Most desktop musicians are accustomed to hype. We're inundated with a generous Minimum Daily Requirement of it, and we frequently dish it up ourselves. But even in today's jaded marketplace, when Apple paraded out OS X to its eager fans in March 2001, the excitement was palpable. OS X was hailed as possibly the most significant development in Apple history since the Macintosh itself. The new operating system boasted the bedrock stability of a Unix code base with core-level multimedia integration, while still maintaining the smiling mug and useful simplicity of a Mac. As described by Apple, OS X would change the face of Mac computing practically overnight.
So here we are, more than a year and a half later, and only a handful of audio applications have placed both feet firmly on the good ship OS X. Hardware drivers and plug-ins also remain a crapshoot. While the rest of the Mac world has been moving to OS X with only minor hiccups, most audio-related companies have lagged far behind. What's the holdup? The answer, as you might expect, depends on whom you ask.
HOW DID WE GET HERE?
The Mac, for all its renown as a media-production platform, was not originally created with that intent, any more than Wintel machines were. The Mac's Sound Manager audio protocol was woefully limited, and its provision for handling MIDI and plug-in DSP was pretty much nonexistent. But the Mac's architecture was very accommodating to third-party add-ons for audio, MIDI, and other media-creation functions, and over the years, a number of innovative technologies have been developed to work that magic within the Mac operating system.
Unfortunately, the various entities creating those technologies have, for the most part, done so independently (and competitively), with precious little leadership or direction from Apple. The result is a laundry list of occasionally compatible formats and protocols for audio, MIDI, and plug-ins: OMS, FreeMIDI, DAE, Direct I/O, ASIO, EASI, MAS, TDM, RTAS, and more. And attached to each of those “standards” is a structure of products and a base of users built up over many years (and representing a lot of cash).
Enter OS X. Much more than just a handful of tweaks or a new coat of paint, OS X is a major paradigm shift for the Mac. It promises to integrate multichannel audio and MIDI at core level, reducing latencies to single digits. (Note that I said “promises.” More on that in a moment.) In fact, it's fair to say that OS X has the potential to outperform all other operating systems, without any third-party extensions (see the sidebar “Speed Demon”). By integrating all aspects of multimedia at the core level, Apple's goal is to drive this ubiquitous functionality from the application layer down to the system layer. That has the potential to create what should have existed all along — a true single standard for all Mac media functions to be built upon.
A move in that direction, however, places the future of many of the existing formats and protocols in question by potentially eliminating the need for third-party extensions. Developers are faced with the difficult choice of either abandoning their hard-earned protocols in favor of direct integration or somehow merging their technologies to run within, alongside, or on top of OS X. It's not a simple task either way. Even veteran Mac programmers have said that it has taken them several weeks just to get their heads around OS X. That's a lot of time to devote to engineering resources before the task of adapting products even begins.
Hardware-based systems such as Digidesign's Pro Tools have an even more difficult task: many developers have found that writing hardware drivers for OS X is considerably more complex than for OS 9. And plug-ins raise a new set of issues. As manufacturers create native OS X versions, users will (justifiably) gripe about having to repurchase what they already own.
IS IT SOUP YET?
The biggest hurdle in getting audio developers on board has arguably been the development pace of OS X itself. While OS X promises a lot, much of that has until recently been just that — promises. The original release had no real support for multichannel audio; that was gradually implemented over several updates. CoreMIDI support was similarly crippled; although the infrastructure was there, there were no access hooks for programmers. Not until the release of version 10.2 (Jaguar) has much of this support been fully implemented. According to several developers, the main reason for the delay in building OS X support was that the operating system wasn't fully baked yet.
At the time of this writing, Jaguar has just begun shipping, so it's too soon to tell if the new version will resolve all the open issues. The superiority of CoreMIDI over the present options is generally agreed upon. But although Apple feels that Core Audio should universally replace applications' audio engines, not all development partners agree. Hopefully, many third-party protocols will continue to coexist peacefully.
The issue of plug-in standards is similarly unresolved. Though some plug-ins have been Carbonized to function in OS X, for the time being there are few native Core Audio plug-ins other than a dozen or so bundled in Apple's OS. Audio Unit (OS X's native plug-in format) is reported to be fully implemented in version 10.2, but again, until now no shipping OS has supported it.
A number of smaller issues also remain. For example, many users have reported having problems with transferring large (1 GB or more) files to and from external FireWire drives; others have encountered audio problems within iMovie. While neither issue is directly related to writing audio software, many developers have pointed out that Apple's slowness in resolving loose ends doesn't help to inspire their confidence.
In Apple's defense, there has never been, and probably never will be, any operating system with a flawless version 1.0 release. The sheer magnitude of this project makes some stumbling inevitable. Nevertheless, a number of audio and media developers and their users have been understandably vocal in criticizing Apple's pace on this issue. At least in the audio world, the full potential of OS X still has yet to be realized or proven. For all its zeal in wanting Mac users to adopt OS X, Apple has been a major obstacle in getting it done.
Anyone who has followed Apple's history knows that the company is famous for surprises and sudden changes of direction. As if the whole OS X transition weren't interesting enough already, on July 1, 2002, Apple redrew the maps with its surprise purchase of Emagic. The venerable creator of Logic Audio has been an industry leader since the Neanderthal days of MIDI, and Logic has always been among the top choices for audio professionals.
Apple's purchase of Emagic has potentially far-reaching consequences for Mac- and PC-based musicians. The most visible and immediate results are on the Windows side; Apple has dropped Logic PC, leaving many Logic users stranded and removing a major innovator, competitor, and motivator from the ranks of Windows audio.
In the Mac world, the future is far from certain whether you use Logic or not. What will be the long-range effect of Apple's move from host-hardware manufacturer to direct competitor? How will Apple's relationships to software rivals such as Steinberg and BIAS fare over time now that the developers of Logic have the inside track on Apple's plans? What about hardware developers? How will Apple's handling (and probably integrating) of Emagic hardware affect companies such as MOTU and Digidesign?
Even for Logic users, nothing is guaranteed. In the short term, Logic's audio engine will be a great complement to everything from iTunes to Final Cut Pro. But how will being a part of Apple affect future development? Certainly there's enormous potential for Apple to take a leading role in the development of operating-systems media technologies, and that would be an industry first. We'll just have to wait and see whether Apple gives the new “iLogic” division the resources and autonomy to be a driving force or whether it waters it down to a consumer-grade shadow of its former self. (For more on Apple's purchase of Emagic, see “First Take: Big Mac Attack!” in the September 2002 issue.)
Hopefully, we're nearing the end of a protracted transition that has been tough on everyone. Apple has developed an operating system capable of transforming the Mac into a far more powerful multimedia machine — something even Apple admits should have happened 15 years ago. Assuming Jaguar has all the pieces finally in place, it's likely we'll finally see the migration begin to accelerate. (Of course, Apple's stated intention to make all pre-X operating systems obsolete in the not-too-distant future tends to wield some leverage.)
Developers, for their part, have been in the unenviable position of having to essentially start over. Much of what they've created for the Mac of the past — enormous amounts of work and expensive research — simply won't work under OS X. For all intents and purposes, they are now writing for a brand-new machine that handles tasks in a completely different manner. Entire code bases have had to be scrapped and rewritten.
Meanwhile, most users are still watching the dust settle. Mac users of graphics and other applications have largely made the jump already, but for desktop musicians and Mac-based studio owners, the passage is as agonizing as watching slugs race. We run OS X on partitioned drives, going back to OS 9.x for the stuff we still can't do in OS X. Nonetheless, things are slowly changing. As of this writing, Propellerhead's Reason, Ableton's Live, BIAS Deck, and a handful of other applications run under OS X; by the time you read this there will be many others.
STATE OF THE STATE
I got in touch with several of the major players in the world of Mac audio in order to get an update on where they are with OS X.
BIAS was one of the first companies to support OS X, with the 3.0 release of Peak, its 2-track mastering program. Once multichannel audio support was fully functional under OS X, BIAS was again among the first to update its software (see Fig. 1). Version 3.5 of BIAS Deck supports Core Audio as well as ASIO and Sound Manager. It also supports VST plug-ins, including Carbonized versions. Vbox, BIAS's multi-effects control environment, is also OS X compatible.
MOTU has announced an OS X version of Digital Performer, slated to ship by the end of 2002 (see Fig. 2). While it's a fair bet that this new version of Digital Performer will do away with FreeMIDI, it's likely that MOTU's MAS protocol will continue for audio and plug-ins. MOTU has also announced OS X support for its 828 FireWire interface, as well as its Clockworks control-panel software for MIDI interfaces.
Digidesign has had what is arguably one of the biggest tasks because it provides a truly integrated hardware-software system. At the October 2002 AES show in Los Angeles, the company demonstrated its newly redesigned OS X-compatible Pro Tools 6.0. The new version is slated to ship by January 2003. Pro Tools 6.0 will still be based on Digidesign's DAE, but of course CoreMIDI will replace the very dated OMS. RTAS and TDM plug-in protocols will also remain a part of the Pro Tools feature set. There's been no word yet on whether there will be an OS X version of Pro Tools Free.
The Mac version of Steinberg's Cubase SX (see Fig. 3) now supports OS X; Nuendo is scheduled to follow shortly. Steinberg will continue its ASIO protocol for audio, as well as its VST format for instruments and plug-ins. The company is also writing OS X drivers for its hardware products.
Developers of more specialized apps also reported various stages of development. Propellerhead released Reason 2.0 with OS X support early in 2002; no word yet on ReCycle or ReBirth. Ableton, also an early adopter, already has an OS X version of Live.
BitHeadz was another company quick to support OS X, with the release of Unity Session 3.0 in January 2002. Unity AS (formerly Retro) has also been updated, as have some Unity modules. Phrazer 2.0 will also support OS X. Native Instruments has not yet made an official announcement of its OS X support, though it's likely that updates of Traktor and Kontakt will be available by the time you read this. U&I Software is in the process of porting its full product line and will release OS X-native versions of MetaSynth, VTrack, and ArtMatic Pro.
Rocket Network has announced full OS X compatibility for its products. Although Rocket's technology is not directly dependent on OS X's audio implementation, there's considerable interdependence on individual audio manufacturers as they migrate their applications. Interestingly, a good bit of attention is being given to Rocket's potential for transferring audio data between OS 9 and OS X, much as it has been used between Mac and Windows applications.
Waves is working on OS X versions of its entire product line. In addition to providing Core Audio support, the company will continue to support all third-party protocols, including TDM, RTAS, MAS, and VST. Cycling '74 is working on an OS X version of Max 4.0/MSP 2.0. Kind of Loud and DUY have no information available yet, but Arboretum has announced upcoming OS X support for its Hyperprism and Ray Gun plug-ins and is introducing Montage, a new OS X multimedia development application. Spectrasonics expects to have an OS X version of Stylus, its vinyl-groove-module plug-in, by January 2003.
Sibelius Software has already released an OS X compatible version of its Sibelius 2 notation program (see Fig. 4). MakeMusic's Finale 2004, scheduled for summer 2003 release, will be a native OS X application. PG Music has no immediate plans for an OS X version of its popular Band-in-a-Box software; its upcoming release will run in Classic mode.
TC Works has already released OS X versions of its Spark product line. Updated TDM and VST/MAS versions of its plug-in line should soon be available, as should drivers for its PowerCore DSP unit.
As for hardware, RME has released OS X drivers for its Hammerfall card; M-Audio has done the same for most of its product line. As noted earlier, MOTU has released OS X drivers, and it should surprise no one that Emagic was among the first to release OS X drivers, for its EMI 2|6. Event reports that its EZbus is fully OS X native. Apogee has drivers for its Mini-Me USB preamp, as does Tascam for its US-428 and US-224 USB interfaces.
Daniel Kelleris an engineer and producer living in Southern California. His work in digital multitracking and computer audio has included second-story gravity tests on a number of systems.
The Peabody Institute, in a report titled Audio Latency Measurements of Desktop Operating Systems (by Karl MacMillan, Michael Droettboom, and Ichiro Fujinaga; August 2001), cited OS X's Core Audio as being one of only four systems registering throughput latencies of less than 2.9 ms. (The other three systems hitting that milestone were all Pentium III/933 MHz machines, one running Windows 2000 with ASIO and two running versions of Linux with a third-party software patch.) The OS X system was the only one capable of achieving those extremely low latencies with no add-on hardware or third-party extensions.