on unix will set the volume level for all mp3s below the current directory so that they sound similar. it sets metadata, so it doesn't change the signal itself (ie it's reversible).
as for the site linked - if you're going to have a blog about your business, have an easy-to-find link to the business on the same page (i couldn't find it and ended up editing the url).
It's slightly more complicated than that -- depending on how the track was compressed or limited you end up having to make a decision to apply that light compression/limiting (possibly altering how a musician wanted their track to be heard) or 'compromise' by lowering the volume of all tracks to give enough headroom for the especially quiet ones to be scaled up. The more 'uniform' you want your tracks to be, the more aggressive compression you have to apply.
But that is exactly how both Replay Gain and EBU R128 behave.
Initial set of conditions:
- we want music to have uniform volume
- we want louder parts be slightly louder
- we want quieter parts be slightly quieter
In the end, over compressed songs are penalized. Songs that are recorded too quiet are pumped up. But it is a per song setting, leaving song to song progression on same level.
> In the end, over compressed songs are penalized.
OK, that's what I'm saying though. You can attach 'loudness scores' to songs all you want, but eventually you have to pump it through a fixed-range DAC and have to make some tradeoffs :).
as for the site linked - if you're going to have a blog about your business, have an easy-to-find link to the business on the same page (i couldn't find it and ended up editing the url).