So it's basically like a broadcast compressor, right?
No, not a compressor. It doesn't compress the signal at all. It's not a limiter either. It doesn't squash the dynamic range of the audio in any way. All it does, is detect the highest amplitude in a song and then raise/lower the playback volume for the entire song when it starts playing, so that at the loudest point in the song never surpasses the Replay Gain setting for loudest peak. The dynamic range does not change at all. So for example, if we have our device set to maximum volume, and we have a song like this when played back normally:
Quietest point = 40 dB
Average loudnes = 90 dB
Loudest point = 100 dB
If we listen to the song as is, it's going to damage our hearing, because it's beyond safety threshold.
And if we have another song like this:
Quietest point = 10 dB
Average loudnes = 60 dB
Loudest point = 75 dB
When we play this much quieter song back-to-back with the other much louder song, one of two things will happen:
1) If the louder song plays first, we would need to adjust the volume so the song isn't too loud and hurt/harm our ears. But when the quieter song plays next, we can barely hear it, and have to raise the volume.
2) If the quieter song plays first, we have to raise the volume so we can hear it at satisfying levels. But when the louder song comes on next, it is ridiculously loud and instantly damages our hearing, and we'd have to fumble for the volume control to bring down the level before we experience permanent hearing loss.
Now, if we activate ReplayGain, this is what happens:
Both songs are analyzed, and both playback so each song's loudest peak does not go over the set limit (I would set this at around 85 dB. If you don't know how to do that, I'll explain later). Now, the first song would playback like so:
Quietest point = 25 dB
Average loudnes = 75 dB
Loudest point = 85 dB
And the second song would playback like so:
Quietest point = 20 dB
Average loudnes = 70 dB
Loudest point = 85 dB
The dynamic range stays exactly the same, and its only the entire song's volume level that's been raised or lowered.
You can learn all about ReplayGain easily: h
ttps://www.google.com/search?q=replay+gain&oq=replay+gain&sourceid=chrome&ie=UTF-8
As for setting the volume level exactly to a number. This is how you can do it:
1) Buy a sound level meter:
Sound Level Meters | Instrumart
2) Pick a song that is one of the loudest songs in your library, with consistent volume throughout (as opposed to a song that has quiet and loud moments). Usually really aggressive and pounding heavy metal or electronic dance tracks during its most dramatic moments, where all frequencies have ample presence is best.
3) If listening to speakers, simply use the sound level meter (set to A-weighted, fast or slow isn't that important) at where your ears would be and point at the speaker(s) and measure, play your loud song, then set the volume of your player/speakers until the loudest peak does not go over 85 dB. This is a pretty safe threshold in general.
4) If listening to headphones, then you need to measure the headphone's driver directly. If it's an over/around the ear headphone, use an optical disc of some kind (CD, DVD, Blu-Ray) and cover the headphone's earcup, as if the disc is the side of your head. Now place the measuring mic of the sound level meter at where the hole of the disc is, and then measure. If listening to earbuds or in-ear-monitors, then just put the measuring mic right up against the driver and measure. If you can seal off the sides to block off the leakage of sound (maybe with a piece of cloth or tissue paper), then do that.