[LAU] Normalizing Audio Levels using RMS

[Justin Smith]

On Wed, Apr 15, 2009 at 12:24 PM,  <hollunder at gmx.at> wrote:
> On Wed, 15 Apr 2009 10:37:29 +0200
> Raffaele <raffaele.morelli at gmail.com> wrote:
>
>> 2009/4/14 Viktor Mastoridis <viktor at mastoridis.co.uk>
>>
>> >
>> > Hi Linux Audio Geeks
>> >
>> > In my musical prehistory, while I was on Windows, I used to use a
>> > program called SoundForge that had one very useful feature:
>> > normalizing audio levels with RMS, even using the Equal Loudness
>> > Contour
>> >
>> > For a whole year I am struggling now to find something similar on
>> > Linux, without much success.
>> >
>> > Any help/hints will be much appreciated.
>> >
>> > Viktor
>> >
>>
>> Hi,
>>
>> if the the goal is to have a sound file play as loud as possible I
>> would use Jamin, so you can take care about frequencies, if needed.
>>
>> If you are going to normalize a telephone conversation maybe you
>> don't need such accuracy.
>>
>> regards
>> -r
>
> I don't really see how RMS can be that useful, at least not when you
> try to get stuff as loud as possible. You still have to watch for peaks
> because of clipping.
>
> If your goal is to have songs or albums at the same loudness for
> playback use replaygain. It's probably not perfect but the best thing I
> know of. I think it uses something like RMS and something to adjust it
> to the human hearing, but it also has an option to not clip the signal.
>
> for more information check http://replaygain.org/ and
> http://hydrogenaudio.org.
>
> Best,
> Philipp
> _______________________________________________
> Linux-audio-user mailing list
> Linux-audio-user at lists.linuxaudio.org
> http://lists.linuxaudio.org/mailman/listinfo/linux-audio-user
>

According to replaygain's website, they do a weighted filter to
compensate for unequal loudness perception, then use the calculated
RMS power to determine the appropriate gain for replay. So this looks
like it does exactly what the O.P. asked for.