ProAudio Review - October 2008 - (Page 48) MASTERING MATTERS by Alan Silverman Where’s The ‘Air?’ CD reissue of At Last by Etta James is playing in my mastering room, and it sounds as vibrant and brilliant as any recording released in the 47 years since the song was first a top ten hit in 1961. Correction: it actually sounds more alive than most of what we’re hearing today. There’s a feeling of what sound geeks like to call ‘air.’ That’s not to say that exciting, realistic presence is absent from all modern recordings — just far too many, it seems. In my mastering work, I do a good number of vintage reissues and in such sources the perception of air is the norm rather than the exception. What is this quality? How have we managed to lose so much of it, even as superior recording technology becomes ever more commonplace? ON AIR In asking this question of colleagues, a few explanations have been given that could be regarded as conventional wisdom on the topic. One idea is that there is distortion inherent in analog tape recordings that our ears interpret favorably as “presence.” Yet standard resolution digital is obviously capable of communicating this presence or it would not be audible on a proper CD reissue. There are many alldigital recordings that have ‘it’ too, so while analog tape coloration can contribute to the effect, it isn’t essential to it. Another line of reasoning is that recordings of the 50s and 60s were done with far simpler signal paths. A typical session utilized a small number of tube microphones, themselves far simpler in circuitry than FET and chip-based designs, recorded through mixing consoles that were much simpler and more direct than the large-scale desks of today. When it comes to signal stages, ‘less is more’ has always been an audio axiom. I think there is a lot of truth in that, yet many recordings today are made without a console at all, using the very direct approach of ‘mic, preamp, converter, DAW’ with little or no processing on the way in. Even though the direct approach can help preserve ‘air,’ the end product from such a scheme all too often lacks vibrancy anyway, so something else must be at play. WHAT’S MISSING? I have my own theory of what might be missing. It’s a theory that would require scientific testing to be proven, but it can be easily applied should you want to try it for yourself. It has worked for me. (Note: For the purpose of this article, I’ll assume that we live in an alternate universe where the life and dynamic range of recorded music is not routinely obliterated by the insane use of extreme digital limiting or at least that our own reality will someday grow beyond the practice.) Start with the observation that vintage recordings were usually done with a lot more going on in the room at once. Parts were not tracked one at a time. Even with the advent of three-channel recorders in the late 50s, multiple instruments were tracked on each pass. Before anyone gets their hackles up, it’s a given that tracking instruments separately is here to complex and chaotic difference-tones that occur when instruments play together live may well be essential to the perception of what we call ‘air.’ If a recording is going to be made primarily through over-dubbing, as most recordings are today, we can apply this principle or ignore it. IT’S SO EASY Ignoring this sonic aspect of live music is easy – simply overdub everything at 44.1kHz sampling. Proper digital filtering requires that all frequencies above 1/2 the sampling rate be filtered out = thrown away. Failure to abide by this mathematical law results in ugly digitalalias distortion. All properly designed digital converters adhere to it. So we digitally track a cymbal at 44.1k, discarding its inaudible 30kHz harmonic, and later track a trumpet without its 28kHz harmonic. Individually, both tracks sound perfectly fine – the lost harmonics weren’t audible anyway … not a problem, at least not a problem until the mixdown. When we later combine the two tracks, the 2kHz difference-tone that would have happened live in air never happens. The 30kHz and 28kHz tones that would have produced it were thrown away during A/D conversion. This aspect of realism will not be part of the mix. Applying the principle is easy, too – simply track at 88.2kHz or higher. Many of the inaudible high-order harmonics are now preserved. At the mix, whether in the box or through an analog summing mixer, the audible differencetones that help cue the ear that something real is happening will be produced. To get to CD resolution afterwards, use a high-quality SRC utility such as Wave Editor or iZotope for the Mac or r8brain pro for the PC. An excellent dual-platform SRC is the new SARACON from Weiss Engineering. Again, this idea is far from proven, but storage is cheap these days, double-sampling converters are everywhere, and most DAWs are up to the task. So give it a try; let the air in. Alan Silverman is the owner of Arf! Mastering in New York City and a multiple Grammy Award nominee, including for Album of the Year. stay. The majority of modern recordings totally depend on being done this way. If ‘air’ is going to be preserved, however, there is a way to do layering and a way not to. It is a fact that harmonics well above the range of human hearing are produced by acoustic instruments and are picked up by microphones. You can easily see such aboveband high-frequency activity on a real-time spectrum analyzer, but harmonics above 20kHz are inaudible, so why should they matter? Go back to the word ‘air’ — when sound waves mix in air, difference-tones are produced. Say there is a cymbal producing a harmonic at 30kHz and a trumpet producing another at 28kHz. Neither frequency is heard directly by the ear, but when those instruments are playing together in a room, they mix together in air producing a difference-tone 2kHz, and 2kHz is audible. The multitude of 48 | ProAudio Review | October 2008 www.proaudioreview.com http://www.proaudioreview.com
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.