-audition sample.

Now I would like to go in and remove (if possible) or at least reduce the slight "echo" of the room itself. You can hear the "echo" in this attached wav file.

I realize I might not be able to remove all of the "echo" but if I could remove some of it it would be nice.

Thanks

Rowby

audio-test.zip

This topic has been closed for replies.

Correct answer SteveG_AudioMasters_

rowby wrote:
I just downloaded a trial of Audition CS4 for windows.

No such animal - Audition is emphatically not part of the CS suite. It used to be, and that was a right pain. Fortunately for all of us it was removed from it a while back, after AA2.0 was prematurely released. Since this was a direct result of a release date determined only by marketing, it was hardly surprising, really...

Now I would like to go in and remove (if possible) or at least reduce the slight "echo" of the room itself. You can hear the "echo" in this attached wav file.
I realize I might not be able to remove all of the "echo" but if I could remove some of it it would be nice.

Sorry, no chance. It's like trying to unbake a cake; you simply can't do it. Especially in this case where it's a short reverb, and it inevitably contains only the same frequencies that you want to keep. (This is a reminder to anybody about to suggest EQ, like they usually do, that it can only possibly make things worse, like it always does...).

You are stuck completely with this - no software on earth can get rid of that sort of problem. You have to take rather more care with the recording in the first place to prevent this, I'm afraid. That's why people use treated studios for anything serious - just to get over problems of this nature.

The only exception to this in any way is that if you have a stereo recording where the echo is in the stereo field rather than the summed mono one, you can use the Audition Center Channel Extractor, and effectively extract just the wanted mono part. But with a mono recording (which contains no vector information at all) you simply can't do this.

E

ExpoToddC

Participant

Lol. I just thought SteveG might like to know that his helpful reputation has grown beyond the mere confines of Adobe Forums. Removing Echo from Audio with Adobe Audition

C

CRGreenDotCom

Participating Frequently

http://dl.dropboxusercontent.com/u/3675908/echo-removal-test2.wav

R

robertchapin

Participant

As a music major, I was introduced to an amazing workshop on room tone. The presentation was a simple voice recording that was played in a room with a slight reverb. The recording was played over and over again, each time recorded and played back in the same room. After about 100 iterations, the voice became music - a series of overtones. The room absorbed certain vibrations and reflected others back. Seems to me, an advanced custom filter could detect those overtones to cancel them out. This is why some problems can be minimized with a simple EQ - the reverb of every room has a different resonance, and those frequencies don't always contain the full spectrum.

SteveG_AudioMasters_

Community Expert

Rooms do not reflect back anything that isn't present in the stimulus signal - which means that whatever you cancel out also comes out of that signal, which is exactly where your problem with this approach lies.

The only system I'm aware of that works on signals like this (and you don't need 100 re-recordings to achieve it) is a feedback killer. Feedback in a room with a speaker and a mic is predominently based on the greater reflection of some frequencies than others, and this builds up very quickly. Feedback killers work by putting a small real-time pitch shift on the feed to the loudspeaker, so direct reinforcement doesn't occur. And the loudspeaker is the other significant contaminent, of course - possibly more than the room in some situations. And the response errors in this (both time and frequency-dependant) build up significantly, being reinforced each time around...

So what your workshop actually showed you wasn't really about the room, I'm afraid - it was about the system being used in it, and the room was only incidental. This is because the room, unlike the speaker, won't cause frequency distortion, only temporal disturbances.

Experiments/demonstrations like this very rarely attempt to control all the variables - and in this case it may well be that they didn't even explain them all to you, I'm afraid. How do I know? Rooms and reverberation were a significant part of my Acousics Master's thesis...

R

robertchapin

Participant

Correct. The room was reflecting back the stimulus signal, but only frequencies which resonated with the room, which also included overtones. So it was like music. If it was pure reflection, it would've just been noise. So the room was acting like a sounding board - like a guitar or any other instrument. Really cool presentation. Sorry if you don't think so.

And I should've known to post in a thread where there are so many people willing to jump in immediately and say "no, you're wrong" without even thinking. Oh, wait, this is the same guy who was the first to say "it can't be done"? Should've guessed. Dude, looking back at every single one of your posts, there is nothing that you say is even remotely positive.

I'm no "audio master", but I know that every room has its own acoustic property, regardless of what system is being used, which resonates based on specific frequencies. Has nothing to do whether it's an amplifier or a human voice. And I'm sure there's plenty of online sources and peer-reviewed journals to back that up. If you want to argue anything further, you might want to cite sources other than yourself. And you might want to talk to people instead of talking down to them. They tend to listen better and believe in what you say as opposed to thinking that you're just a bitter know-it-all.

J

Joe12south

Known Participant

This plug-in does the so-called "impossible" quite well: http://www.zynaptiq.com/unveil/

It's pricey, but if it saves even one session, it can be worth it.

SteveG_AudioMasters_

Community Expert

Joe12south wrote:
This plug-in does the so-called "impossible" quite well: http://www.zynaptiq.com/unveil/
It's pricey, but if it saves even one session, it can be worth it.

Yeah, we did that one a while back in this thread... and it's not that good.

J

Joe12south

Known Participant

It's not magic in the way Jesus is, but it does work well in a lot of cases. I just hate when people say something is carte blanche "impossible."

A

Anonymous

For what it's worth, here's a suggestion made on another forum

"You can try using Effect -> Reverb -> Convolution reverb... to remove echo: Create a new track using the convolution reverb effect and play it against the old track, with phase inverted (search help for Invert a waveform about that).

What settings you will use and if this will be successful or not, depends on the settings that you will enter at Convolution reverb filter and if you're lucky enough to match the echo. There is no success guarantee!"

Any thoughts about this approach?

Rowby

AskMrScience

Participant

Adding reverb in an attempt to remove it? You'd be adding a new layer of reverb to the existing unwanted reverberance was well as to the signal that you're trying to isolate. It quickly becomes a hall of mirrors - and phase shifts.

A

Anonymous

I deal with this all the time

use the Graphic EQ

you can't kill everything but a great deal can be removed just by adjusting the bands

here's the settings I used with just a quick pass

and I've posted the modified file I reversed the name to avoid overwriting the original

load them both and compare

EQ is fine for speech audio but not the best idea for trying to fix music

Message was edited by: Richard FDisk

test-audio.zip

SteveG_AudioMasters_

Community Expert

Richard FDisk wrote:
I deal with this all the time
use the Graphic EQ
you can't kill everything but a great deal can be removed just by adjusting the bands

You clearly didn't read or understand my first post. You have killed nothing, and removed signal at the same rate as reverb - which makes it harder to hear the content and almost certainly makes things worse, not better. This is a complete waste of time.

*update* I've now listened to what you did, and it conforms absolutely to what I just said - you've made the intelligability significantly worse, not better, and all the room sound is still there.

THERE IS NO FIX FOR THIS.

A

Anonymous

why do you attack me?

► it actually turned out not too bad oy my system,

► the "echo" is just re-amplification & recording of certain wave lengths that happen to make it back to the mic, by reducing those signals it's less "tinny" or "echoey"

► if EQ is a useless tool why include it in AA?

► the OP asked to reduce the problem not remove it completely,

► yes the "room sounds" are still there, but not as distracting or "IRRITATING" as when "untouched"

► the average joe listening to any recording doesn't listen with an Audio Engineer's Ear, why do you think that mp3's and other forms of lossy compressed garbage are so popular?

► yes you can't "Unbake" a cake, but you can make a bad cake more paletable with a little bit of "bakers touch"

► sorry I tried to help

Message was edited by: Richard FDisk

AskMrScience

Participant

For what it's worth, there are deconvolution techniques that are effective at echo removal, but they mostly apply to simple cases, such as telephony, where the echo is a simple time-delayed copy of the original signal. Such methods applied to natural reverberance most often make things worse. For now, and probably well into the future, there's no good substitute for controlling the recording environment.

SteveG_AudioMasters_

Community Expert

AskMrScience wrote:
For what it's worth, there are deconvolution techniques that are effective at echo removal, but they mostly apply to simple cases, such as telephony, where the echo is a simple time-delayed copy of the original signal.

I've altered the reply above to say 'reverb' - as what's in the sample doesn't begin to qualify as an echo. It's what I meant to say in the first place, but it got overlooked in the reply process.

SteveG_AudioMasters_

Correct answer

Community Expert

rowby wrote:
I just downloaded a trial of Audition CS4 for windows.

No such animal - Audition is emphatically not part of the CS suite. It used to be, and that was a right pain. Fortunately for all of us it was removed from it a while back, after AA2.0 was prematurely released. Since this was a direct result of a release date determined only by marketing, it was hardly surprising, really...

Now I would like to go in and remove (if possible) or at least reduce the slight "echo" of the room itself. You can hear the "echo" in this attached wav file.
I realize I might not be able to remove all of the "echo" but if I could remove some of it it would be nice.

Sorry, no chance. It's like trying to unbake a cake; you simply can't do it. Especially in this case where it's a short reverb, and it inevitably contains only the same frequencies that you want to keep. (This is a reminder to anybody about to suggest EQ, like they usually do, that it can only possibly make things worse, like it always does...).

You are stuck completely with this - no software on earth can get rid of that sort of problem. You have to take rather more care with the recording in the first place to prevent this, I'm afraid. That's why people use treated studios for anything serious - just to get over problems of this nature.

The only exception to this in any way is that if you have a stereo recording where the echo is in the stereo field rather than the summed mono one, you can use the Audition Center Channel Extractor, and effectively extract just the wanted mono part. But with a mono recording (which contains no vector information at all) you simply can't do this.

P

phamtruojng

Participant

thank for share

Sign up

To post, reply, or follow discussions, please sign in with your Adobe ID.

Sign in to Adobe Community

To post, reply, or follow discussions, please sign in with your Adobe ID.

Scanning file for viruses.

This file cannot be downloaded