Automatically balancing volume over time

[tomu]

I have a recording of a dicsussion class in which the microphed leader's voice is well recorded but everyone else's voice is quite. Audible, but quiet. Is there any tool which will automatically increase the volume of these quite sections?

Thanks,

Tom

[Gerard Bik]

I have a recording of a dicsussion class in which the microphed leader's voice is well recorded but everyone else's voice is quite. Audible, but quiet. Is there any tool which will automatically increase the volume of these quite sections?

Thanks,

Tom

------------------------

What you want is one of the compressor/limiter plug ins.
When you search the forum you will find more discussions of this problem.
Personally I rather correct these things by hand with help of the actions palette.

--
______________________ _
Gerard Bik grafische vormgeving
Van Aerssenstraat 263
2582 JM Den Haag
070 3554081
_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

[tomu]

What you want is one of the compressor/limiter plug ins.

Which one, what settings? I have searched a little bit and got some help. I was not able to get the AUDynamicsProcessor to expand quiet sections. I can get it to clip out the really loud bursts, which is very helpful.

I suspect automatic processing will have undesired results of amplifying undesired quite background noise. Difficult to distinguish between quite voices that I want amplified and other quite noise that I do not. But this is a common tasks with many sections to amplify and too time consuming to do by hand.

We are working on getting better recordings where there is less difference between volume of speakers but there will always be some quite speakers that need amplification.

Tom

[CDJonah_alt]

I like to use a compressor on any speech. I have used the mda dynamics
compressor with default settings but that is just because I started
before au plug ins existed.

I would probably:
1 normalize the entire track
2 select a quiet section, run statistics on it.
3 amplify appropriately (use the dB settings for ease) where I would use
an appropriate (.5 to 1 sec ramp up/down of the amplification.
4 use the dynamics plugin (take your choice) to smooth things out
5 Call the result good enough.

Chuck

tomu wrote:

What you want is one of the compressor/limiter plug ins.

Which one, what settings? I have searched a little bit and got some help. I was not able to get the AUDynamicsProcessor to expand quiet sections. I can get it to clip out the really loud bursts, which is very helpful.

I suspect automatic processing will have undesired results of amplifying undesired quite background noise. Difficult to distinguish between quite voices that I want amplified and other quite noise that I do not. But this is a common tasks with many sections to amplify and too time consuming to do by hand.

We are working on getting better recordings where there is less difference between volume of speakers but there will always be some quite speakers that need amplification.

Tom

------------------------
- Tom




_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

--
Charles D. Jonah CDJonah@anl.gov
630-252-3471
Chemistry Division
Argonne National Laboratory
Argonne, IL 60439



_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

[Sonic Purity]

I suspect automatic processing will have undesired results of amplifying undesired quite background noise. Difficult to distinguish between quite voices that I want amplified and other quite noise that I do not. But this is a common tasks with many sections to amplify and too time consuming to do by hand.

I would imagine that the goal is intelligibility… this sort of arrangement is not going to win high fidelity awards. As such, in addition to aggressive compression and similar items already discussed, i’d add bandwidth limiting: filter out the frequencies above and below speech frequencies, where there will mostly be unwanted noise. The result may sound “low-fi” due to the missing frequencies, yet should be easier to understand what the people are saying, which i would guess is the goal.

Not sure where in the processing workflow would be best to “narrowband” the sound… this is something i know about yet have not done much… others here may know from experience where the optimal place to do this is. If nothing else, make copies of the unprocessed original and experiment with different workflows.

As long as no one is expecting a work of sonic brilliance, it should be quite possible to set something up within Amadeus to do this automatically.