Audacity Forum

Hi there
I'm creating a medley from 8 different songs. I have already faded them in and out and into each other, but is there a way to AUTOMATICALLY have them all at the same volume?
I understand that I can do this manually by simply turning the gain up and down on each track, but I'm looking for a faster solution.

Can audacity analyze the entire song and take a mean volume and apply it to all tracks so that it wont ever go past that volume?
Or something similar?

Thanks

Shippo wrote:Can audacity analyze the entire song and take a mean volume and apply it to all tracks so that it wont ever go past that volume?
Or something similar?

Not yet, though I did have a go at making such a tool.
If you are particularly interested in there being an effect for this and are willing to do some testing I'll probably have some time next week to re look at this project.

With only 8 songs this is not too hard to do by hand with Audacity.

1) decide which level you want to amplify to - I would suggest a maximun of -2dB

2) select each track in turn - if you have labelled the tracks making this selction shoub be easy (if you have point labels zoom out to fit the project to the screen and then select between the label pairs, the label points are "stick" to help you with this. If you have range labels for each song you should just be avle to clivk in the range label to select the labelled range).

3) use the Amplify effect on each track in turn to bring it to your required taget level

job done ...

But with more than a few tracks this would quicly become tedious

WC

It's usually best to do this manually by ear. Most audio editing requires listening and human judgement...

I haven't tried equal-loudness.ny, but that's the best approach if you want to do it automatically. If that doesn't work, WaveGain* and MP3Gain try to standardize the perceived loudness. The idea is to WaveGain the tracks individually, and then create the medley from the adjusted tracks. (ReplayGain and SoundCheck are similar, but they won't work because they change the volume during layback, instead of "permanently" changing the file.)

hen after creating the Medley, you may need to boost the gain with Amplify, since these tools tend to reduce volumes.

waxcylinder wrote:With only 8 songs this is not too hard to do by hand with Audacity.

1) decide which level you want to amplify to - I would suggest a maximun of -2dB

2) select each track in turn...

3) use the Amplify effect on each track in turn to bring it to your required taget level..l

That will set the peak level, but it won't necessarily make the parts sound equally loud. Perceived loudness correlates better to the average level, with the equal loudness curves applied (to compensate for the ear's different sensitivity to different frequencies).


* As you may know, it's always best to work with uncompressed formats such as WAV. If you want/need MP3 or another lossy format, you should compresson once as the final step after all editing.

DVDdoug wrote:I haven't tried equal-loudness.ny, but that's the best approach if you want to do it automatically. If that doesn't work, WaveGain* and MP3Gain try to standardize the perceived loudness. The idea is to WaveGain the tracks individually, and then create the medley from the adjusted tracks. (ReplayGain and SoundCheck are similar, but they won't work because they change the volume during layback, instead of "permanently" changing the file.)

The idea of equal-loudness.ny is similar to ReplayGain and SoundCheck as a Nyquist plug-in, but to apply (amplify) the audio to a standardised loudness. Unfortunately it doesn't work (yet). When standardising the loudness of multiple tracks it is necessary that some tracks will have peaks below 0 dB. If amplified after adjusting the tracks to a standardized level, all of the tracks must be amplified by exactly the same amount (the same number of dB) so that the relative loudness of each track to each other track remains the same. The difficult part of implementing this in Nyquist is the statistical analysis (as described here: http://wiki.hydrogenaudio.org/index.php ... cification ). I'd welcome help with that if anyone want to volunteer.

DVDdoug wrote:It's usually best to do this manually by ear. Most audio editing requires listening and human judgement...

I haven't tried equal-loudness.ny, but that's the best approach if you want to do it automatically. If that doesn't work, WaveGain* and MP3Gain try to standardize the perceived loudness. The idea is to WaveGain the tracks individually, and then create the medley from the adjusted tracks. (ReplayGain and SoundCheck are similar, but they won't work because they change the volume during layback, instead of "permanently" changing the file.)

hen after creating the Medley, you may need to boost the gain with Amplify, since these tools tend to reduce volumes.

waxcylinder wrote:With only 8 songs this is not too hard to do by hand with Audacity.

1) decide which level you want to amplify to - I would suggest a maximun of -2dB

2) select each track in turn...

3) use the Amplify effect on each track in turn to bring it to your required taget level..l

That will set the peak level, but it won't necessarily make the parts sound equally loud. Perceived loudness correlates better to the average level, with the equal loudness curves applied (to compensate for the ear's different sensitivity to different frequencies).

Yes indeed, a good point Doug - so for perceived loudness it may be better to replace the Amplify in step 3 with the use of a compressor (and I would probably recommend Chris's Dynamic Compressor for this rather than Audacity's Compressor effect - see this sticky threas: http://forum.audacityteam.org/viewtopic ... 42&t=58958 ). But that does assume that you do indeed want to compress the music rather than just amplify it.

One of the problems with a straight "Amplify" is that any single high transint peak will keep the rest of the music at a relatively low amplitude even after using the Amplify effect.

WC

My preferred method is the manual method that Shippo described in the first post. It is the most accurate method, and you get to enjoy your music while doing it.