Subject: Re: Software for loudness equalization? From: Matthias Milczynski <matthias.milczynski@xxxxxxxx> Date: Tue, 10 Apr 2012 19:31:55 +0200 List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>--f46d044287a6d1ccee04bd567f2a Content-Type: text/plain; charset=ISO-8859-1 Dear Dave, you can try replaygain (http://replaygain.hydrogenaudio.org/proposal/), this web-page includes Matlab-code, I think they use a pink-noise presented at 83 dB as a reference, replaygain is used to achieve an equal loudness percept for e.g. mp3-recordings. If you have access to adobe audition you can use the "group waveform normalize" function. Both, replaygain and adobe use (approximations to) equal-loudness contours, in adobe this is optional. Hope this helps Regards, MM 2012/4/10 Dave Benson <davehbenson@xxxxxxxx> > Dear list, > > I have a largish (600 item) collection of short musical sound files that > I'd like to equalize in loudness. Specifically, I'd like to scale each file > so that the peak of its loudness function is equal to a particular value, > in sones, that I specify. Is anyone aware of a software package that can > perform this sort of equalization? > > I've found several implementations of loudness models online (see below) > but they all seem designed to solve a slightly different problem. Namely, > they take audio files as input and produce loudness values as output. I'm > interested in the inverse operation: taking a target loudness (and a set of > unscaled files) as input, and producing set of scaled audio files as output. > > I'm aware that this problem could be solved using an iterative technique > (i.e., by repeated adjusting each file's scaling factor until its loudness > is close to the target), and writing a script to do this would be > relatively straightforward. Before writing it, though, I wanted to make > sure that a similar script (or perhaps a more elegant solution to the > problem) hadn't already been written by someone else in the community. > > Again, is anyone aware of an existing software package for equalizing a > set of audio files in loudness? > > Many thanks for your help, > > > Dave Benson > PhD Candidate, Sound Recording department > Schulich School of Music of McGill University > > > * The loudness model implementations I'm referring to are > - command line utilities from Prof. Brian Moore's lab at the University of > Cambridge (http://hearing.psychol.cam.ac.uk/Demos/demos.html) > - Psysound3 (http://code.google.com/p/psysound3/) > - the Loudness Toolbox from Genesis Acoustics ( > http://www.genesis-acoustics.com/index.php?page=32) > --f46d044287a6d1ccee04bd567f2a Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Dear Dave, <br>you can try replaygain (<a href=3D"http://replaygain.hydroge= naudio.org/proposal/">http://replaygain.hydrogenaudio.org/proposal/</a>), t= his web-page includes Matlab-code, I think they use a pink-noise presented = at 83 dB as a reference, replaygain is used to achieve an equal loudness pe= rcept for e.g. mp3-recordings. If you have access to adobe audition you can= use the "group waveform normalize" function. Both, replaygain an= d adobe use (approximations to) equal-loudness contours, in adobe this is o= ptional.<br> Hope this helps<br>Regards,<br>MM=A0 <br><br><div class=3D"gmail_quote">201= 2/4/10 Dave Benson <span dir=3D"ltr"><<a href=3D"mailto:davehbenson@xxxxxxxx= l.com">davehbenson@xxxxxxxx</a>></span><br><blockquote class=3D"gmail_q= uote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1e= x"> <span>Dear list,</span><br><br><span>I have a largish (600 item) collection= of short musical sound files that I'd like to equalize in loudness. Sp= ecifically, I'd like to scale each file so that the peak of its loudnes= s function is equal to a particular value, in sones, that I specify.=A0 Is = anyone aware of a software package that can perform this sort of equalizati= on?</span><br> <br><span>I've found several implementations of loudness models online = (see below) but they all seem designed to solve a slightly different proble= m. Namely, they take audio files as input and produce loudness values as ou= tput.=A0 I'm interested in the inverse operation: taking a target loudn= ess (and a set of unscaled files) as input, and producing set of scaled aud= io files as output.</span><br> <br><span>I'm aware that this problem could be solved using an iterativ= e technique (i.e., by repeated adjusting each file's scaling factor unt= il its loudness is close to the target), and writing a script to do this wo= uld be relatively straightforward.=A0 Before writing it, though, I wanted t= o make sure that a similar script (or perhaps a more elegant solution to th= e problem) hadn't already been written by someone else in the community= . =A0</span><br> <br><span>Again, is anyone aware of an existing software package for equali= zing a set of audio files in loudness?</span><div><br></div><div><span>Many= thanks for your help,</span><br><br><br> <span>Dave Benson</span><br><span>PhD Candidate, Sound Recording department= </span><br><span>Schulich School of Music of McGill University</span><br><b= r><br><span>* The loudness model implementations I'm referring to are</= span><div> - command line utilities from Prof. Brian Moore's lab at the University= of Cambridge (<a href=3D"http://hearing.psychol.cam.ac.uk/Demos/demos.html= " style=3D"color:rgb(17,85,204)" target=3D"_blank">http://hearing.psychol.c= am.ac.uk/Demos/demos.html</a>)</div> <div>-=A0Psysound3 (<a href=3D"http://code.google.com/p/psysound3/" style= =3D"color:rgb(17,85,204)" target=3D"_blank">http://code.google.com/p/psysou= nd3/</a>)=A0</div><div>- the Loudness Toolbox from Genesis Acoustics (<a hr= ef=3D"http://www.genesis-acoustics.com/index.php?page=3D32" style=3D"color:= rgb(17,85,204)" target=3D"_blank">http://www.genesis-acoustics.com/index.ph= p?page=3D32</a>)</div> </div> </blockquote></div><br> --f46d044287a6d1ccee04bd567f2a--