I am in the process of importing, cleaning up, and exporting as WAV or AIFF all of my original music coming from different sources - cassette, CD, digital file produced from DAT. I have a several different projects going - some for a single song, others for an entire album or live recording. The goal is a single “anthology” of all my original music.
What I’m looking for is advice on unifying the volume of all these tracks, so that when people listen to my anthology, the volume won’t change drastically between tracks (where now, due to recording circumstances, some are much quieter/louder than others overall).
I’ve taken two tracks for a test drive using the peak amplitude analyzer - amplifying them both to just below 0.0 dB, but that isn’t working to get them to a similar overall volume - I assume because the loudest parts, which aren’t necessarily representative of the overall volume of the track, are driving the peak amplitude.
Is anyone aware of a tool or method (even one you just figured out yourself) for achieving a uniform volume level for a whole collection of songs across multiple projects?
Any advice would be greatly appreciated. Thanks!!
The very best of luck with that. Music production over time has gotten louder and louder, not because the volume got turned up in the delivered product, but the sound got more dense and forceful through aggressive sound processing. This is the “loudness wars” thing. To cut to the chase, you’re probably never going to match older and newer commercially issued songs except possibly one by one and by ear.
Amplify and Normalize don’t “know” anything about music. They look at the tips of the blue waves and take action based on those. Because it’s cheap, inexpensive, easy to program, and it semi-works about 75% of the time.
So I don’t know an easy way to do what you’re doing automatically, and even manually you can get into trouble. If you start out life with one of the newer, louder productions and then go to match a very much older but similar song, you will find that you can’t make the older one loud enough without overload and clipping distortion.
They’re just very different as I assume you’re finding.
Try [u]ReplayGain[/u]. For WAV files you can use [u]WaveGain[/u]. Or if you play back on iTunes, an iPhone or iPod, turn-on Sound Check.
Note that any of these algorithms (or volume matching manually) will require that you make your loud songs quieter because, as you’ve discovered, some quiet-sounding songs are already maximized/normalized and these particular quiet songs can’t be made louder without clipping/distorting.
I assume because the loudest parts, which aren’t necessarily representative of the overall volume of the track, are driving the peak amplitude.
Yes. Plus our ears & brain don’t respond to loudness instantly, so a few short-term “loud” peaks here and there don’t necessarily sound loud.