Mastering and Loudness

I didn’t get the chance to pick Bob Ludwig’s brain about what he thought of the files that I used during my demonstration at the DEG, CEA, NARAS and label event at Jungle City Studios last week. I was very pleased that he stuck around until my presentation and heard my presentation and demo tracks. As a former mastering engineer, I know the type of work that he’s been doing for many years AND I know how the tools and techniques has evolved over that period. Mastering engineers are the final arbiters of fidelity in the music production process and use both analog and digital (PCM only as DSD doesn’t allow any processing) to alter final mixes to meet their clients’ needs.

But the tracks that I played during my presentation didn’t undergo any “mastering” other than to be sequenced and adjusted for relative amplitude. I had a number of people come up to me following the presentation and tell me that they thought my tracks “sounded the best” and that they enjoyed my presentation. I know David Chesky and I agree in trying to avoid any manipulation of the real world musical dynamics associated with a music selection. That’s the difference between the commercial pop/rock/jazz records that you’re used to and the unprocessed tracks that AIX, and a few others, release. It makes a difference but it’s a very tough sell…louder is always perceived as better.

I spent hundreds of hours mastering the mixes of other artists. From crooners to classical guitar to Bad Company and The Allman Brothers, I tweaked EQ and dynamics processors…stripped out hiss and distortion with NoNoise from Sonic Solutions and made loud CDs. When the artists/clients were in the room, I would play the mastered version vs. the unmastered versions back to back and ask for feedback. In virtually every case, the louder version was preferred. The level change may have been subtle or large…but it seems everyone likes his or her music loud.

So I’ve spent some time mastering “Mosaic” three different times. The files are up on the FTP site in a folder called Mosaic Mastering Comparison. I’ve also included the original unmastered file AND a version that contains short segments of all of the files edited together so that you can have an active switched version.

Take a look at the spectrograph below. This illustration shows the differences between the four versions of the same track. The mastering involves dynamics processing and identical EQ treatments.

Figure 1 – The spectra of 4 versions of “Mosaic” from unmastered to heavily mastered. [Click to enlarge]

You’ll notice that the rightmost plot is almost devoid of dynamics…but I’ll bet that many people will like it better than the others because it “kicks” harder. We respond to loudness more than dynamic range…so go figure, where do you think mastering has taken us over the past decade or so.

I did speak with a number of mastering engineers in NYC and there is a distinct trend towards less dynamic squashing…let hope. Take a listen to these files and let me know what you think. The amplitude difference is over 10 dB!