Mixing the sound

By far the most important decision taken by the producer and mixer of virtually any surround music project is which of two primary perspectives to employ: frontally oriented with surround enhancement called the direct/ambient approach, or a “middle of the band” perspective. In fact, I believe there is a case to be made for each of these, and that ultimately it is the intersection of the aesthetic values of the project with surround sound capabilities that should make the decision.

For music, the decision whether to use a direct/ambient or in-the-band approach might be weighted by audience research (especially if you are a crass marketer), which the CEA sponsored in a blind telephone survey of 1000 households. When asked the relevant question most people wanted a “best seat in the house” perspective. On the other hand there is probably little familiarity in the general public about the issue, and this is likely to be the first time they’ve ever heard anything like the question, so they may be responding with what is more familiar to them. Also, when broken down by age, younger people seemed to take to the idea of “in-the-band” in greater numbers than older people, so depending on what you are mixing and its audience, you may find that the decision based on marketing is a toss up.

Of the two, the in-the-band perspective is the more controversial one, especially for some listeners. John Atkinson, the editor of Stereophile, recently wrote that having percussion coming at you from the back is unnatural and disturbing. A woman in a CEA focus group on surround also found it to be uncomfortable having direct sounds coming from behind her. So, for a wide spectrum of people, the experience of being embedded in the music is at least unfamiliar, and perhaps even disturbing. It is useful to know this and factor it in along with some knowledge of psychoacoustics to enhance mixes for the widest range of listeners.

I’ve been criticized in the past for my surround mixes. On the one hand, some of the critics think I’m stuck in my ways as a direct/ambient natural classical music mixer and wonder aloud why I don’t exercise the freedom that surround sound offers to put the listener at the conductor’s podium and wrap the sound around more. Recently I’ve been criticized for exactly the opposite: playing too much with space making mixes too gimmicky. Interesting. You’d hardly think I could do both! I guess I’m an equal opportunity offender. I’d like to explain how examples of each type came about, and where things were panned, because this is the central question of perspective.

The classical music mixes I’ve worked on involved spaced multi-microphone technique using a basic ORTF pair of cardioids, with head-type spacing and angled at 110 degrees centered over and behind the conductor’s head, an outboard set of left and right cardioids more or less in line with the ORTF pair, cardioid spot mics, and omni and cardioid hall mics. The left and right outriggers were assigned hard left and right, the center ORTF pair just to the left and right of center, and the spot mics where appropriate to match into the soundfield. I have used time delay on the spot mics, because otherwise as you bring the level up you are also adding in a sound that arrives at the spot microphones earlier than at the main ones, which leads to exaggeration of the spot mic due to the precedence effect. The hall mics were omnis directed to the surround speakers only, and the hall cardioids were pointed away from the orchestra, adding a deeper perspective than even the omnis. With the right perspective on the surround mics, which comes about as a function of placement in the hall, they may be too far away from the stage in time. So again time delay (or its postproduction adjustment editorially in a digital audio workstation, which is what I used) is useful. In the 10.2-channel version of these mixes, the omnis are sent to the dipoles and the cardioids to the center back in order to make a more fully enveloping soundfield. There’s a lot more about the natural, direct/ambient style approach in the long paper described in Relevant Research this month. In fact, the paper goes beyond anything I’ve tried, but I do mean to try its techniques over the next year or so. Stay tuned.

One of the happy accidents we got along with the music was stage hands setting the stage of a hall while a piano was being tuned. The reviews? Here’s one:

Finally I heard a system that is light years ahead of the rest when it comes to re-creating a sound space and placing sounds within it. This was a 10.2-channel system, from a company called TMH, designed by Tom Holman. Unlike much of our present-day surround sound, ambient cues did not sound like anything specific. That is, for those demo recordings that were simply trying to re-create sounds that were happening on stage in front of you, there was no indication that any sound at all was happening to the sides or to the rear. Yet, it sounded as if you were in the acoustic space of the recorded venue. Truly incredible. I didn’t have to imagine that I was in the recorded venue because it sounded as if I was. 1

Yes there is an advantage here of the 10.2 system over standard 5.1 in producing greater envelopment because of the greater number of directions of arrival we’ve got, but even the 5.1-channel version of this program material works quite well. The approach is strictly direct/ambient, and the reviewer obviously got the point: you don’t hear the surrounds obviously, but shutting them off makes the soundfield collapse into the front. I have found experimentally that the total level of the surround sound vs. direct sound is very critical, with a ±1 dB change making a huge change in frontal “focus” vs. surround “envelopment.” It is easy to go wrong, so watch out for your monitor system calibration, as it turns out to be critical here.

Now maybe I’m getting looser as time goes by, but it is early in the use of multichannel sound for music, and I think experimentation is still in order. But perhaps there needs to be some organizing principles applied to the logic of experimentation. For instance, there are some known psychoacoustics that mitigate against attempting imaging everywhere in the 5.1-channel system, or even in quasi-6.1 (Surround EX), 6.1 (DTS ES), or 7.1 (Logic 7). This is because it is a mighty wide gap between a left loudspeaker at 30 degrees and a left surround at 110 degrees, and the different frequency response in your ear canal for these directions of arrival makes them not match, even if your speakers and room are perfect. This means that panning down the sides in 5.1 is fairly hopeless, with different notes from the same instrument popping back and forth due to the different frequency responses. So side imaging should probably best be left out of the equation, although we hear about it in workshops and seminars in the field. Some believe in such a simplified world that there is a stereo soundfield in front of us, complete with phantom imaging even if the center channel is not used (“it’s the devil’s work — don’t go there,” as one editor put it to me), and one behind us, and one to each side. Yes, phantom imaging works in front and behind, but it does not work to the sides!

So knowing these facts, and consolidating them into a mix, here is what I did. Herbie Hancock loaned us the 46 original tracks of his tune “Butterfly.” As he pointed out at CES, the album flopped in the marketplace because the record company demanded a stereo mix, when the content was too big for stereo and demanded a multichannel format. I started by listening to all the tracks, some in groups.2 Then I broke down the tracks by type and made some basic decisions. There were certain tracks that are attention-getting percussion; others are more warm and fuzzy. The tune is in the jazz idiom, where different solo instruments take over from time to time and are highlighted. Then the whole thing comes to an end with a flute solo without accompaniment, after having a wall of sound up to then. This was a 10-channel mix, but the same principles would apply to a 5-channel version. I started by placing the percussion across the front stereo stage and centering the soloists where you’d expect they might move and take the spotlight. Actually, maybe this is a little weird because soloists don’t usually move to the center, but having the audience know where each soloist was going to enter in this complex mix kind of helps them in sorting things out, I think.

Having the percussion up front reduces that feeling of uncertainty that occurs when transients come from all around you. Next, Herbie’s keyboards produce a warm sound and I felt they should be enveloping. So I put them principally into the left and right wide (±60 degrees) and left and right surround to embed you inside them. Certain spot sound effects I left for center back, as they were intermittent and adding a certain zing to be put there, without being annoying, at least to me.

I theorized that certain instruments might be better if they were moving. Following a lot of sound around is tiring on the listeners, but moving some sound can be exciting, I think. So I picked just a couple of things to move. The first is a set of hand chimes. Usually played by hanging them in the air about at head height of the musician, running a glissando across them really seemed to make sense putting them into the left and right high channels (±45 degrees in plan and +45 degrees elevated). Then the listener hears something like the player would hear, with appropriate motion. Finally, the killer flute solo occupied the spot of the soloists in center front when it played earlier in the tune, but at the end it’s heard alone and exposed. Here it made sense to me to bump it around the multichannel soundfield, as it is alone and completely exposed.

I played this back for Herbie, and he liked it. He told me that the flute was the “Butterfly,” and that it takes off at the end. Duh — of course. I hadn’t thought of the title at all, being buried in the mix details. How did it go over?

During the presentation I attended, jazz great Herbie Hancock gave a short talk about how multichannel sound can be a boon to musicians…It’s nice to hear from a leading musician such statements as “stereo is unnatural, surround is natural,” and “the more channels, the better.” He cited some of his own stereo work that was too sonically complex to fit comfortably within a two speaker medium and demonstrated a 10.2-channel remix of one of his tracks that stunned all in the audience with its beauty and musicality.3

The sounds to the sides were those that did not have a particularly hard edge to them. Further, I’ve reported earlier on side imaging in the 10-channel system. It seems that, by “filling in the gap” between 30 and 110 degrees with a speaker at 60 degrees, the disjuncture is avoided, and side imaging, even at 90 degrees, can work.

In the end, I think both styles of work have a great deal going for them, and that the choice between them should relate to the aesthetics of the content, more than forcing one solution or the other.

1. www.planethifi.com
2. Assisted by Andrew Turner.
3. www.soundandvisionmag.com

Surround Professional Magazine