The Music Technology Pages Return to MusicWords Home Page

[Note: It's funny how your mood can change your perceptions. I now feel that this essay is almost embarrassingly grumpy and negative. Still, I have to admit that, as far as I'm aware, every observation in it is factually correct, and I can't actually disavow any of the opinions. Maybe I ought to stop taking so many antidepressants.]

Sometimes I think about Beethoven. As I sit in my cramped but not entirely modest home studio, with half a dozen rackmount multitimbral synthesizer modules crowding my right elbow and the PC-based multitrack digital audio recorder looming up on the left, I think about how 18th and 19th Century composers had to struggle to get adequate performances of their works. No bassoons this month, so we'll substitute (shudder...) an ophicleide. Egotistical soloists inserting their own show pieces between movements of concerti. Audiences gossiping, eating, wandering in and out.

And Beethoven was one of the lucky ones. Not only did he have a reputation that would sell tickets, he lived in a city where ordinary people knew and cared about "serious" music. Classical composers today have trouble getting their works performed at all.

I can't help thinking Beethoven would have given a great deal for access to the kind of music technology I take for granted. My palette of sounds is far wider than any orchestra's. I have control over every nuance of the performance. I can audition a work in progress, so I never have to guess whether a melody or an orchestration is going to work. And when I'm finished, I can press a few buttons and burn a CD that listeners can enjoy at any hour of the day or night: No logistics of rehearsal to manage, no inadequate concert hall acoustics, no fluffed notes, bad intonation, or sloppy entrances.

So we've reached a musical Paradise, right? Well, not quite. Truth is, there's a big fat snake slithering around in this garden. Far too often, the process of creating music in a project studio like mine is a parade of headaches, frustration, compromises, and wasted time. At best, it's a tedious and fiddly process. At worst, it can turn into a nightmare.

The name of the snake is Technology. Like the original snake in the original garden, it promises wonderful things, but what it bestows on those who allow themselves to believe the promises is a very mixed bag.

I'm sure you could make your own list of problem areas -- places where technology has let you down, or worse. You don't need me to point out the obvious. But sometimes it's useful to take a fresh look at where we are and what's going on. Maybe you'll find it gratifying to read about someone else's frustrations. Maybe you'll find it heartening to think, "Oh -- it's not just me! Other people go through that too." Maybe you've been so busy beating your head against a brick wall that you haven't yet stopped to notice why you're not getting anywhere, and taking a fresh look at the situation will inspire you to take three steps to the left and open the door. Or, just possibly, if I articulate a problem in a fresh way, maybe some clever lad or lass will come up with a solution. I'm happy to play devil's advocate. Even if I'm wrong about the details, maybe I'll be able to help stimulate some useful discussion.

For whatever it's worth, then, what follows are my assorted ruminations on what's wrong with music technology in the seemingly auspicious year 2000. I realize this is a sensitive issue. From time to time I hear rumors that one or two manufacturers' representatives consider my product reviews too negative, my attitude too cynical. Once in a while, such a person requests that my boss give a product to "somebody other than Jim" for review. I have even been accused of bias. For the record, I do have a bias. I'm biased in favor of wonderful products -- products that sound great, perform flawlessly, and are designed with a clear understanding of musicians' needs. This is not the forum in which to make a detailed explanation of how I approach product reviews, so I'll let it go at that.

I need to make it absolutely clear, before we go any further, that I'm not writing this essay in order to attack manufacturers, much less any particular manufacturer. In developing and marketing viable music products, manufacturers face incredible challenges, both technical and economic, and the results they achieve are often not only credible but inspiring. If I seem a little too enthusiastic about pointing out a few of the shortcomings of assorted products, I hope you'll put it down to my desire to air the issues frankly rather than tiptoeing around them.

Many of the problems discussed in this essay are not the fault of manufacturers at all. Sometimes the universe just has these little kinks in it as it unfolds. Sometimes the economic imperatives and sheer shortsightedness of consumer-driven capitalism are to blame. And sometimes it's our expectations of technology that suck; the technology may be just fine.

Having said all that, I'll cheerfully admit that there are times when pointing the finger at manufacturers is probably appropriate. Beyond the borders of the rather insular world of Linux, the far smaller community of Max/MSP developers, and the back rooms where the modular synth folks are wiring together their circuit boards, music products are not user-customizable. The buck has to stop somewhere. Bad design decisions are made, and not only made but defended, ofttimes with great vigor. Optimism and self-interest blind otherwise intelligent people to flaws that could be dealt with early on, flaws that later turn into extended disasters. Corporate egos ride roughshod over common sense.

With no more than a handful of exceptions (and no, I'm not going to name names), everybody in the music manufacturing community is doing the very best they can to make musicians' creative lives easier and more fruitful. We're in this industry because we love music. But between the ideal and the manifestation falls the shadow....

Too Many Choices

The ten synthesizers in my studio (admittedly that's too many, but I'd hate to give any of them up) average more than 200 presets each. That's well over 1,000 sounds at my fingertips. When it's time to add a chord track or a fill to a sequenced tune, which sound do I use?

Hmm, maybe I should use something from a sampling CD instead. Almost forgot about those, because they're more trouble. You have to listen for half an hour, hook up cables, truncate the waveform, assign it to the keyboard, all that rigmarole. But maybe there's something magical tucked away on one of the CDs in the stack, something that could raise my tune to a whole new level.

Even when he wrote for a full symphony, Beethoven had fewer than 50 sounds to choose from. He scarcely had to think about it: "Okay, this phrase belongs to the winds. Let's give the melody to the clarinet for a change." With that choice out of the way, he could think about the music itself.

When I'm poking around among the presets, I often have no clear idea what sort of part I want to write. I just want to fill in a hole in the arrangement. So I grab a patch, fling my hands at the keyboard, recoil in disgust, and grab another patch. This process continues until the moment when there's a happy surprise. "Ooh, I like that. Let's record it."

In essence, I've ceded my creative decision to the machinery. Rather than having an idea and developing it, I've allowed the silicon magician to force a card into my hand. If I do have a clear idea what sort of sound I want to use, finding it amongst all these memory banks is likely to take so long that the inspiration will be gone. True, there are well-known classes of sounds. If I want a Lyle Mays lead synth or a Jaco Pastorius fretless bass, I know where to find one. But it may not be quite right for the track. I may have to do a little programming. And at that point, my choices multiply exponentially. Add a second oscillator? Change the filter mode? Longer envelope decay? Faster LFO? More reverb? Less reverb?

I know what you're saying: "Why doesn't he have all his synth sounds categorized in his editor/librarian? He could find the right program in 30 seconds." Well, maybe you're not saying that, but two or three software marketing managers of my acquaintance are saying it. Here's why: First, I have more interesting things to do than spend ten or twelve hours adding category tags to a thousand sounds. Second, and more relevant to the topic at hand, the editor/librarians I've looked at aren't really happy unless they're operating in an environment where there's bidirectional MIDI communication between the computer and every single synth. As it happens, I don't have a 16x16 MIDI interface. So creating a database of sounds would involve a flashlight and a mirror, those being the tools with which one identifies a MIDI out jack on the back of a rack module in order to cable it back to the interface.

In the real world, then, an editor/librarian wouldn't actually make my creative process any easier; on the contrary, it would complicate the process. Someday editor/librarians may be smart enough to inquire of synthesizers what category their ROM and RAM sounds belong to. Assuming we can all agree on the names of the categories ... no, I don't want to think about it. The point I'm groping at is this: No matter how powerful a piece of technology may be, if it makes unwarranted assumptions about my equipment setup or my approach to music-making, it sucks. (I'd say the same about a sequencer that assumed my tunes were all in 4/4 time. Thankfully, we've progressed beyond that point, at least in the software world. Or have we? As of version 2.0, Acid has no provision for time signature changes.)

So half an hour later, I've recorded a new synth track with the sound I stumbled onto, but I'm still not done with the choices. Maybe it's a little too prominent, or downright brash. Gotta fix that. Shall I lower the filter cutoff, or reduce the velocities in the sequencer, or change the EQ on the mixer, or lower the volume? If I lower the volume, shall I do it at the mixer or with a MIDI volume change command? And what about the rhythm of the track? Shall I quantize the part, or quantize it with 50% strength, or re-record it to get the groove tighter, or edit a few of the more aberrant note start times (and maybe note durations too) by hand?

Beethoven didn't have to worry about any of this. He could write "dolce" and a few staccato marks over that clarinet part, and his job was done. True, when the piece was being rehearsed he might throw a tantrum because the clarinetist wasn't doing what he wanted -- but even then, he didn't have to guide the player through the part one note at a time. He could indicate his displeasure in more general terms, and trust (or hope) that the player would see to the details.

In any case, the music was down on paper. It wasn't affected by a bad clarinetist at the premiere performance. A hundred years later, we can still enjoy the music. If you're adept at reading scores, you can even enjoy it without hearing it. Try that with an audio waveform display.

Unless I bring in some musicians to do overdubs, I have to play every instrument in my "orchestra" myself. This has its advantages, especially for a control freak. On the other hand, playing with real musicians would improve not only my recordings but my social life. Still, it's a decidedly non-technological solution, and I'm not 100% sure I have the energy to tackle printing out legible parts. (Notation software. Let's not go there.)

Brian Eno has suggested that it would be useful to have a synthesizer that only made a few interesting and useful sounds. I'm not sure I'd go quite that far, because I'm not impressed by the quality or playability of the presets in most of the electronic pianos I've tried. But maybe if somebody built a great-sounding synth that had a Lyle Mays lead, a Jaco bass, a TB-303 bass, a Clavinet, a luscious string pad, and another dozen sounds, all with perfectly tailored realtime controller inputs, I'd be happy to use it. Only then they'd bring out ... the Expansion Boards. And I'd be back where I started.

Anyway, I doubt anybody is going to market such an instrument. Every keyboard player has slightly or radically different taste in sounds. Manufacturers have to pack their synthesizers with a variety of presets to sell as many as possible. And they have to make them programmable, even though most musicians do only the most rudimentary programming, because there's a stigma attached to preset instruments. We want to be able to program them, even if we never do.

I'm not complaining about having an embarrassment of riches, not really. But I do think the fatigue factor that sets in when you've been searching among the presets for an hour gets in the way of music-making. Plus, as noted above, all too often I choose the right sound for a part instead of having a creative idea.

I know I'm not alone in this. I hear CDs all the time, sent to the magazine by aspiring synthesizer players, that are full of lovely sounds, impeccably mixed, and yet largely devoid of musical merit. The ideas just don't speak to me.

We can look at the bright side, if we like. Here are thousands of somewhat talented musicians who would never be able to create their own music if it weren't for synthesizers, samplers, sequencers, and digital recording. Or at least, it would be a whole lot harder for them. They'd have to study arranging and orchestration with a teacher and then find a local band willing to play their charts. They'd have to work their brains instead of pushing buttons at random.

Some people just can't do that; it's too hard for them (or more likely, there's a bad parental voice in their head telling them it's too hard for them even though it isn't). But those who stuck with it would learn a lot. They'd interact with more knowledgeable musicians, and they'd get useful -- even inspiring -- feedback. Writing for synths and samplers doesn't teach them nearly as much about what works musically. They have the illusion that they've created something that other people will enjoy listening to, when most likely they haven't.

What's worse, their listeners may have the illusion that they're hearing something that's musically profound. Synthesizer music made by inadequately trained but technologically savvy musicians is like fast food: It's vaguely unsatisfying, but you can't quite put your finger on the missing ingredient. If you're having that reaction, it's because there are no spiritual nutrients in the music, just empty sonic calories.

This is actually a separate issue, now that I think about it: Technology sucks because it makes it easier to disguise bad music. Credit for first noticing this issue goes to Bob Doerschuk. Back in the early '80s, Keyboard had a yearly reader tape competition. After plowing through a stack of cassettes one afternoon, Bob said, "You know, in the old days you used to be able to tell within five seconds if a band was no good, because the drumming sucked. A good drummer would never waste time playing with bad musicians. But now, thanks to drum machines, any musician can have a drummer who keeps perfect time." His point being, with the advent of drum machines you had to listen longer and harder to pick up the telltale whiff of bad songwriting and bad arranging.

Sampled loops have made the problem ten times worse.

Unrealistically High Standards

Ever since the 1920s, musicians and their audiences have been mesmerized by the recordings of great artists. Long before punch-ins and tape splices became feasible, long before pitch correction algorithms created an artificial level of perfection that few real singers could ever hope to achieve, we flawed, limited musicians were already measuring ourselves against a painfully high standard of performance created by a few world-class virtuosi.

When this happens, the focus in music-making moves insidiously away from artistic sincerity in the direction of mechanical execution. That's why piano competitions are full of technically flawless players with safe, cookie-cutter interpretations. It's why having a radio-ready mix is more important than being able to play an impassioned solo. Technology sucks when it encourages us to focus on the wrong things.

Rigid, Inexpressive Instruments

When MIDI was introduced in 1983, a number of nay-sayers complained that it wasn't good enough. They were right. It wasn't. It still isn't. The bandwidth is so narrow that you just about can't force a truly expressive musical performance to travel down a standard MIDI cable. Plus, MIDI is keyboard-centric. It doesn't deal at all gracefully with the kinds of articulations or pitch gestures you can make on a violin, for instance.

But the real problem with MIDI isn't in the Specification. There are ways to fudge the Specification -- controller LSBs, high-bandwidth Firewire connectors, and so on. The real problem is at the receiving end of the cable. Considered as a class of musical instruments, MIDI synthesizers are about as expressive and sophisticated as a 2x12 pine plank.

(In case you came in late, I should reiterate that I own rather a lot of them. I compose and record music with them.)

MIDI synthesizers are inadequate for several reasons. Most of them have poor performance interfaces. Their tone-generating circuitry often lacks important dimensions of control. Many of them are badly constructed. And all of them become obsolete far too quickly.

Leaving aside the velocity-sensing keyboard, the performance interface of the typical synth hasn't improved much since the Minimoog. There was a long stretch in there where the interface got worse, in fact. Remember the no-knobs era? We're getting more knobs and sliders these days, and many times the knobs are assignable to various sound parameters, which is wonderful. So why is the resolution of the typical synth knob/slider limited to the 128 data increments allowed by MIDI? This isn't enough resolution; control movements with knobs and sliders tend to sound grainy and/or stairstepped.

The resulting sound is fairly unmusical (unless you happen to like grainy and stairstepped). It's also unnecessary. Even if the instrument is going to transmit only 128 values over the MIDI cable, there's no technical reason why it can't sense, and utilize internally, 512 increments or even more. If the sensor attached to the knob is limited to 128 values, the voicing software can still be written in such a way as to interpolate more smoothly between adjacent values.

To add to the problem, the knob output values may not be updated often enough. Move the knob quickly from one end of its travel to the other, and you may not get a stream of 128 data values. You may get only 37 different values, or 11. I haven't tested this observation on a lot of the newer synths, but I know for a fact that the pitchbend joystick on my Korg 01/W ProX (not a cheap instrument, and not first-generation technology) transmits only eight or ten discrete values when you move it quickly through its full range. If I want a smooth bend, I have to draw the data into my sequencer with the mouse. (See below for more on the mouse.)

Then there's the keyboard. Synthesizers are built with keyboards not because the keyboard is an ideal performance interface but because there's a huge pool of keyboard players who will buy synthesizers. I'm not sure the keyboard needs more than 127 possible velocity output values, but a smoothly operating, reliable mechanism that will let players who are willing to put in some practice time use all 127 values is almost unheard of.

Aftertouch sensors are a joke. Either you have to lean so hard into the keyboard that you risk wrist and finger injury, or they spew out gobs of data when you play normally, clogging up the MIDI data stream and potentially adding unwanted vibrato or filter sweeps to every note. I happen to have an answer to this one, if anybody is interested: Write the synthesizer's OS in such a way that it will filter out aftertouch data that occurs within the first n milliseconds after a new note-on event (n being user-programmable, of course). At the end of this time interval, interpolate smoothly between the previous aftertouch value and the current one. Combine this software with a sensitive aftertouch sensor, and you have a more playable keyboard.

We could talk about how there ought to be keys that sense side-to-side motion, in-and-out motion, or finger position. Some of these ideas may lead to useful performance techniques, and some may just be gimmicks. More significant is the physical layout of the keyboard itself. In the mid-'70s a young man named, if memory serves, Paul Vandervoort built a close to ideal keyboard, adapting and improving the ideas of the 19th Century von Janko keyboard. Its advantages include uniform scale and chord fingering in all keys, the ability to reach larger intervals comfortably, and less chance of hitting a wrong note. With the advent of MIDI, there's no reason, other than economic inertia, why the Vandervoort keyboard or something similar couldn't be mass-produced.

It won't be. For the foreseeable future we'll continue to buy, and use, MIDI keyboards that suck. (A few years back, a Japanese company brought out a von Janko-inspired keyboard, but they made at least one huge mistake. Maybe more, but this one was visible in the photos: They made all the keys white. Vandervoort quite sensibly retained the existing black/white grid, thus making it easy to see what key you're in.)

Now about sound. I'm still a big fan of Yamaha's FM synths. FM patches can sound incredibly expressive. There were several reasons why FM never achieved its potential. First, it was proprietary technology rather than open-source, so when Yamaha dropped the ball, nobody else was allowed to pick it up. Second, the phrase "cheesy FM" got attached to cheap soundcards, and it infected the rest of the market. Third, irrespective of the somewhat constricted nature of the DX7's programming interface (a dead horse that I'm not going to beat), good FM programming requires a knowledge of acoustic instrument theory that few musicians possess. Fourth, sample playback synthesis happened. (See below for more on sampling.) Samples can sound more realistic than 6-operator FM, and keyboard players got hypnotized somehow -- maybe they hypnotized themselves -- into thinking that a synthesizer sounded best when it sounded like something else. Weird idea, but then Ronald Reagan's entire presidency was one long weird idea, so why should the triumph of sample playback synthesis surprise anybody?

With some of the newer "modelled analog" synths like the Waldorf Q and the Novation Supernova II, we're finally getting back to the level of sonic sophistication that FM offered 15 years ago -- though for the record, of the current crop of instruments only the Clavia Nord Modular is capable of true 6-operator FM. The static 2-op FM in the standard-issue virtual analog synth is not going to win any converts.

In the intervening years, we've been saddled with many, many instruments that, in retrospect, sounded pretty darn awful. (Listened to a D-50 lately? I sold mine.) The awfulness was disguised somewhat with chorusing and reverb, but it was still there.

In a conventional sample playback synthesizer, you just can't do much to make the tone genuinely responsive. You can add vibrato (under the control of a monotonously regular or, what's worse, randomly irregular LFO) or sweep the filter. And let's not forget amplitude modulation from an LFO or aftertouch, and maybe effect output level. But that's just about it in terms of the dimensions of expressive control. The ability to modulate individual envelope segments, if it's included at all, is generally implemented in such a primitive way that you can't rely on it for any kind of subtle expression.

Besides, what are you going to control with the envelope? The filter. Digital filters are getting better, but basically a filter subtracts overtones, which limits its usefulness. Yes, you can modulate the filter cutoff with an audio-range wave -- but that's FM. Which proves my point.

With a high-quality FM instrument, polyphonic aftertouch, and maybe a couple of spring-loaded knee levers, it would be relatively easy to create a synthesizer on which one could do wonderful, expressive performances. But there's one other essential criterion for such an instrument: It would have to have a long enough shelf life that players could refine their performance techniques. And that's not going to happen anytime soon, because the economics of the industry don't allow it. Manufacturers are forced to keep building new models in order to bring customers back into the store. Never mind that most of the new models are nothing but a slightly different combination of the same old features. Change, even meaningless or counterproductive change, is what keeps the economy pumping.

Gimmick Parameters

Most synths have one or two parameters that aren't implemented well. To truly suck, a parameter has to actively interfere with your music-making.

My perennial favorite example is Roland's random panning parameter. They've been including it on their synths for years. It has always sucked, the reason being that it destroys any sense of where the sound is positioned in the stereo field. Technically, it's not the parameter itself that sucks; it's the fact that Roland's factory programmers insist on using it. In order to use that otherwise stunning electric piano patch in a song, you have to copy it to a RAM location in order to defeat the random panning.

Not a parameter, exactly, but definitely a voicing feature that sucks bigtime: velocity cross-switching. It can be done well, but most of the time it's done badly. More often than not, the bottom drops out when the fortissimo sample kicks in. I'm not sure whether the culprit is inappropriate normalizing on the raw sample data, or whether the lack of overall dynamic range in the synth is at fault. Either way, velocity cross-switching sucks. With badly matched samples, it's impossible to control your articulation from the keyboard, and even editing the sequencer data after the fact won't necessarily save the track.

When a synth can't fire most or all of its ballyhooed 64 or 128 voices with reasonably tight timing, the timing definitely sucks. But if you're expecting me to gripe about timing latency per se, I'm going to disappoint you. Bach made great music on pipe organs that had, in some cases, as much as half a second of latency (that's 500ms, for you techie types) between key-down and an audible sound. Latency is a problem mainly when it's inconsistent and unpredictable from note to note.

Even when it's consistent, though, latency can be a problem. If you have to fudge the automation data in your audio recorder, for instance, by drawing it on the screen two beats early because of the latency in the audio playback engine, you have my permission to tell whoever asks you that the automation sucks.

Premature Shipping

It doesn't just happen in the music industry, but we're not immune. Sometimes the marketing department makes unrealistic promises and the engineers are forced to back them up. Sometimes the accounting department says, "If we don't have some cash flow in the second quarter, there isn't going to be a third quarter." For whatever reason, a product is kicked out the door long before it's ready.

No, that's being too kind. There are victims here. A product isn't just kicked out the door, it's foisted off on unsuspecting customers, customers who plunk down their hard-earned money on something that they're promised will work at least at the level described on the box and in the ads. And it doesn't work. It's riddled with bugs. It sucks. I recently heard a story about a music software company, now defunct (and no, it wasn't Opcode) that was so desperate to show a cash flow before the end of the quarter that they shipped shrinkwrapped boxes that were empty. The program didn't work; they didn't dare ship it; but they didn't dare not ship it.

The other side of the coin is when musicians put down a deposit on an expensive instrument and it still isn't shipping five or six months later. We can do better, people. I know your company wouldn't do this sort of thing, but it happens.

Sampling

Earlier I mentioned that sampling technology makes it easy for bad musicians to sound superficially plausible. But sampling suffers from another glaring defect: A sample is a rigid entity. Set in concrete. By definition, it's not possible to add musical nuance to a sample.

Since I haven't been sleeping under a tree for the last hundred years, I know it's quite feasible to load a sample into Kyma or Reaktor and do granular synthesis with it. I've heard some wonderful expressive sweeps of sound color that were generated using this technology. But while samples can be used as the source material for very expressive sounds, the expressive aspect of the music consists precisely in mangling the sample, often beyond recognition. To put it plainly, this type of synthesis proves my point. If you have to run a sample through a meat grinder to get anything meaningful out of it, it's because it wasn't meaningful to begin with.

And at the risk of descending into social criticism, what percentage of sample users go to all that trouble? Most people who use samples just slap them into the track, maybe add a little distortion or some filter sweeps, and congratulate themselves on how hip their music sounds.

Sample-based music thrives, in point of fact, on rigidity. Repeating something, be it a drumbeat or a saxophone squeal, eight times without the slightest variation is, in the modern world, the bee's knees, the cat's pajamas. How well this music will stand up to the test of time is certainly open to question. Will we still be listening to it in 30 years, or will we laugh uproariously and pelt the host with hors d'oeuvres if he dares slip a hip-hop or techno CD into the stereo?

Good music will survive, and some of it will use samples. It will be good in spite of the samples, not because of them. It will be good to the extent that it subverts the rigidity of samples, to the extent that it skirts repetition rather than embracing it. Most sample-based music will be, deservedly, forgotten.

Anyway, sample-based music isn't made for posterity. It's disposable. It's made with the ironic postmodern understanding (call it a sense of dread if you like) that all cultural artifacts are temporary. The further we plunge into the 21st Century, the more temporary, the more fragile and evanescent, such artifacts become. If Venice is sinking into the sea, how long will Beethoven be able to keep his waistcoat dry? Using samples is a way of abdicating one's creative responsibility. It's a way of saying to one's listeners, "Look, you and I both understand that this isn't real music. But it doesn't matter anymore."

As an aside, the latest trend I'm seeing along these lines is the compilation CD packaged by a dance DJ. The DJ's name is in big letters on the front -- but this person has made no discernible creative contribution to the proceedings. He or she chose the tunes and figured out how to segue from one to another, and that's about it. Can you imagine a Frank Sinatra album where the A&R man got his name on the front cover, while Frank was listed in small print on the back? The times they are a-changin'.

I want to acknowledge the good side of sampling: With the aid of a soundware CD, I can collaborate creatively with talented musicians whom I've never met. While I'm at it, I can take advantage of their great mics, their racks of effects processors, and so on. That's a fantastic option to have.

But there's a dark side, too: In their quest for fresh audio, sampling CD producers have spread out around the globe, taking their battery-operated DAT recorders into remote villages where local musicians still play in traditional styles. You may not see that as a problem. Indeed, if you're assembling (I won't call it creating) "world music," those CDs may be among your most treasured musical assets. Here's the problem: Most often the musicians who are contributing their highly skilled and heartfelt chants and drumming to the sampling session are not credited by name in the liner notes of the CD. And if they're not being credited, how can we be sure they're being paid?

In simple terms, a "world music" sampling CD can be yet another way in which light-skinned people exploit and rip off the economic resources of dark-skinned people. I'm sure some CD producers are very ethical about paying local musicians in remote areas -- but from the defensive reactions I've heard more than once when I've brought up this issue, I'm fairly sure some of them are not that ethical.

Okay, enough philosophizing. Let's talk about something concrete.

The Mouse

I love a synthesizer panel studded with knobs, or a mixer with lots of long-throw sliders. What I don't love, not even a little bit, is music software whose user interface consists of a bunch of graphic "knobs" or "sliders" that I have to operate with the mouse.

The mouse is a very inadequate input device for almost any type of music data. True, the mouse was a wonderful leap forward in 1980. I'm not advocating a return to the command-line interface. But by now the mouse has been around long enough that its shortcomings have become glaringly apparent.

I have two problems with the mouse, one practical and one musical. The practical problem is that gripping a mouse for hours on end is an open invitation to tendinitis and carpal tunnel syndrome. From which debilities I intermittently suffer. I hate the miserable mouse. And please don't talk to me about track balls; they're worse. The mouse was not invented by or designed for 50-year-olds. It may be perfectly swell if you're 25, or if you use it a couple of hours a week. If you use it all day, every day, you're begging for physical trouble.

Give me a music program in which every single function can be handled by the QWERTY keyboard, and I'll at least stop grumbling for a while. Not that the QWERTY keyboard is a great music performance interface either, but at least my hands don't get numb from gripping it.

Does that sound like a ringing endorsement of stand-alone hard-disk recorders with button-studded panels? When somebody builds one with a computer-sized monitor (or even an output jack for a monitor), a zoomable track waveform display, and a mouse input so I can drag chunks of audio around on the screen, I'll consider using it. Until then, no. The mouse is good for some things, no question.

Nor is it the only problematical piece of hardware ever invented. Many (perhaps most) musical instruments put physical stresses and strains of one sort or another on the body. Why should the computer be any different? Trumpeters, or so I've read, run a greater risk of stroke than most of us, because hitting those high notes puts pressure on the blood vessels in their brains. Violinists suffer twisted necks. Hammond B-3 players get hernias and slipped discs, though not from playing.

If the mouse were a wonderful input device for music, I'd grumble a bit about how my hands hurt and then put a cork in it. The real trouble is, the mouse is an awful input device. You have two dimensions of more or less continuous control -- the X and Y pixel position -- and one or two non-velocity-sensitive buttons. Attach two footswitches to a Theremin and you've got the same amount of control. And with apologies to Bob Moog, the reason the Theremin hasn't been a huge success is not that its tone is uninteresting. The real problem is that its input structure is hard to use.

Music software usually has MIDI inputs, and some manufacturers let you define MIDI equivalents for keystrokes. But except for a few specific commands, such as arming a track to record, MIDI equivalents aren't an improvement, because you have to swivel away from the computer to use the MIDI keys. Why aren't software designers producing programs that use MIDI slider boxes for data entry? I could position one right here, next to my QWERTY keyboard.

One reason, I suspect, is because programmers want to follow a uniform interface design. If they depart too radically from the lowest-common-denominator interface, they run the risk of making their program hard to learn. At least, it may look hard to learn. If the program looks hard to learn, they may not sell enough copies to stay in business. To some extent, commercial survival seems to dictate that everybody use more or less the same awful mouse-based interface.

Beyond that, we need to remember that the computer is a very young device. It's still at an early stage of development. Remember what automobiles looked like in the 1920s? No turn signals, no seat belts, no power windows or power steering. Someday the computer itself may have sweep footpedals, velocity-sensitive function keys, and half a dozen pressure-sensing X/Y buttons that the user can link to various tasks.

Just don't give me a touchscreen and try to convince me how great it is. Or voice activation. There's a technology that really sucks. Unless, of course, you need it because of a physical disability. And speaking of disabilities, if you're blind, practically all music technology sucks. A few manufacturers have addressed this issue, with varying degrees of effectiveness, but most are oblivious. The acid test is, if the rest of us can't operate it blindfolded, from a blind musician's perspective it sucks bigtime.

Computer-Based Music Instruction

Periodically I get a letter from some poor soul who wants to study music and is hoping to find a piece of instructional software that will make the process painless and fun. While I'm not an expert on instructional software, I'm not reluctant to offer such people the benefit of my off-the-cuff opinion. My opinion is, there is no meaningful sense in which you can learn music from a computer program, and anybody who tells you you can is trying to extract money from your wallet, pure and simple.

Don't get me wrong: A computer sequencer is a wonderful laboratory in which to learn about composing and arranging. But a sequencer doesn't try to teach you arranging. When it comes to teaching, computers are very good at types of instruction that involve repetitive drills. You can improve your ability to identify intervals, rhythmic patterns, and chord voicings using a computer program. But a computer can't evaluate a student's posture or fingering. A computer can't suggest pieces that are appropriate for a student's current level of development. A computer can't respond to a student's vague, ill-formed questions with questions that probe more deeply. A computer can't discuss the realities of a practice schedule, or praise the student's tone, or organize a recital.

Actually, what I tell people who ask about instructional software is that they should ask their teacher what software the teacher recommends. The usual response is that they don't have a teacher, and don't want one. They want to use the computer instead of a teacher. When I hear this, I can only sigh. This response suggests one of two things: Either their commitment to playing music is almost nil, and they're hoping the computer will somehow magically allow them to learn music without making a commitment; or they're so fearful of being judged that they can't bear to make mistakes in front of a teacher.

A music instructor that I talked to about this issue told me I was missing something important. Kids love to play with computers, she said. Kids who will turn sullen or rebellious if their mistakes are corrected by a teacher will cheerfully correct their own mistakes and improve their performance when a computer makes its little "oops, you missed" noise, because the computer isn't threatening to them. Learning music has been transformed into a game.

If true, this is extraordinarily sad. What have we done to our children, that they so fear and hate us? And how will they learn what it means to be human -- to be a musician -- if they can only learn from machines?

The Latest and Greatest

The other afternoon one of the editors at Keyboard lustfully pulled the plastic packaging off of his latest acquisition -- a very expensive stereo compressor/EQ. I suggested to him that composing the right melody, bass line, and chord voicings might possibly be a more important factor in the success or failure of his next recording than whether or not he uses this particular toy. He told me he hoped the people competing with him for jingle work agreed with me, because he feels that particular gadget will give his mixes a competitive edge.

He may be right. And if he can justify the purchase by the fact that he's actually making money in his studio, who's to say he's wrong? In fact, the clients he's trying to impress may be much more easily swayed by a punchy, sparkly mix than by the quality of his musical ideas. Fresh musical ideas might even alienate them.

Having decent equipment is important; don't get me wrong. I wouldn't want to be forced to spend the rest of my life working with my 1981 home studio setup, which consisted of a Serge Modular, an 8-track reel-to-reel, and a spring reverb. (Actually, that might be kind of interesting. Wasn't it Stravinsky who said he loved having a commission that gave him definite limits for his next piece?) But there is definitely a point beyond which adding more equipment is not only useless but actually becomes counter-productive.

My concern is that a lot of musicians are seduced into spending a lot of money, money they can ill-afford to spend, on equipment that will make no conceivable difference in their creative lives or their professional prospects. They're wasting money on dreams.

Maybe I shouldn't complain too loudly about this. If all the amateur musicians in the world woke up one fine morning and said, "You know what? I don't need a 128-voice sampler. I think I'll spend the money on some piano lessons instead," a lot of companies whose cash flow depends on the maintenance of wannabe musicians' rosy expectations would find themselves in serious financial trouble. Those companies would stop taking out magazine ads, and I'd have to go out and find honest work.

It's a chance I'll have to take. Here's the point: Technology sucks to the extent that it doesn't deliver what it promises. The greatest rackmount tube compressor/EQ in the world won't make a trite lyric into an emotionally moving song. It won't make a stiff, gawky arrangement into a gem.

The technology doesn't actually promise any such thing, of course. But we're encouraged to believe that it does. We're encouraged to seek the solutions to musical (or, at a deeper level, spiritual) problems outside ourselves, in physical objects, and in never-ending frenzies of consumer behavior.

Like other addicts, the music gear addict always plunges into the low that follows the high. Suddenly and unaccountably, that synthesizer that you knew six months ago was destined to magically elevate your music to a higher plane now sounds weary, stale, flat, and uninteresting. You've gotta have the new latest and greatest thing to give you a fresh rush.

Admittedly, having a set of fresh sounds at your fingertips can be truly inspiring. But why not try learning to program the synth you already have, rather than spending money on a new one? Or give yourself some composing exercises that will teach you new ways to use the same palette of sounds you've been using all along?

Popular music is the triumph of style over substance. If you buy into the idea -- and it may be true -- that having a trendy haircut, hair color, tattoo, and/or facial hardware adornment is going to attract the right sort of attention to your music, why should having the latest equipment be any less important? I can't argue with this way of thinking, but I can (and do) feel sorry for people who have been sucked into it. Thinking you have to have the latest and greatest isn't just an invitation to waste money. It can also be a way of courting disaster. Most of us can tell war stories about the Upgrade From Hell. You know, the new software that was supposed to give you more tracks, more plug-ins, more blah blah blah, but instead toasted your system and forced you to spend weeks rebuilding everything from scratch. I could tell a few stories like that, but the details don't matter.

What matters is this: If your current music system allows you to make music in an efficient and gratifying manner, think long and hard before you rush into an upgrade. As they say in Missouri, if it ain't broke, don't fix it.

Lack of Effective Standards

MIDI was an extraordinary achievement, not for technical reasons but because an entire industry actually managed, for once, to agree on an effective and open standard. This led to a proliferation of creative possibilities beyond anything musicians had ever imagined. So why didn't manufacturers learn the lesson? Why did they immediately go back to guarding their private turf like spoiled little boys quarreling in a sandlot?

No point talking about OMS vs. FreeMIDI. That's old news. Or is it? If the entire music industry had agreed from the start on one open-source protocol for MIDI-to-Macintosh I/O, is it possible that Apple might have supported it in the MacOS? Is it possible that if they had, OS9/USB MIDI communications might not be in quite such a snarled-up state this year?

Yeah, that's possible.

The Standard MIDI File works pretty well. So why have sequencer and multitrack audio recorder manufacturers not yet agreed on a standard file format for songs that include audio? Sure, it's a more complicated data format to define, but it can't be that complicated. Some people would have to sit down and hammer out the details, and that would take time and cost some money.

One reason it hasn't happened may be because the manufacturers don't want to make it too easy for their customers to switch to the competition. Never mind the advantages that such a file format would provide. Never mind that musicians who happen to have several sequencers and/or audio recorders on their hard drive could use each one for what it does best. And never mind that if (when!) a company shuts its doors, any musician who has years of creative work tied up in files created on that company's sequencer is in deep shit.

I'd love to see musicians strike on this point. I'd love to see us band together and refuse to buy any more recording software until a universal file format that includes MIDI, audio, plug-ins, and automation data is hammered out. Just say no. At this writing, there's apparently a proposal called AES31 on the table for file exchange among multitrack audio programs -- but whether it will include MIDI data or plug-ins, and how widely it will be adopted, remain to be seen.

Isolation

Historically, music has always been a communal activity. Players and singers rehearsed together, and then they performed in front of audiences. Technology doesn't have to sever those connections among people, but quite often it does so. Even when you go to a pop concert, there's no guarantee that what you're hearing is actually being played at that moment in time; it may be sampled. If it is, your connection with the performer has become illusory. You think there's a connection, but really there isn't one.

The artificial nature of studio overdubbing and punching in is an old topic. No need to rehash it here. In a traditional recording studio, though, at the very least the player was working in the presence (behind glass) of a producer and engineer. In my studio, I'm entirely alone. The freedom to pursue my own vision without compromise is a wonderful thing, but it's purchased at a high cost: I no longer have that vital connection with other people.

Many musicians who are producing CDs at home today have never performed in front of audiences. Technology makes it easy for them to realize their visions, but at the cost of robbing them of the opportunity to interact with others.

Isolation is a bad thing not merely because it may deprive the individual artist of useful creative feedback on what parts of his or her project excite listeners, and what parts fall flat. Nor am I mainly decrying the lack of moral support, though that can lead to discouragement and loss of interest in music-making, which is a problem too. More crucially, I'm suggesting, the musician working alone risks losing some essential spiritual dimension of his or her art, a dimension that, if present, is bound to inform the creative work in ways that are intangible but nevertheless quite audible. I'm sure many of us can still access that dimension when we're alone. Painters and novelists manage it, after all -- why not musicians? But it's harder.

Non-Realtime Music-Making

I recently returned to my first instrument, the cello, after 25 years pursuing other dreams. The cello is a triumph of 18th Century technology, just as the piano is a triumph of 19th Century technology. Nothing natural about a cello: Just the fact that it's made of wood doesn't mean it grows on trees. But there's an important difference between a cello or piano and a computer-based studio: With a so-called acoustic instrument, you have no choice but to generate your music in so-called real time.

(Reminds me of a snappy comeback made years ago by a Delta bluesman named Big Bill Broonzy. Asked by a white musicologist whether he considered his blues a form of folk music, Broonzy replied, "It's all folk music. I never heard no horse play none of it." The point being, it's all real time. What other kind of time is there? But let's let Bill remind us that synthesizer music is folk music too. That might be an interesting way to look at it.)

The process of making music on a cello involves the whole body. Every note has to be shaped by hand: You can't set a parameter and have it affect a dozen or a hundred notes at one go. Intense concentration is required in order to avoid mistakes -- and the fact that you played it perfectly this time is no guarantee that you won't screw up royally next time. Yet paradoxically, in the midst of this intense concentration, the process of playing music can never become wholly conscious. The unconscious, the intuition, is arguably making a larger contribution at any given moment than the conscious mind.

The contrast between playing the cello and editing MIDI or audio in my sequencer couldn't be more striking. In the computer, if I fix a mistake once, it stays fixed. Which means that once I get a given line edited the way I want it, I can safely ignore it. (In fact, ignoring the tracks you recorded early in song development, which is an ever-present temptation, can lead to bad writing and worse mixes.) If I'm playing a keyboard, I'm more or less separated from the process of shaping individual notes, so I can ignore that aspect of musicmaking too.

In a MIDI-and-loop-based composing environment, every musical decision and operation has to be made consciously; there's no room for intuition. Yes, I can react intuitively when I sit back and listen to the tracks I'm working on -- but at that moment I'm not making music. I'm a listener. Who is making the music? Nobody.

In this sense, even when the computer is functioning flawlessly, it puts up a barrier between me and my music. The cello doesn't. The computer removes me from the process, not merely by imposing its inevitable biases on my aesthetic decision-making (though it does that too) but by eliminating my body as a source of musical expression. Music is always experienced in the body, but the computer doesn't know about that. When I use the computer for anything other than recording the cello, I'm strictly in my head. If you have a physical disability that prevents you from playing music without computer assistance, the computer's lack of somatic bandwidth is a wonderful thing. But if you're physically capable of playing an acoustic instrument and choose not to do so, the computer is cheating you of part of the musical experience.

Oh, well. As Joni Mitchell said, something's lost, but something's gained. I just want to be clear about the fact that something is being lost.

The Internet Music Glut

Where better to end these dyspeptic ruminations than with the Internet? I would have been perfectly happy, I think, to be the conductor of a brass band in a small town in the Midwest in the 1890s. When I wrote a new piece, I'd have some experienced musicians to play it, and a bandstand in the park for our concerts, and every Sunday afternoon my friends and neighbors could come to listen. They would have come, too. In those days the radio hadn't yet been invented, and the gramophone was a long way from being a hi-fi. If there had been a Musicians' Union, they wouldn't have had to use, "Always insist on live music," as their motto, because live music was music. There was no other kind. Today, I can have a web site. (Come to think of it, I do have a web site.)

The trouble with having a web site is, fifty thousand other musicians have web sites too. And they all hope to sell CDs to folks who visit their site. Imagine being on the bandstand in the park, only there are fifty thousand other bandstands in the same park competing for the attention of the passers-by. What cacophony!

In the so-called information age, we're laboring under an information overload. The scarce commodity is attention. Sure, millions of people use the Internet every day. But what's going to attract them to your site, or to mine? Free mp3 downloads? Everybody has those. In all likelihood, nothing will bring them to your site. If they're actually quixotic enough to go surfing for new music by unknown artists, they'll quickly discover that most of the music being offered for download is dreck. They'll give up in disgust long before they unearth your gems.

The Internet is less a promotional gold mine than an endless, foetid swamp of self-publishing. Having a web site may indeed be necessary as an adjunct to other, more traditional promotional activities, but it's not a substitute for them. All it is, for most of us, is a money-drain and a time-sink. In addition to putting up posters for your gig and paying a photographer to take shots for your press packet, now you have to have a web site. (Everybody who chose their band's name after checking to see whether the URL was still available, raise your hand.)

And we haven't even started talking about lost royalties yet. I'd rather not get into the whole Napster/RIAA thing. It's gotten pretty boring. We can fulminate all we like about piracy, but it's not going to go away.

Still.... In the end, digital piracy of music may turn out to be the single most significant reason why technology sucks for musicians. I'm not losing too much sleep over Metallica's lost royalties, or Michael Jackson's -- they have the machinery of big business on their side, and they're going to continue to reap royalties from radio sales, pick up fat checks for celebrity appearances, and so on. In the long run, though, when the value of a commodity (be it music or wheat) is debased by the fact that it's being given away for free, all producers of the commodity suffer, be they large or small. If the mp3 "revolution" cuts Michael Jackson's yearly net from (let's make up some numbers here) $20 million to $5 million, it could easily cut the yearly net of your favorite local new age recording artist, or of a struggling songwriter in Nashville, from $40,000 to $10,000 -- and that's the difference between making a halfway decent living and having to go out and get a day job.

There are ways around this deadfall. You can gig like crazy (hey, it worked for Benny Goodman). You can sell your production services to other musicians. You can write soundtrack music for dirty movies and get paid at a flat rate. In the plus column, the World Wide Web can help you make more money at any of the above. But in the long run, the Web is going to profoundly change the nature of mechanical licensing royalties. Pay for play has been a problem in L.A. clubs for a long time. Now we're being told we'll have to give away our recordings for free in order to promote our careers. This is not a good trend.

And you can't put the genie back in the bottle. Can't be done.

But then, Einstein couldn't put his genie back in the bottle either. If it's any consolation, the technology he inspired with one elegant equation sucks a lot worse than anything musicians have to contend with. In fact, a lot of things suck worse than anything musicians have to contend with -- even unemployed musicians saddled with inadequate or badly designed gear. Maybe we should all stop whining and get on with celebrating life. (I certainly should.) Having acknowledged that the world isn't perfect, maybe we need to take a deep breath and remind ourselves that no obstacle, be it large or small, can possibly stand in the way once we make up our minds to make the best music we're capable of. The Beatles made great music with a four-track reel-to-reel, and right now there's a kid in kindergarten somewhere who will make great music someday on a hand-held, voice-activated 27GHz computer that isn't even on the drawing board yet.

It's not about technology at all, not really. Technology is the railroad train, not the journey. If the train rattles and sways a little going around the curves, don't waste too much time worrying. We can still look out the windows at the mountains and the sky -- and someday soon we're bound to get where we're going. --JA

Footnote: The lead paragraph refers to hardware synthesizer modules, but as of mid-2005 those are all in storage. Software synthesizers now do it all. Someday soon I'll get around to updating this essay: What's that stuff about OMS doing in there in 2006?

Acknowledgments: Thanks to Marvin Sanders for suggesting the phrase "technology sucks." Needless to say, all of the specific instances of sucking cited herein are my opinions, not Marvin's. And thanks to Peter Gorges for commissioning this essay for a book he was planning to publish through his company, Wizoo. The book never happened, alas, but thanks to the Internet, even projects that never saw the light of day can find eternal life. (At least, it's eternal as long as I keep paying $29 a month for the server space.)