Festival97 Report
PARSONS AUDIO's 7th annual expo/conference
October 14 & 15, 1997 at the Newton (MA) Holiday Inn.
Exhibits: NEW TOOLS
Workshop: PRODUCING & ENGINEERING IN SURROUND SOUND, with Tom Bates, David Moulton, and Tom Jung*
Panel Discussion: COMING SOON TO SURROUND US, with Tom Bates, Tom Jung, Bob Ludwig, and David Moulton*
Presentation: DOLBY MULTICHANNEL AUDIO, with Jim Hilson of Dolby*
Open Discussion: AUDIO FOR VIDEO -- WHAT IS COMING?, with Tom Bates and Jim Hilson*
*Panelists and Workshop leaders
Too briefly to do them justice: Tom Bates is a producer/engineer whose projects have won eight Grammies; also a former audio director at NBC, for Emmy winners Saturday Night Live, Live From Lincoln Center, and Live From the Met. Jim Hilson is Dolby's applications specialist in multichannel products. Tom Jung is a noted producer/engineer, and founder and guiding light of the Digital Music Products (DMP) label; also a well known equipment reviewer, technical consultant and digital audio expert. Bob Ludwig is one of the world's most renowned and respected mastering engineers; has mastered many dozens, perhaps hundreds, of best-selling recordings; has won numerous TEC Awards; and founded Gateway Mastering in Portland, Maine. David Moulton is an accomplished teacher (Berklee, SUNY/Fredonia, etc.), writer (Recording magazine), author of the Golden Ears critical listening CD's, loudspeaker and acoustical designer, and a studio owner and engineer. These folks know our topics well. They have been working with them, and helping to shape them, for years.
There's a lot developing in our industry these days, which many of our customers are finding both thrilling and unsettling. Excitement in all corners of our event this year was appreciably above the norm. Forces within the computer, video, film, telecommunications and home entertainment sectors -- forces vastly beyond the scale of the audio business -- are clearly making certain that we audio dweebs will face dramatic changes over the years just ahead. To hype Festival 97 we described the present and near-present as a frontier moment. In that spirit we brought in some leading explorers to speak with us future settlers about what is happening.
Exhibits: NEW TOOLS
In one space at our annual event were exhibits by many of our industry's leading equipment manufacturers. Crucial supporters of all of our enterprises, they invariably have impressive developments to show. This year, as usual, they did not disappoint us.
Many key manufacturers exhibited: Akai, Avid/Digidesign (new 24-bit, 32-track Pro Tools 24), Dolby, Earthworks, Lexicon (with the new Studio workstation), Panasonic (with Merging Technologies' Pyramix workstation), Roland, Sonic Solutions, Tascam, Troisi, Yamaha (02R, 03D, etc.), Bay Roads Marketing (representing Audio Accessories, Focusrite, Genex, HHB, KRK, Sony, Soundscape, t.c. electronics, Troisi, Tubetech, etc.), NETA (Crest, EAW, Joe Meek, Sennheiser, Spirit, White, etc.), and On the Road Marketing (Mackie, Symetrix, Tannoy, etc.). The many new developments by these manufacturers are well reported elsewhere in the trade press; also on their own Web sites. Contact us us for any particulars.
Workshop: PRODUCING & ENGINEERING IN SURROUND SOUND
For the Surround Workshop, over 120 people gathered within a potent surround sound system consisting of Genelec 1037's and subwoofers, a Yamaha 02R enhanced by a Troisi 20-bit 8-channel D-to-A, and a variety of source machines. The Genelec and Troisi were provided courtesy of the manufacturers. The Workshop consisted of presentations by three producer/engineers: Tom Bates, David Moulton, and Tom Jung.*
The heart, soul, and inspiration in the Workshop resided more than any place in the thrilling surround works played for the audience by the three presenters. The room and audience were too big for ideal listening to surround, and there was too little opportunity to tweak the playback system in advance, but it still sounded great. No words this reporter might write here can convey either the quality of those pieces or the excitement of the audience upon hearing them. But in addition to playing sublime recordings, each of the three said interesting things.
Here, in notes form, is some of what they said:
Tom Bates
I have been working in discrete surround for several years; releasing in stereo, of course. Still, everyone is a beginner now, even though, especially as one's accomplishments mount up, clients expect that all of God's knowledge of audio is included in your genes.
Some things work. Others don't. For example, I have yet to establish an image inside the ring of speakers. Anyone done it?
I like the overhead speaker, though I realize it's not likely to become a reality.
When I was young I would put on a record, sit down, close my eyes and let the music wash over me. It was a major source of enjoyment and satisfaction for me. The same can't be said of many young people these days. For several reasons there's a problem with record sales in general and with putting music in the forefront of people's consciousness as a form of entertainment, including: 1. There's a glut (even a confusion) of product on the market. So it's hard for a music and its audience to find its market. 2. We purveyors of music are losing a battle in sensory assault. Our competition is Nintendo, Internet, rented movies, etc. These things offer much more sensory input. Surround may help to change that.
An analogy: When we first recorded music, in terms of soundstage mono was like looking at it through a window. Stereo is like a wider window. Surround is like going out there and being in the same space with it.
Miking rooms, there are basically two types of surround production. 1. Record multitrack mono or stereo and then mix it, creating a false room. We don't yet have 5.1 channel compressors and efx devices. There's nothing too difficult about making them, though; they will happen. 2. Do production involving capturing a room.
3 mic techniques I have used:
Five spaced omnis, placed roughly where the speakers would go in a playback space, as a starting point. Then I cheat on that, move the mics to where they sound good. It sounds good in many circumstances, but is hard to justify mathematically.
Tom Jung and I have done many experiments together. We like surrounding a head with five cardioid mics. Often, however, halls don't have one place that sounds that good.
Model halls using an MSM technique. This images well if you don't combine it to mono. You can capture it in the field with only three tracks. This has fallen out of favor with me.
(Tom played several examples. One was a recording of a rehearsal by the Hartford Symphony -- 100 orchestral members, a 100 voice choir, and 3 brass bands in the Balcony; captured by Tom and Tom Jung using five spaced omnis.) The orchestra, appreciating that we're in it together regarding surround, gave us permission to record for free.
(Then Tom played a 5.1 soundtrack recording of a gospel group recorded live -- 80 members approximately, with funky band.) The setup had 54 lines coming in, using two 02R's, including a couple of audience mics. ...I felt that the recording was only a qualified success. I found it to be missing front to back depth, sometimes with an ill-defined center image. You couldn't feel that the audience yells were right next to you. Instead they tended to seem like they were along the wall. I think I anchored the lead voice successfully by putting 90% in the center speaker; otherwise there's only kick and bass in the center speaker , also sent to the sub. I used a phantom image for the choir.
(From David Moulton, during a Q&A:) In stereo you have the median plane. The sweet spot is more ambiguous in surround. In surround there is a stereo zone. As you move backwards and forwards the musical experience changes - - the front of house sound differs markedly from what comes to listeners at the rear. One production challenge is to manage that factor for the most interesting array of sensations.
TB again: While producing I tend to make an adjustment and then walk around the room. I tend to stand when mixing in surround. That's unlike stereo in both respects.
I am petrified of bass management. There is no standard for setting a subwoofer. I would shy way back on subs, preferring there to be too little bass rather than too much.
I likes my monitor speaker to be full range all the way around, using the same speakers.
I often use matched mics too -- Tom Jung's custom Shures, for example.
To create space for a renaissance record for BMG I used a stereo reverb (one of the 02R's', with dense, short decay settings) returned to the front; another reverb returned to the back; then a Lexicon 300 as overall reverb to all channels.
(To end, Tom played a tour de force: Sonny's Car Wash.) To capture it I tried 7 car washes, looking for "the Stradivarius of car washes". (laughter) The looks I got driving in with the world's most spotless car, with C-24 attached! (more laughter)... The most remarkable sonic gem occurs at the end of the recording. When the car exits the car wash, I lowered the driver-side window to adjust the side mirror. Especially when listening at the sweet spot, you can sense the change in pressure as the window drops and then rises again.
One can easily move the apparent walls out vis-a-vis the listening room's walls. I think it should be possible to move them in too.
In monitoring surround, the direct sound from the speakers far outweighs the sound bouncing around the listening room.
Dave Moulton
Since quad in the 70's I have been intrigued by the compositional possibilities of surround.
·
Like Tom, I want all the loudspeakers to be identical. I would like phantom images to be able to come from any direction, including overhead.
I have developed speakers that are flat to 15k up to an 180 degree angle. The behavior of images with those speakers is very different than with speakers with narrower dispersion. They work very well for surround.
When I first put up a full-range overhead speaker I thought it might prove to be no more than cute. It turns out I liked it immediately. Then Tom Bates made some recordings with overhead information. The results were powerful. You seem to be listening to a real room. It may not be an accurate representation of a real room, but it's a real room. The overhead somehow couples the front and the back in a meaningful way. It's mostly revealed when you take it out. With an overhead, when you take it out the cathedral you were listening in just got a drop ceiling. You have a sensation that what's overhead is white fiberboard.
Once exposed to recordings that include an overhead speaker, musicians tend to feel that you must have the overhead. It is more important, they say, than the center speaker. I hope that an overhead will become part of the surround format. People say it will never happen; that no movie theater will install an overhead speaker. But I keep fantasizing the first movie with a helicopter that comes from the rear, moves overhead, and then proceeds forward onto the screen. Whenever someone first produces such an effect well, the gold rush will be on -- theaters will rush to put an overhead speaker in.
(Dave played a recent surround recording by Curt Wittig, in Swarthmore PA, of George Crumb's Ancient Voices of Children.) Curt: In all 14 mics were used, to allow room for decisions in post-production. We used two Soundfield ST250's. One was placed as normal for acoustical classical recording -- somewhere behind the conductor's podium, to gather the hall and orchestra. The second mic faced to the rear, moved back to the critical distance of the room so that reverberant information could be at least as loud as direct info. Think of the rear-facing mic as a reverb return, conveying the recording space in real time. In stereo, summed to mono, the recording managed to keep reverberant values, and a sense of the details. The mics were less than 15 ft apart, so as not to get a slap. A Schoeps omni was placed in front of the forward MS, for the center image. An additional cardioid mic was placed 12 feet up, five feet behind the main MS, pointed to the ceiling.
Dave again: Surround recordings should go for a large, generalized, fuzzy sweet spot.
The Crumb was a discrete 6 channel recording. Central to the production was an effort to capture the sound of a hall with musicians. Issues: 1. the sense of envelopment, which comes in large part from the rear. So (for this piece) you want a strong reverberant wash in the rear. 2. With the musical sounds, as the direct sounds happen, there's a migration of direct-to-reverberant sound from front to rear. It is very satisfying to hear that motion over time. (During playback, upon taking the ceiling speaker in and out, the audience felt that it was important well beyond being subtle.)
We don't know yet how to bring sources into the room. How to have someone in the center speaker walk right up to you cognitively. It will be a powerful illusion whenever someone learns how to do it.
Curt Wittig: Crumb has been composing in surround all his life. This was the first time he has heard it played back to him that way. He was very excited.
(Dave then played a recording by Curt of the The 20th Century Consort at Washington's National Cathedral.) As it was being recorded there was a carillon concert outside. It is audible right down at noise floor, around the dither. Beautiful. I'm beginning to think about the possibilities of using musical dither instead of shaped noise!
(Turning to pop music and an A+B, A-B mixing aesthetic:) The central message of pop recording: lead vocal, bass and kick drum in the Center (that's A+B); the A-B stuff off to the side -- spatial and musical counterpoint to the A+B. The Center becomes discrete center. Instrumental leads, stereo leads, parts etc. are outside them, as a big stereo spread. Some reverb at the back, and some strings and rhythm. And occasional hockets -- signals bouncing front to back.
(Dave then played an Alan Parsons DTS disk.) Notice that there's no middle, just a standard stereo panning routine with nothing in the Center channel. There's nothing wrong with that necessarily. It points up that for surround you can't just reliably pan from L to C to R. We have a lot of work yet to do in that.
Tom Jung
(The audience was knocked out by Tom's recordings, most of which are available commercially.)
The Yamaha 02R console with v2 software can hard-assign to the Center channel, or create a phantom channel, or vary the ratio of phantom to C. That's nice!
Most of the 5.1 systems out there are primarily designed for home theater. You will find usually that the Center speaker is not the equal of the L and R; it is a lesser speaker. ... If you put important info in the Center, you'll get burned. A quasi-center "covering yourself" approach is the one I'm going to use for now.
(Tom played one of his big band recordings.) I used 3 cardioids on 3 stands -- 2 UM-95 Soundelux large capsule vacuum tube mics; 1 C-12VR as the Center or forward mic. I had the band blowing into a concave RPG performance shell. I pointed the MS mic at the shell to catch highly diffusive reflections; fed most of that to rear; some to feed a stereo reverb which also went to rear, to add to the tail. It was not a huge room, so the tail needed enhancing.
(Glenn Miller band recording.) The sax section was positioned in a "V" with the brass section. We put a stereo mic pair on brass, and a stereo pair on the saxes. We panned the Center to Left and the Center to Right, respectively. For the rhythm section I used spaced omnis.
(17-piece percussion ensemble.) A Soundfield mic was used onstage, and a couple of front facing cardioids for the Center channel, and 2 omnis in the hall. We ended up moving the (hall) mics in. There was too much separation at first. The mics ended up just 7 or 8 rows back.
I often usesome experimental studio cardioids acquired from Shure. They're smooth and quiet.
Q&A period
Dave Moulton, question: I'd like a show of hands from the audience as to whether the overhead speaker was of some considerable value?
A: (Most and perhaps all hands went up.)
Q: Is working in surround a lot more work than expected?
TB: It is not a relevant question. There are many rewards to working in surround. It's a pleasure to spend time at it.
TB: This (producing in surround) is all learning curve at this point. There is much lack of predictability as far as how to get good results.
DM - There's lots of brainwork to panning and bussing issues if you use a normal console. In some respects projects are easier in 5 or 6 channels than in 2 -- there's much more space to work with.
TB: When energy is distributed into 5 or 6 speakers, you gain clarity. Into 2 spkrs things can combine too much. Sometimes, of course, there can be too much clarity, so that you lose musicality. Something from overhead is better than nothing.
TJ: There are not enough places to try surround mixes yet.
TJ: It doesn't work to have a big band wrapped around the listener. Solos from the rear, etc. can be annoying.
DM: In certain spaces and music, however, instruments originating in the rear can work well. The Junior Wells' DTS disk, for example.
TB: The human brain doesn't like to have things happening behind you. If there's something interesting you want to turn and face it. Which maybe means we should all listen from swivel chairs! (laughter)
DM: Surround monitoring is if anything less difficult in terms of listening fatigue than stereo. ... It doesn't have to be as loud either. ... From the dispersion standpoint, a full-range regular ceiling speaker (low-fi) might be fine. But a safe answer is to use a speaker equal to the others; then let the producer control what goes there.
Dave Griesinger of Lexicon: All the cues to the ear for elevation are over 1000 Hz. In terms of avoiding symmetrical room modes, it might be better not to use an overhead speaker generating much below 1K.
TB: I prepare stereo and surround mixes as if they are two separate records.
TJ: I may do a matrix mix, but concentrate on the stereo mix.
Panel Discussion: COMING SOON TO SURROUND US
On the evening of October 14 100+ souls filled the room for this discussion. The panelists were Tom Bates, Tom Jung, Bob Ludwig, and David Moulton* -- highly accomplished and influential users and observers of recording/production technology. Mark Parsons moderated. They and the assembled assessed some of the technological changes that are coming upon us, including the advents of 20- to 24-bit word sizes, higher sample rates, DSD, and surround sound.
Here is some of what was said:
(What does the seemingly perpetual quest for improved technology amount to? Mark Parsons began by holding up a 1904 Edison Gold Moulded Record imprinted with the phrase "Echo all over the world." He observed that we are still striving to fulfill that promise to deliver music in as beautiful and compelling a fashion as possible to the rest of the planet; that the most discriminating listeners of earlier eras like Edison's insisted that they could not distinguish between a recording and its source; that perhaps a century from now listeners will consider our best recording/reproduction technology, which we tout as transparent and realistic, to be as low-fidelity as we now consider Edison's, 78 rpm records, the LP, analog tape, first generation digital, and everything else that has gone before. That possibility raised, the first question was "What is happening now?")
Tom Bates: Changes in the audio industry are being driven by developments in much more powerful industries -- film, video and computer. .... The record industry is basically broken. It needs to reinvent itself, and it will. .... At this point it is about impossible for a music and its audience to find each other. That's worrisome, but technologically there are lots of promising things going on. .... It is certain that in a hundred years people will hold up a CD and say, 'Can you believe that back then people actually thought these were good recordings?' We're nowhere near realistic reproduction. .... Early Bell Lab experiments, with test subjects listening to table radios and a live orchestra, they preferred the table radio! .... But the coming changes will bring much improvement, especially surround sound. .... Using an overhead speaker with 5.1 is...a real step forward, exciting. So is the probable 96k sample rate, though I'd be happy with 64k, and more bits. Every bit you add adds much more impact.
David Moulton: For multichannel audio, the pump has been primed by the millions of Dolby Pro Logic decoders in people's homes; ...the coming into being of tens of millions of home theater systems. In that sense there is a hardware base now for which we can release product which is at least vaguely surrounding. More exciting and important to me will be the natural coming of surround sound to car audio. Then the gold rush will be on. Pop artists are already producing some excellent material for surround release (he cited the Boyz to Men CD in DTS). When the rush begins, then will come a complete reissuing of the discography, this time in surround.
As for the the 'facsimile event', wherein you can't tell the difference between the source and the recording, where you'll be able to close your eyes and not tell whether what you're listening to is a piano or a loudspeaker, I think we're miles away from that. The best we're doing now is simply not a facsimile of the event. We have no way to really do reality. And if we did I'm not sure we would like it, because we would get the reality we don't like along with the reality we do like. I'd like to work towards a stronger musical reality -- loudspeaker music as opposed to symphony orchestra music, or jazz group in a bar, etc. Loudspeaker music has its own particular sound and style. That's what we're working with, and will be for a long time.
(Bob Ludwig began by holding up a Pioneer 96k 24-bit DVD player and a disk burned at 96/24. His studio, Gateway Mastering in Portland, was one of the first studios to get the 96/24 capability. It is also one of the first mastering studios in the world to have acquired a complete DVD authoring system.)
BL: There's definitely a learning curve involved with all these technologies. The stories of the pioneers getting all the arrows in the back are true. .... We've already done a number of surround sessions that pretty much just involve doing transfers. But we're getting more into it. I've already got a sessoin booked for December for a 5.1 release on DTS. There's a lot coming. From one point of view it's scary (beginning with uncertainty and lack of standards about where to place speakers). THX wants one thing. Audiophile recordings another.
Philips/Sony's DSD offers the best surround sound I've ever heard. Two fronts 30 degrees off the center axis; the rears only 20 degrees back. (But there are things to beware about surround.) Back in the 70's when I was working at Sterling Sound the broadcast consultant Eric Small and I came up with a button that said, 'Back to mono.' That made its way to the West Coast where Phil Spector picked up on it; it ended up on the cover of one of his records. That was about,'Maybe quad isn't such a good idea.' You don't necessarily want to be in the center of the orchestra. That can produce disastrous results. On the other hand, when (surround is) done properly, you never want to go back to two channels. .... But it is a scary proposition to see how this is going to be handled. Marketing is going to be a key issue here. The Sony/Philips dual layer DVD, with a CD layer and a DVD-Audio layer, is a crucial step. There's enough momentum now that surround will not go away the way quad went away. It is safe to say that surround is here to stay in some form. But the acceleration of that form would be tremendously helped if there were a single-inventory disk in your record shop that would play on both systems (CD and DVD) for quite a while. To say it out loud, that other layer (on the disk) would not be 96 in the case of audio, but would be 88.2. That's desirable because the decimation needed to go from 88.2 down to 44.1 for the CD version is much easier to accomplish than from 96k.
We're still waiting for (suitable) toys to be built - de-essers, reverbs, etc. (Pointing to the 96k, 24-bit Pioneer DVD player:) The output of this 96 is going to sound better than 44.1, but won't hold a candle to a Pacific Microsonics or dB Technologies at 44.1.
For sure the average Joe on the street with a Walkman, if he could hear a 44.1 vs. 96, maybe would prefer 96. But anyone on the planet would recognize that 5.1 is a dramatically different experience.
Hopefully when the DVD Audio spec comes in.... Remember that with CD things were 14 bit until just before it came out. We do have time. I hope they don't legislate the new technologies too fast. I hope they look at a way to incorporate the DSD, even though PCM consoles and other PCM gear won't work with it. (Observing how dramatically superior the sound of DTS is to Dolby AC-3) it would be nice to hear AC-3 run at a more decent data rate, more like DTS.
(Jim Hilson of Dolby responded from the audience saying that Dolby is obliged to abide by the spec, since its AC-3 standard is at the heart of it. Dolby reportedly doesn't expect to raise its rate any time soon, but points out that others, like DTS, are free to develop as they wish.)
MP (question to Tom Jung): Tom, what do you see happening in the real world that you inhabit?
TJ: (mildly) Well, I'm terrified. I been working in surround over the past three years. Just about everything I've been working on has been in surround. Matrix surround because of its compatibility. But at the same time I have been following the work of the DTS people and have three releases in discrete DTS. Quite frankly I had hoped it would be a little further along at this point. I don't see that. I like the DTS system a lot. Even though it is lossy compressoin, it is less than 3:1, and it's a smart algorithm so it really sounds good. One thing that really frightens me about DVD audio is that we're about to take a step into new hardware, new software, new tools to make recordings, and my biggest fear is that we won't take a big enough step. So far what I've heard from some of the higher sample rates and longer word lengths is that I want more than that.
People at Sony let me use a prototype DSD machine. Back to 2-track, by the way. I think there are things about it that make a lot of sense; others don't. You really do need new tools; all these wonderful PCM tools we've developed over the past 10-15 years would suddenly become obsolete. It is important to think as far ahead as we possibly can. It would be a big mistake to settle on 88.2 or 96k 24-bit until a lot more study. Once the pro side gets things together, I'm hoping the consumer thing will get together. So it's a little scary.
MP (question asking the panelists to characterize their DSD experiences; replies spoke of transparency, and an abilty to hear around instruments three-dimensionally; then:)
BL: One of the concepts behind DSD is to make it a Holy Grail. Especially from an archival standpoint, Sony's and Philips' goal was to make the most perfect recording technology that can reasonably be done now. This system that samples 1 bit at 2.8+Mbits/second. ...It turns out all these high-sampling things require impeccable analog circuitry. It's almost the weakest link right now. That's ironic, because we know how to make great analog circuitry. .... But it is digital designers who are doing it. .... The latest iteration of DSD is extremely ear friendly. It does not give you that digital sound. But it could be better. .... You can super-down convert it to 88.2 or 44.1 and maintain a lot of the chatracteristics of the great 1-bit recording.
TB: DSD really is a high resolution system. .... (Tools for it are coming.) Sonic Solutions showed a first-generation DSD DAW at AES.
MP (question) : What is the reality of DSD? Will it happen?
BL: At one point it was definitiely inferior to 96k. No longer. DSD has got to sound considerably better than the other technologies before we go ahead with it. It's getting there.
MP (question): To what extent was buying your DVD system a risk?
BL: Buying the DVD system was a risk. Having heard 96k 24-bit recordings compared to CD, you really don't want to go back. I wanted to be able to offer it to others (record labels, producers, engineers) who want to work in it .
MP (question): You're in a position, then, to lobby labels on behalf of better audio?
BL: Yes. We've already done several 24/96, some pop, and classical. .... Regarding sample rates, it seems to me from a theoretical standpoint 64k would be about ideal. A problem is the antialiasing filters. .... People who have heard the dcs 192kHz sampling converter have told me report that it's as much better than 96 as 96 is than 44.1. Scary! And it doesn't make sense. The idea of having to store 192k!
TJ: I agree that that magic number doesn't have to go all the way out to 96. 64k is OK. Longer word length, I agree, is very important.
TB: It turns out the DVD disk isn't big enough to (incorporate) all the suggestions. I'm not saying 64 is better than 96. But as a trade-off, I'd rather live with it than lose bandwidth for other purposes. But this is moot since 96 will evidently be (the standard).
Jonathan Wyner of M Works, from the audience: Who is driving the train? When 16 bits was new, it sounded marginal. Now it's good.
MP (observing that most of the recording/production community is still working in 16 bits, asked how significant the difference is between 16 bits and more)
BL: Only when we got 20-bit A-to-D's -- that's probably when we really finally got 16-bit converters. There is no 24-bit converter now. 22 bits of true real world dynamic range is as good as I have ever heard. 24-bits are marketing bits.
The difference is dramatic vs. 16 bits. I push the rock engineers. Nick Gazauski, (who) works with an AT&T digital console...now sends me mixes that are 48k 24-bit right from the multitrack.... It really does make a difference, even at plain old 44.1 or 48, having those extra bits of dynamics. Echoes seem a lot more true. My ear feels more relaxed. Higher sample freqencies make that difference yet again to me. ....
TJ: When digital first came in one thing changed. Professionals stopped working above the threshold of what the release format was. I think it is important to be working in much higher resolution than the release format. You have more to work with when someone can bring you a mix or master which is at a higher sample rate, or 24-bit or 20-bit. Then you have a fighting chance to master a full 16-bit product. Pros need as many bits and as much data as possible. At that, we can probably do a lot more to make 16-bit 44.1 sound better. I don't think we have peaked out there.
DM: There has to come a point of diminishing returns. There has to come a point where more bits don't matter. Same regarding sample rates. My current instinct is that we would like...around 21 or 22 bits. And a frequency response of 10Hz to 40kHz for the human auditory system. At 96k that's about what we're looking at. A big range. Storing a lot of data. How much makes sense from a business standpoint (is another matter). 192k really scares me. I'm not sure in a blind testing situation we would come up with a significant difference.
On the other side of the coin, we're getting better at data compressoin. On the one hand is a purist stance that compression is evil. Recently I have been called upon to do some listening tests, using listening panels testing codecs with 12:1 or 16:1 compression rates under really good conditions (to determine) what differences are audible but not annoying -- i.e., differences that don't matter to (the listener). I have to struggle to hear the difference with 16:1 data compression. Such codecs are coming! The people who build them are getting better and better. So at the same time as we go the other way, we're developing better data compression. We have to for the Internet -- I think it will be a long time before the Internet will support 24-bit, 96k stereo, much less 5.1 -- you know, download that for your car!
MP (question about mixing for surround -- control room layout and acoustics)
TJ: I just built a new surround room. I had a limitation in terms of ceiling height. Peter DiAntonio came up, did measurements, consulted, etc. We put in a fair amount of diffusion, and a lot of Abfussors for broadband absorption. As an experiment I tried the room much drier than I ever thought I'd like, but ended up liking it. Mixing in surround I really don't want to superimpose the sound of the mix room on top of the space I'm trying to create. My experience is the drier than normal room works better for surround.
BL: Stockhausen used to recommend an anechoic room with lots of speakers.
DM: With broad-dispersion speakers, surround can work well in a fairly reverberant room. Symmetry front to back is not necessary. I like the rear speakers more than 20 degrees behind the side. Now (for the front) I put speakers at dead center, then at 45 degrees, which is wider than normal, but with wide-dispersion speakers it works well. I still believe in live-end-dead-end with an acoustic sink in front of room, absorbing early reflections, etc. I believe it will continue to work well for surround.
Jm Hilson of Dolby: Most people's living rooms aren't exactly dead. In a typical room, the side speakers will be about even with you, and the room will be more live than you want it to be. How you mix will vary with the liveness of the room.
DM: We have to have a reference room when we mix that works for us. Then we must test the mix -- a "focus room" approach (i.e., living room, car, headphones, etc.); then go back and fix it in the reference room. There is no one end-user environment. Tom Bates says that until you have made the mix sound good on every speaker system available to you, your work isn't done. ... Five- or six- channel digital radio will come to cars by 1999.
MP (question): Will surround change consumers' tastes in music? (Panelists respond to the question with amused but resolute exclamations of horror.) I mean, it might spark more consumption in general because it is so absorbing, so interesting. More music will mattter more to more people around the planet. It will lead heretofore minority tastes in music to become more popular -- jazz, classical, world and other sorts because they're so spectacular to listen to in surround. Is there anything to that?
TJ: I think music in general can use a shot in the arm. If surround can spark a little more creativity..., I'd like to see a little growth in that area. With surround I think, especially for some of the younger musicians and composers, there's a possiblity of writing for the medium. Right now we're doing pretty basic stuff. I think if some talented composer sat down and really wrote for the medium, it might get us off the lull we're in.
MP (question): Do you have pet peeves about what you hear coming at you in terms of engineering skills?
BL: I have just seen the average quality of tapes go downhill over the past decade as people get more and more into home studios.
MP (question): What are the symptoms?
BL: ADATs. (general laughter)
TB: There's something to be said for craft.
Someone in the audience: (That problem will be) self-correcting in the long run. Craft will catch up.
DM: I suspect we all agree that the problem is not the ADAT per se; it's its broad availablility. It can create a surface veneer of audio acceptability, which is being bought by the record companies.
TB: The answer is to stick to your standards; do your best; make good records. Go into every recording date with the dream that this will be the best recording you ever made. Then fight every step of the way for that vision to remain intact. (Toscanini once told his players, "I hate you all, because you destroy my dreams!" -- MP) When I first got in this business, the general level of recording wasn't that high. There were many wonderful engineers then; there still are. But the average level wasn't all that high. Now the average level of the professional recording engineer, not counting home productions, is really quite high. (Murmurs of agreement from the others.) And I'm exhilirated by it. We all really need as high a level as possible. (Apropos worrying what tools one can afford or should use) wouldn't you rather hear great music recorded to cassette rather than not-great music recorded to whatever?
DM: I don't think any of us know what to do with surround. We're all struggling. Just put some speakers in the back of the room and experiment.
TB: We're only a few years ahead of you in terms of experimenting. You can catch up immediately. Just try it. Experiment with surround on a demo level.
BL: It is early on. Today is the first time I got to hear a speaker in the ceiling. And my first reaction was "Curse word! It really makes a difference!" And I was behind the rear speakers.
DM: It's nice to be able to put a full-range sound up there. Also, I have found that creating a phantom image between the front and ceiling speaker can be very nice. Elevating the sound stage 5-10 degrees (for example). (As for the question of how many speakers is best) five is not nearly enough.
David Griesinger of Lexicon, from the audience: How many channels you need is one question. The real question is how many loudpspeakers do you need. The channels-to-speaker ratio need not be one to one. It's a big unanswered questoin.
DM: Four channels may be enough, a la Ambisonics. Eight channels and twelve loudpseakers might be plenty of room to grow.
(The question arose of how best to archive for the future. Bob Ludwig noted that there are now five competing MO drives, all incompatible. He recommended an analog pass. Conrad White of Harvard reported "a lot of shed from analog in our archives." David Greisinger said, "The Library of Congress has gone to DAT." David Moulton said perhaps DSD will prove appropriate; "Good for at least two decades?" Mark Parsons closed the evening remarking that no doubt some way will arise for our recordings to survive another century, to inspire and amuse the listeners of that era.)
Presentation: DOLBY MULTICHANNEL AUDIO, with Jim Hilson of Dolby.
We didn't manage to record Jim's Power Point presentation, which he presented to a full house. He thoroughly covered the nuts and bolts of Dolby's current multichannel systems -- how they developed, what is coming for them, how best to make use of them, etc. We have a wealth of Dolby info at Parsons, so you can contact us with any questions. (We are the Dolby dealer for New England.) Or visit Dolby's fine Web site.
Open Discussion: AUDIO FOR VIDEO -- WHAT IS COMING?
This was an open discussion of what's coming to audio-for-video, especially the audio implications of digital television, facilitated by Tom Bates, Jim Hilson of Dolby, and Mark Parsons. Here are notes:
MP: The audio aspects of developments in audio-for-video are being driven by the video and computer industries, which apparently could turn out fine. For example, as was more than a little evident at the surround workshop yesterday, surround sound will bring audio producers and engineers into a very much more active and key role than we've ever played before, perhaps, with the artists and talent that we work with. ... What are you doing about the coming of digital TV and its audio ramifications? And what are your questions?
Conrad White, Harvard Univ.: Some practical questions about one aspect: How good is surround sound if it isn't loud? If the space is small? If the speakers are small?
TB: Those recordings are mixed in small rooms; always with the consumer in mind. ... It works better in small rooms. Ideally in rooms with some kind of diffusion -- bookshelves, etc.
MP: On the whole obviously the object environment for digital television will be home theaters of one kind or another -- in living rooms, etc.
DH: Feature films are mixed for large spaces -- theaters. But mixes for home theater are done for small rooms, where they work very well.
Monte McGuire: One concern I have is how many formats of picture will we have to deal with? Computer video? Compressed in varying degrees? On disk; digital or analog tape?
Andy Munitz, Sony Professional Audio: Just one aspect of that is what will happen at TV stations. Sony, for example, has a new HDCAM product line -- a series of switchers, cameras, decks, etc., about 30% up in cost from prior technology; putting high definition into a single camcorder, etc. Something like that is what broadcasters will end up migrating to. ... One consideration in that is worldwide distribution -- a concern about loss of quality going to various distribution formats . HDCAM, as an example, downconverts beautifully. .. There will be a lot of audio channels too -- 4, 8, even up to 32.
TB: Basically there are two kinds of production. One is where your audio person is there doing the audio -- integrated into the system. A concern then is whether you're using digital or analog audio? Do you have word clock that needs to be locked to video black? You need to be absolutely synced. The other situation is Post, doing music, ADR, or Foley. In that case I'm getting picture with code. I come up with music, feed them stems, then they'll do final mixes. Then I can normally specify the format that I get. I can ask for picture in whatever format -- VHS with time code on the FM track, for instance. If you don't want to buy one of everything, just be vocal with the video people to get the format you want. Video people are good at doing that.
MM: There are lots of DAWs now that can handle picture in QuickTime format.
TB: We are poised for more of that as we get the bandwidth and storage capacity, and as more capable systems get cheaper. ... Leading edge stuff always gets you advantages -- gets you some work, perhaps, before it becomes more commonplace.
AM: I think some of the things you're still going to have to pay attention to are issues of long format, time code, film-to-tape transfers, 10% pull-ups/pull-downs, going back to high def, going back to DAWs.
TB: Yes, if you're mixing film and video together to any degree, and you're in audio, you'd better know all about pull-ups and pull-downs, time code, synchronization. I learned about that the hard way.
Nancy Byers-Teague, Dolby Labs: SPARS has an excellent little guide, the Time Code Primer, for I think $19, which takes you through time code, pull-ups and pull-downs, and so on.
Adam Rosen: I'm concerned about when things aren't perfect in surround systems. I would guess that many home systsm aren't well set up. What should you check in surround mixing, so as not to make it worse in a non-optimum consumer system?
MP: As we were saying yesterday, everyone working in surround is necessarily pretty much a begiinner. There aren't that many home theater systems; certainly no firm sense of a consumer standard to use as a reference. No sufficient sense of control.
AR: Some consumers just want to hear something in the back -- that's their test. That's how they set up their system.
Phil Adler, Fox: Lots of customers don't know how to set up. As professionals we have to be sure the mix is proper -- the balances correct, that it can collapse OK to mono. We have no control of how end users set up. ... The manuals of those systems are important.
MP (question): Phil, you have been working in surround for a couple of years now. What do you think about?
PA: When Fox went on the air with surround one of the first things they did was go to the stations and pull out the stereo synths, and the phase chasers, and as much processing as they could so that when the signal was delivered from the network to the affiliates, it was as close to proper as possible. What you don't have control over is when the cable companies get to it, and everybody else down the line. All we can do is deliver it as best we can -- set up at the event using Dolby guidelines, mix acordingly, etc. ... I come home and listen on my cable system to the mixes we do and they are absolutely stepped on. Compressed video, compressed audio.... I know that's my cable system; every one may be different. They all don't care about how it sounds; they just want to get the level right for commercials. ... I just went to my neighbor's house. I took my meter. He wanted to know how to set up. What do you think I found? As we said a minute ago, he had the attitude that, well, there are rear speakers here, so we gotta have plenty of signal out of there. And same thing with the subwoofer.
TB: You're talking about distribution systms, and what they do is criminal. At the consumer end you do want manuals for people who want to set up right. But ultimately we cannot set standards for everyone. You can't get rattled by that. You have to mix to a standard so that the very best listeners have a shot at hearing what you did.
Jim Hilson: The real fault is salespeople -- at the stereo store, etc. You get someone who doesn't know what he is talking about. They don't know the proper way to set things up. We have tried to train them. We trained a whole bunch of people at one of the major stereo store lines and six months later the problems are back. There is constant staff turnover. We need to get those people trained; need some accountability by the store owners to say I have to get quality salespeople here to serve my public.
MP: Perhaps now we're going through a learning curve about that. Hopefully at a certain stage even the most inept salespeople will be OK.
JH: We made a training video for stores. It is a step in the right direction, we think.
PA: If we audio producers do it right to start with, as consumers catch up they'll be able to go back to old material and find that it sounds good; that it was done right.
MP (question): Turning to standards...WGBH, what do you see coming? What are your concerns?
Gerry Field, WGBH: I work with the Descriptive Video Service at WGBH, a service for blind and visually impaired people. We have been on broadcast, cable and home video for the last 7 or 8 years; we add additional descriptive narration. One of the things we're interested in is the audio possibilities that are at least accommodated in the standard, though not implemented so far. Concerning alternate audio tracks simultaneous with the multichannel sound mix, for example, receiver manufacturers are now looking at making the receivers less smart than we would like. The original idea was a dual decoding scheme -- the full 5.1 mix delivered plus the option of additional narration. ... The biggest question for me in production work is how many channels? What format will I be using? At what point will we be compressing and encoding? Stay discrete and uncompressed how far down the road? And what will I be delivering things on? My mind boggles when I think of what the tape floors at TV stations are going to look at. How will networks handle transmission -- what compression level -- from the network to homes, and what about the affiliates? This has real implicatons for simultaneous multiple language delivery. The reality, it appears, will be considerably less than what is possible within the standard. ... We'd like to be able to give our target audience -- the blind and visually impaired -- a complete 5.1 mix. Our effort now is to get that issue back on the table.
As for DVD, that spec is a little different.
DH: Yes. The Selena disk is the first disk I've seen with three 5.1 mixes, in three languages.
MP (question): What are the TV stations anticipating. How worried are they?
AM: In New York, where I work, the networks are getting going. By and large stations in middle to larger markets are at least slightly freaked. As ever, audio takes the back burner, of course.
MP: Around here on the whole we find the stations preoccupied with concerns about the video and transmission aspects of digital television, instead of audio. But the big change in audio will come. It's going to hit them like a hammer.
Richard Bock, WGBH (where he does mixing to picture, mostly long format): In the spring Iwas very worried about all this. This seemed like a snowball that had started its trip down the hill. But now, not so much, with all the stories about stations not going to high definition. It's tempting for TV stations to put four simultaneous channels on instead. ... What was bothering me was...what will high def TV look like? Will all high def TVs have all the loudspeakers? Where is the narration going to go? Into the Center channel? What if the customer hasn't plugged in his Center speaker? I read the AC-3 standard. There's lots of possibility for an intelligent sensor in the receiver to determine what is coming to the system --
MP: And flash an alarm in the living room if the Center speaker is not hooked up? (laughter)
TB: Part of the manufacturers' task is to get us as good a quality as possible out of that Center speaker, since the TV will be located there too.
RB: This is a Dolby question.
JH: We have some stringent guidelines -- building receivers, using Dolby surround decoders, etc. ... There is a requirement that a system either use external speakers or use the speakers in the TV for Center. If you don't have a Center external speaker you must choose between creating a phantom C image with L&R external speakers, or use the speakers in the TV. For surround speakers, you can ignore their information, or downmix it and listen to it in the L&R spkrs. A mono decoder will combine it all. ... Some older shows will only transmit stereo; the decoder will deliver that. So, no, they won't lose Center channel information for systems that don't have a Center speaker. ... Similarly, when it comes to the LFE (low fequency channel) you can set the system up to tell the receiver that you have a small Center speaker. I can select the crossover frequency for my bass spkr -- 80, 100, or 120 Hz. I can reroute the bass to a subwoofer if I have one, or to full range L&R speakers. It will depend on what the receiver manufacturer implements. But all systems have to be able to handle the surround and Center and bass. The smart manufacturers write user-friendly software, not bit-head software, to make systems easy to set up. ... With the manufacturers, we try to think of all the possibilities. For example, every DVD player that you buy includes a 2-channel analog output on the back, to connect with high fi systems.
AM: Homes are not set up for media rooms. Speakers are going to end up where they are least obnoxious looking.
PA: But everybody's audio consciousness will go way up. Surround sound is coming, with even the cheapest systems. I don't mix for the lowest (non-surround) common denominator any more. People are more aware. They may not be able to articulate, but they know it when they hear it. The more this new stuff catches on, that's gonna drive it.
TB: These changes take time. I remember that the transition from mono to stereo took a number of years. But it's likely to be quicker in this case. The installed base of home theater systems may be larger than many people think. I measure that informally: There are a number of magazines on the stands that deal with that subject; they're no longer only on the best stocked stands; somebody is buying those magazines! My intuition is that people who would buy those either own a system or are thinking of buying one within six months. They will all have 5.1 spkrs in a room... My guess is over a million systems are out there right now.
MP (question to Richard Bock): Richard, what were you worried about back when you were worried? Concerns that will surely revive.
RB: We converted to mixing in stereo in recent history. Producers' first question to me was, well, do I have to use two tracks of audio for that? They asked why do we need stereo? Fortunately PBS said, well, because we're now broadcasting in stereo. In that sense going to 5.1 will be same thing. ... We'll have more tracks for storage. In documentaries frequently the only thing stereo is the music. I'd love to have more stereo location sound work done. Producers are reluctant to do it. The location sound people are reluctant to do it too. They don't understand it; it's risky. People are used to slapping a lav on someone or use a boom, but it's not stereo.
MP (question): You'll want to encourage those people to record in surround, then -- to get ambiences, to get live music into multiple channels...?
RB: Yes. But the first thing to come in surround will be (post-produced) music. ... And more stereo library sound effects... There's a tremendous amount of inertia to overcome.
MP (question): Are you excited about what you will be able to do if you can get surround to work with?
RB: Sure!
MP: Having seen so many productions on Ch 2 (WGBH-TV) which you have worked on, and been thrilled by the sound, when people start to hear your productions in surround -- ! When the consumers of the world hear what you and people like you can do, I believe it will drive other stations to get into it as fast as possible -- over a period of some years, sure. It seems there is a certain inevitability to it.
TB: I'm now planning a video production in Nepal. Part of what I want is cameras moving through the streets of Katmandu and complete surround sound. We need a single mic that will image reliably and record onto the number of tracks I have available. Obviously the music that will accompany this will be in surround. That excites me. ... For location record machines I often use a DA-38 mounted in a case with mic pre's, inverter, etc. That's pretty heavy and clumsy. There is no elegant answer. ... As for miking technique, the MSM (dual MS miking) technique I spoke of yesterday, with an MS facing forward and an MS mic facing backwards -- doesn't sound as satisfying as I'd like, but a virtue is that it can be recorded onto 3 channels. You want the "gee whiz" factor to be huge. And if you want to get into the marketplace early, you look for opportunities like this. Maybe I'll need 6 or 8 channels. ... A lot of people don't want to double their work load. It's not public-driven yet. ... Audio for film and video production will be fairly predictable. 20-bit converters will be fine. 48k sample rate is wanted (96 easy to decimate, if that's what you get). You'll lock to 6 channels. ... For facility changes, you will be handling all your audio in digital format -- digital in trucks, etc. You'll have word clock in addition to time code and video black. The changes aren't really big. ... As for cost, it's no money in comparison with video costs. .. Our end of this transition will be much easier than the video aspect. .. At the moment the DA-88, timecode intelligent , is the recorder of choice in this industry.
N B-T: Not all documentaries are going to be in 5.1 to start with. Its all OK. Movies will be first to broadcast in 5.1. The rest will come along later.
TB: Sure. For spoken word we'll even want mono. Watching the news anchor I don't want to hear the sound of cameras behind me. (Laughter) ... Those of you who do production... for documentaries, yes, you'll want surround. For sports you want to feel that you're in the crowd. For music -- symphony broadcasts, etc. -- you'll want to model the hall, etc. That's how surround will start. ... Of course documentaires are always underbudgeted. So that can be a problem.
GF: I keep hearing that multiple standard will begin it. HDTV will at first just be the Sunday night movie of the week, and special occasion broadcasts.
Don Puluse, Berklee: Quad didn't succeed. I'm just wondering about what success we might find piggybacking off video. Might it fail to happen?
TB: None of us in audio are driving this train. Film is driving it. Without film I don't think there would be more than a few multichannel listening systems in the country.
DP: I remember when stereo came, (record mogul) Clive Davis said all records from now on must be in stereo. We (engineers and producers) said, "What a jerk!" We were doing all our work in mono.
JH: Things can change pretty quickly. Up until 2 years ago -- we have had 5.1 in movie theaters for 5 years now -- for the first three years 5.1 got 5% of the production time; Lt/Rt got 95% of the time and budget for the film mix. Then 2 years ago everybody realized that regardless of whether it's going to be Dolby, DTS or Sony, it's going to be 5.1 sound systems in the theaters. Then the numbers flipped overnight to 95% 5.1, 5% Lt/Rt. It's to the point where some of these movies spend all of their time on the 5.1 mix; are in big hurry to get the product out. They take a 5.1 channel mix, assign the 2 surround channels together, drop them down 3dB, and pass the thing through and say, "We don't care what the Lt/Rt sounds like any more. That happened because the successful theaters are all using 5.1. Lt/Rt is no longer the majority of systems.
PA: Theaters had to deliver something you couldn't get in the home -- big screen, surround audio. Now that there are some home theaters out there, the theaters still need to get people to come to the theaters, so now we're seeing big effects (explosions; big, dramatic visuals, etc.) that you can't get at home.
TB: Every proposed standard for the future has 5.1 in it. For the next 10 years we know what our job is!
END OF REPORT
MP, 11/97
to top of page
back to learning
Copyright 2002, Parsons Audio. We welcome your questions, comments, and contributions >>> Webmeister@paudio.com.
|
|
|
