Singer-songwriter Phil Maffetone, former physician to the Red Hot Chili Peppers, and an artist with whom I have worked in the past (Livin’ on a Dead End Road, recorded in Oracle Arizona) asked me to remaster some of his on-line offerings. I recently completed work on The Best of Phil Maffetone, and now I am working on songs from his Nashville Bootleg, released in 2012. The process I used for remastering was different from the way I normally work.

This was an unusual request because he asked me to work with mp3’s downloaded from his website, https://maffetonemusic.com. I prefer to work with original or mastered .wav files, however they were not available, being the work of various other mastering engineers, including some high caliber folks.

The one exception was Wizard Shoes, one of my favorite Phil songs. Since no one had touched it since we originally recorded it (back in, 2008, I believe it was), I was able to work with the original 24-bit .wav file, because I keep everything that I have worked on in my archives. After running it through my great analog chain, which includes a Rupert Neve Designs Master Bus processor and A-Designs Hammer EQ, the song ended up sounding better than ever.

But back to those mp3’s… Originally, I balked at the idea of using mp3’s in a mastering context, and if you know anything about the biases of mastering engineers, this is not considered a best practice. Fortunately, the songs were well recorded, and then encoded at the highest bit rate available, 384 KBPS. The results sounded very good. Since Phil is located on the east coast these days, I used my Dropbox account to send him previews. That was made easier by the smaller size of the mp3’s.

You might ask, why bother? Weren’t they already mastered? Well these songs were all recorded over a period of several years. They were all over the place in terms of average level, varying from about -14 dB (average RMS) to some really smashed rocking numbers, at around -6 dB. Some sounded better than others, but there was no consistency. The task at hand was to make them all sound good but at more or less the same average RMS level, so that they could be downloaded as a set. Given that some were very loud, and Phil’s fans seemed to like them that way, I decided on a target level of around -12 dB, similar to many pop tunes.

I ran the songs through my high-quality analog chain, after doing some level processing in Wavelab, and then reencoded to mp3 format at 384 KBPS as the final step. The results still sounded great, but more cohesive as a set of songs, and Phil agreed.

So, for me the lesson was to put aside my biases, and get creative. There are always ways to make the music sound better, provided you have the right gear and know-how.