Thursday, April 28, 2005

I got the music in me

Some things the general public doesn't really know about are some of the behind the scenes changes that have happened in the recording business. Here's a partial update.

Originally, albums (records) were made by recording everything at once and capturing it on a disk made of lacquer coated on metal. This provided a master which could then be processed and turned into multiple copies which in turn were used to make records, you know, the vinyl type.

Recording on magnetic tape came later. Although it had been invented and subsequently perfected by the late '40s, it was really in the '50s when recording studios started to record to tape. At first it was a similar process: record directly to tape, transfer to lacquer with some additional processing called "mastering", then manufacturing. And of course, all of this was done in glorious old mono.

Stereophonic recording soon followed, but it was not available to the casual consumer. The first stereo recordings sold to the general public were on reel to reel tape, sold in the manner of record clubs, for enthusiasts who had early home stereo tape players. Mono was still the main choice for users of records.

But that soon changed. With the invention of the stereo disk, it quickly became the medium of choice for the manufacturers and consumers alike. Recording was done to a 2 track tape machine, and then edited and mastered to the master disk. But that was soon to change.

Several manufacturers of tape recorders existed: Ampex, 3M (Minnesota Mining and Manufacturing, the Scotch Tape folks) in the US, Studer in Switzerland, as well as some other manufacturers with smaller client lists. And they were competitive, as well as being responsive to the recording industry. And very quickly the benefit of more tracks became clear.

First came 3 track. Used extensively in the US by Capitol Records, it allowed real time recording with final decisions put off until later. For example, while I was employed by Capitol Studios in the late '90s, I had an opportunity to play many of the wonderful old tapes in their library. It was virtually a religious experience to play a Nat "King" Cole 3 track master for the first time. Orchestra on 1 & 3, in stereo, vocal on 2 in the center, the experience was enlightening. I realized that the recordings being made today, while technically more complex, sounded no better.

Of course, this material was most often released in mono. The fledgling multitrack process was mostly used to postpone mix and edit decisions until the last moment, the mastering room.

But artists had other ideas. It quickly became obvious that one could record on track 1, then "overdub" on tracks 2 and/or 3, and create musical ideas that were previously impossible. A singer could sing along with him.herself, for example. In the most famous example of overachievement, Les Paul, along with late wife Mary Ford, could create a whole orchestra and vocal ensemble with just 2 people.

Ampex responded by developing first 4 track tape machines, then in the later '60s 8 track mechines. Quickly after this came 16 track, and then, by the very late '60s and early '70s, we had 24 track machines, using 2" wide magnetic recording tape. By the mid '70s, 24 tracks on 2" tape had become the standard. A few alternative formats were tried, but were discarded; 24 track reigned supreme.

During this time we had 3 primary manufacturers of recording tape: Ampex, 3M (Scotch), and BASF in Germany, the folks who perfected the technology of audio recording on magnetic tape. All 3 had good products, and some engineers preferred one over the other, but they all worked well.

Meanwhile, digital recording, by converting analog electronic signals into computer storage information consisting of groups of 1s and 0s arranged into "words" having a numerical value, had been on the ascent.The first commercially usable digital recorders were demonstrated in the mid '70s, and by the early '80s had become, not commonplace, but at least regularly seen in studios. And as Sony & Phillips were perfecting the CD, Sony also developed the digital recording system that all studios had to use in order to manufacture CDs. These early digital units were 2 track systems, and were used for the final mixdown of all the tracks on the multitrack tape, the conventional way of making records at the time.

Soon, digital multitrack recorders came on the scene, and they became truly commonplace by the mid '90s. While they offered an alternative to some of the inherent noise of analog tape, they also presented their own problems. Most people agreed that the early digital recorders simply didn't sound that good. But they offered some electronic editing features, more tracks per single reel of tape, and a lower residual noise floor.

Meanwhile, percolating under the surface were some really new devices: personal computers using 3rd party software and hardware to allow recording digitally into the computer, and storing the digital audio data on a hard disc drive. Starting in the mid '80s, these systems were adopted first by home users and smaller studios, but didn't get much market penetration at first. The editing features were truly spectacular, but these early systems were slow in performance, and hard drive costs were high, so the application seemed limited.

But by the mid '90s, computer based recording systems (DAWs-Digital Audio Workstations) were coming of age. A few systems were available for PCs, but 2 systems designed for the Mac really captured the market: Sonic Solutions, designed for mastering (final processing after all mixing is finished), and ProTools, designed as a complete multitrack recorder and editing system.

2000 was the transition year. ProTools moved from being a clever alternative to the dominant recording system in both the music and TV/film post production worlds. I have seen the use of recording tape, both analog and digital, decline steadily since then. Of course, some producers and engineers still like to use analog tape for the "sound", but the actual amount of tape used has really dwindled. Ampex had spun off their tape manufacturing division several years ago, and newly renamed Quantegy, they bravely soldiered on. Until recently.

In late '04, Quantegy announced that a) they were looking for a buyer, b) they were filing Chapter 11, and c) they were closing. That meant there were NO manufacturers of large format open reel tape for recording studios. Widespread semi-panic ran through the industry, and people called all the distributors trying to buy the last available stock. I personally arragned for a high profile client to purchase virtually all the inventory from a distributor in Chicago.

Meanwhile, another radical change was taking place, this time in content delivery (previously called sales): downloading. Shortly after the start of the internet, people figured out how to copy albums, both vinyl and CD, into their computers. And as internet complexity exploded, the idea of using the web to share music with someone else became a new goal.

As far back as '97, while I worked at Capitol Studios in Hollywood, the record companies were in a dither about downloading. There existed, for a brief time, a joint effort by the 5 (at the time) major labels called The Madison Project, so called because the meetings to initiate the project were held at Sony Record's Madison Ave. offices. The 5 were EMI/Capitol, Sony, BMG, Universal, & Warner Bros. They discussed many ideas to counter "free downloading," including copy protection, watermarking of files, value-added fee based downloads, etc. But reality intervened. Mergers occured, alliances were broken, and all moved on.

Apple has been quite successful with iTunes, a service I feel has established the benchmark for the downloading paradigm, at least in the first round of the struggle. And after much legal wrangling, Napster has emerged somewhat respectable, although much more tame. But the battles aren't over.

Record companies still haven't figured out how to combat illegal downloading. Will they succeed? I doubt it. The 18th amendment didn't stop people from drinking. And while they try to wrap their corporate minds around the idea, the world flows on around them. We have a generation of kids who feel that they are entitled to take anything they want, thus ripping CDs and uploading them for file sharing. Of course, if you tried to take their PlayStation from them using the same logic, they'd freak.

We also have a generation of label execs who want their 7 figure salaries to remain status quo. And we have a group of self congratulating lawyers acting as talent scouts who continue to sign talentless acts to labels. While I worked at Capitol Studios, the head of the label, who was quasi-famous for having grudgingly signed Nirvana while he was at Geffen, spent years pouring money down the drain, and into his own pockets, and at the end, had only Everclear to show for his effort. Well, and a big house in Brentwood, and a golden parachute.

So budgets are being cut for albums. The labels think that since everyone is recording into ProTools, they don't need studios with expensive rent and infrastructures anymore. The fault in that logic is clear. ProTools is merely a medium, one that can be used anywhere, in studios formal and casual, large and small. But because of the perception that it has to be cheaper than "back in the day" they are willing to spend less on recording. And while disk drive space costs less than 2" analog recording tape, the proportion of the big picture is not really significant. Think "Tom Cruise gets 20 mil$ per film, is it important whether he flies first class or business class?"

Producers are told that they did such a wonderful job on artist X's first album, for which they were paid a fee of $100,000, they are being asked to do X's 2nd album. Only this time they are being asked to do it "all in." This means a fee will be paid, from which the producer must deliver the finished product, with all costs inclusive. Thus, all expenses come out of the producer's bottom line. And it trickles down from there.

Management companies call to book studio time, and here's the conversation. "Hey, we had a great time working on the X's project last year. The album shipped platinum, won a Grammy, we all couldn't be more pleased. So we'd like to the the next album with you again. Only, instead of charging is $1500 per day, with ProTools rental extra, and charging us for a 2nd runner to help with the complex food runs (vegan/kosher for the drummer, Zone diet for the singer), we now want to pay $1000 per day with all the other stuff included."

So tape sales have slowed down, so much that Quantegy filed Chaper 11. And studio after historical studio here in LA is closing down. A world class producer friend of mine is buying a house in the Hollywood Hills where his artists can stay while recording on the lower floors. Sad, maybe. Yet the demand for "content" goes on. And labels sign acts, some of which actually make money. And execs still make 7 figure salaries, no matter how badly the labels underperform. And kids want the music for free.

Were does it end? Will everyone with a Mac record something with GarageBand and be famous for Andy's 15 minutes? I dunno. I just spent the last couple of days working for a World Class producer, a nice man, who needed my skills to make sure that the computer music and the tracks on tape synchronized correctly, not a trivial task. In the old days he would have been working at a studio with a full time technical staff. But nowadays he's working at a studio he can afford. Except without that staff, he couldn't be confident that all went well, thus, he hired me. Out of his own pocket. Because he cared.

But who cares at the end of the day? Not the label exec, who is too busy checking his Blackberry at the club. Not the attorney because, well, that's obvious. Not the kid who feels entitled to the music, while compaining that artist X didn't tour last year because, well, he couldn't afford to. And not the studio owner, who just sold the property to a developer who wants to rebuild it into a check cashing business and a laundromat.

So that's some of what's been happening in the recording biz lately. As a friend of mine once said, "It's all tinsel and glitter."

Update: corrected for late night typos.

4 comments:

Jane Hamsher said...

Awesome, Steve. Thanks. That was an amazing effort.

Anonymous said...

A little bit of a contrary view of the early history of digital. I was at CBS Masterworks in the late 70's/early 80's and was one of the first producers to use digital audio.

Everyone agreed that the early digital systems we used - soundstream, sony pcm 1600, mitsubishi, and 3m - sounded vastly superior to analog, even with Dolby. Problems with the sound were noted later, but at the time, to listen to a recording with no wow and flutter and no tape hiss was a mindbending experience. Some people noted a lack of warmth in digital, but we assumed that our ears had become so accustomed to the flaws of analog that their absence was an annoyance. I'm still of the opinion that most of the complaints about digital by my generation of producers, engineers, artists, and listeners stem from this.

No, the problem wasn't the sound in the early days, but editing, which was unbelievably cumbersome. Sony, and later, 3m, utilitzed techniques derived from videotape editing. Typically, you aligned tape machines and assembled a master by re-recording bits from the session master tapes. The early editors were expensive, slow, hard to come by, unreliable, and inflexible.

By this time, razor blade editing of master tapes was a fine art. None of us in classical thought twice about creating a two-inch master of a single movement from hundreds of razor blade splices. Digital editing, by contrast, took far too long, as you waited for these bulky transports to align themselves, copy, and then review.

We were truly going crazy from impatience back then. Mitsubishi released a two track digital recorder that could be edited but it required special tools and uncomfortable gloves. Worse, the edits would pull apart or change sound. Soundstream had the best early solution: you gave them an editing plan and they edited your master in Utah (by transferring the tapes into a mini-computer and using non-linear techniques) and sent you a completed tape. The major problem with that was that last minute changes were a nightmare.

The cumbersome nature of digital editing, combined with the unbelievable expense (in the early 80's a 32 track 3m machine cost $150,000 and tape for it reached similar stratospheric heights) led to a backlash. It was then that complaints about the sound started to be heard.

Some of these were quite legitimate. Soft, sustained, textures were often marred by choppy artifacts. Mastering eq, long standard from the lp days, was too bright. And there were some more subtle flaws in digital to analog conversions that led to problems with stereo image and barely quantifiable crtiteria such as the recording's ability to breathe, and so on.

These days, non-linear digital audio editing is child's play, far better than a rzor blade. And the converters are vastly improved. Pop producers may prefer, for a variety of reasons, to record to analog. The wow and flutter which we worked so hard to avoid now is heard to impart a certain desirable, all-but-unconscious positive effect. And the high frequency rolloff and frequency bumps that are inevitable in analog also impart desirable "distortions" in the original sound.

But a modern digital recording played back from a hard drive is as close as we can get right now to reproducing the sound that engineers hear live through their consoles. And cd's are still the best sounding mass market audio delivery system, if no longer the most convenient.

SteveAudio said...

Thanks for the perspective. I agree, in the early days, music without noise & flutter was pretty cool. Demos were done at AES shows, etc., and all of us were pretty impressed.

And I agree, editing was abysmal. The Soundstream method was clearly not realistic, Mitsu allowed some razor blade work, but it wasn't easy, as you noted. The Sone DAE was the first truly functional editor, as you point out, derived from video editing techniques.

I was working for Stevie Wonder in '86, and he was still recording everything he did to the 3M 32 track machines he owned, and he has pretty good ears. But I don't think the backlash agains digital had anything to do with cost. If anything, I might be inclined to align it with the ascendence of the Sony machines, first the 24 track, then the 48 track.

Also, as you surely remember, Apogee, who make some nice A-D convertors, got their start making filters for the Mitsu 32 track machines. By the mid '90s, Apogee convertors were pretty standard in LA, wrapped around DAT machines. And while many sessions I worked on were recorded to Sony 3348s, I can't remember anyone saying they liked the sound of them. In fact, many engineers complained about the sound, and the not too good noise floor. No, by this time, those machine were values as a practical tool: lots of tracks, some editing features.

When I started at Capitol, in '95, all the mastering and editing room had convertors in front of the Sony 1630s, often the very expensive DB Technologies AD122 (now Lavry Engineering.) And the sonic differences were clear, it didn't take a pro to hear the difference.

So while I don't completely agree with you, I don't completely disagree with you. I have been part of many casual listening tests, as well as some double-blind ones, and convertors make a huge difference in digital reproduction. As the process has gotten better, technology improved (24 bit, 96KHz and higher), the overall sound has improved. But listening to a console's bus versus replay, the differences are still fairly clear.

Does it matter? No, not after all other steps in the chain. By the time it becomes an MPEG-3, in some kid's iPod, all these subtle differences will be washed away. But in the control room, it still matters. At the time I left Capitol, early 2000, all the major film scores we did were still beoing recorded to 24 track, 15 IPS, Dolby SR, 0 level (185 nW/m). Sometimes 48 track digitla, but most often analog. For a reason.

Tape is still highly valued for its "Rose Colored Glasses" effect; it makes everything sound 'pretty.' Is digital more accurate, perhaps. Ultimately, it's about the music.

Anonymous said...

I think you're probably right that most of the complaints about digital in the early days were reasonable complaints. As our ears got more adjusted, the problems became glaringly obvious. And thank goodness, companies like Apogee addressed that market and other companies improved their product.

I think, however, that there were in the very, very beginning some complaints about the sound of digital that were hyped, related more to the expense of retooling, as well as the hassles connected with editing the stuff than real differences. After all, Studer machines cost a pretty penny and competition at the high end has always been fierce.

In classical, however, we had no choice but to go digital as soon as we could edit (actually, I believe there were certain "digital" lp's by other companies that were unedited, released before they could edit at all). The buzz on this new type of recording was very strong and in the early days you could sell 5000 more units (a nice number for classical) merely by putting the word "digital" on the sleeve.

Not to say that everyone liked it. Classical producers were split into two rival camps, the multi-micers and the two-micers. Prior to the release of the 3m 32 track, the multi-micers were furious that suddenly they no longer had access to multitrack - an "all-digital" lp outsold an analog-digital lp handily and in fact, one producer was fired for refusing to record onto two track digital.

However, once the editing technology developed and the converters improved, the recording of classical music thoroughly embraced digital and never looked back. My friends who are still producing (I quit producing and now compose fulltime and my ears are therefore focused on different things than digital vs. analog sound quality) wouldn't dream of recording to anything but a digital recorder.

You're right, though, that in pop recording, analog remains a crucial component in the bag of tricks. I'm not surprised that film scores were recorded the "safe and true way" for as long as they were. Film companies were/are exceedingly conservative, especially in regards to audio. In the 1980's in New York, even major film mixing houses had ADR setups from the 1950's. Maybe, they're more venturesome now, but probably not by much.