FCPX and the Domino Effect Part I: Stacking Them Up
COW Library : Apple Final Cut Pro X Debates : Walter Soyka : FCPX and the Domino Effect Part I: Stacking Them Up
Apple's release of FCPX has shaken the digital content creation industry, and raised a lot of speculation among editors about how much Apple understands or cares about professional users. On the one hand, FCPX sports completely modern media and imaging engines, a powerful and flexible database backend, and a completely new interface infused with new editorial concepts.
On the other, it's missing critical features that many working editors relied on: interchange with other applications, legacy compatibility, multi-camera editing, proper video monitoring, and tape I/O and control. No matter how appealing new features like native media editorial or pervasive metadata may be, the missing features immediately disqualify FCPX from numerous complex workflows.
With the end-of-life announcement for FCP7, and with FCPX being more of a new product than an upgrade, editors and facilities everywhere are now facing migration. That migration looks a little different for everyone. Some users have adopted FCPX and find that it's a great fit for their specific workflows. Some have taken advantage of compelling sales offers from Avid or Adobe and have already moved to another application. Others will stick with FCP7 for as long as possible before deciding on their next step. FCP7's EOL has been an invitation from Apple to editors everywhere to reconsider their needs, workflows, and application choice.
I'm going to describe what my migration is looking like. I started using Final Cut Pro with version 1, and I built my business on it starting with version 3, so I'll start with the context of Apple's rise in post-production. Then I'll get into what the release of FCPX has meant for me and how it's made me reconsider what tools I use.
I learned to edit on linear tape-to-tape systems and Avid systems. At the start of my career, I did editorial and finishing for local broadcast. I briefly used Avid professionally working for others, but I ultimately built my own business on Final Cut Pro during the DV revolution. I think this period of time, starting with the dawn of DV in the late 1990s and extending into the mid-to late-2000s, was a golden age for video on the Mac platform.
A Look Back at FCP and FriendsAt the beginning of that 10-year span, Apple was in serious trouble, prompting Michael Dell to suggest that Apple liquidate and return cash to their shareholders. Sales were down, licensing to clones was hurting Apple's bottom line, and Windows 95 had advanced the PC platform. Avid was threatening to abandon the Mac platform and go PC-only. Steve Jobs was brought back in as interim CEO in a huge corporate shake-up, and found a crowded and confusing product line.
Jobs simplified the line, making a clear separation between consumer products (the iMac and iBook) and professional products (the PowerMac and PowerBook). While bondi blue sold quite a few iMacs to consumers, Apple was intently focused on professional content creators. Apple was just as quiet about their future plans then as they are now, but from around 1998 to around 2008, nearly all of their surprises were good ones.
Apple started shipping Macs with Firewire. This replaced SCSI and laid the foundation for easy DV video I/O on every professional Macintosh.
Apple released Final Cut Pro. Originally developed by Macromedia as Key Grip, FCP may have been Apple's response to the threat of a deeper partnership between Avid and Microsoft. By keeping a good NLE on the Apple platform, Apple could continue to sell high-end, high-margin Macs to creatives. FCP v1 was widely scorned by established professionals, and was not immediately a viable replacement for Avid, but Apple encouraged third-party support to add capabilities quickly. FCP showed Apple's commitment to a market they were in danger of losing.
Apple released Mac OS X. Mac OS 9 was built on a very dated architecture which lacked modern reliability and performance features like memory protection and pre-emptive multitasking. Mac OS X promised huge improvements, but it was also a huge risk; it was a new experience for users, it required more system resources, early versions were so slow and unreliable as to be almost unusable, and there was a major lack of development for the new platform. Recognizing this, Apple offered a gradual transition from OS 9 to OS X, supporting dual-boot machines, porting the Carbon API from OS 9 to OS X, and building OS X's Classic emulation layer. By the time Mac OS X 10.2 (Jaguar) was released, the system was usable and applications were available.
Apple bought Nothing Real, the developers of Shake, which was then the leading high-end compositing package. Owning Shake gave Apple some Hollywood cachet and ensured that blockbuster work could be done on a Macintosh. Apple released the Xserve. After decades of focus on the desktop, Xserve marked a serious push into the enterprise and facility market. With Xserve, you could reasonably have an Apple-made backend for your Apple-made front office. High-performance computing markets took note, and the Big Mac Xserve cluster became one of the fastest supercomputers in the world.
Apple expanded Final Cut Pro, adding capabilities and new tools in every release. New tools like Cinema Tools, Compressor, DVD Studio Pro, Motion, and Soundtrack Pro offered new capabilities, and were ultimately merged into Final Cut Studio.
Final Cut Pro X
Apple won over Walter Murch, who edited Cold Mountain on Final Cut Pro with four Power Macintosh G4s, a Rorke Data SAN, and Aurora Igniter cards. A serious editor cutting a serious film on FCP, Murch proved that FCP was ready for the big time.
Apple developed FCP's XML format. This was like a rich EDL, allowing more advanced interchange between different versions of FCP, or between FCP and third-party applications.
Apple released XSAN. Coupled with a couple Xserves and a RAID chassis, you could build an Apple-powered SAN to feed media over fiber to NLEs across your facility.
Apple moved from a PowerPC architecture to Intel. PowerPC was not keeping pace with Intel, and the switch meant a great performance increase for Mac users. Apple brought out the transition playbook again, adding the Rosetta PPC emulation layer and adding Universal Binaries, which allowed developers to include binaries for both PPC and Intel architectures within the same application bundle. Mac Pro workstations iterated quickly, getting more processing cores and increasing bandwidth. Apple bought Silicon Color, developers of a $25,000 color grading application called Final Touch. Apple rebranded the product as Color and added it to Final Cut Suite at no additional charge, putting a powerful grading tool within reach of a new market segment.
Apple developed and released ProRes, a mastering-quality, lightweight codec that allowed editors to edit HD from a single hard drive, or even on a laptop, instead of being tied to a huge and expensive RAID array.
Apple bought Proximity's artbox, a media cataloging and workflow administration tool. Apple rebranded it as Final Cut Server and marketed it to facilities looking for tighter integration in their predominantly Apple workflows.
It was an exciting time to be an Apple-based facility. Apple's tools had grown in power, reliability, and ease of use, and after a decade of continuous investment and improvement, FCP was embraced across the industry.
Other developers were flocking to the Mac/FCP platform. A large number of developers had written FCP plugins. Other high-end tools like Autodesk Smoke and DaVinci Resolve, previously only available on Linux workstations, were ported to the Mac platform.
FCP's influence should not be underestimated. Apple started with a single new application, bought from another developer a decade prior, and built it into a platform, giving them a huge presence in the content creation industry, and bolstering the perception that Macs were the tools of choice for creative pros.
Content creators had begun to rely heavily on Final Cut Pro and Apple -- but did Apple also rely on content creators?
Maybe not. After all these advances in post-production, things started changing at Apple.
Apple Professional Video Timeline. Please click image above for larger view.
Apple's Shifting Focus?The iPod and the iTunes Music Store had made Apple relevant to new market segments and new generations of consumers. The iPhone, iPod Touch, and eventually the iPad sold so well and changed what computing means so much that Apple Computer, Inc. changed their name to Apple, Inc.
Apple continued product development under their traditional veil of secrecy and delighted consumers with unexpected announcements and launches, but from the perspective of content creators, the surprises stopped being good.
Apple ended Shake. After a few important utilitarian and maintenance releases, Apple announced Shake's end-of-life and sold existing inventory at a steep discount without ever having added a significantly new feature, clearing the way for Nuke to move in and ultimately dominate the high-end compositing market. For a while, rumors about an all-new, high-end compositing application under development at Apple, code-named Phenomenon, swirled around Internet forums. By this point, most of the Nothing Real engineering team had already left Apple, and it's unlikely that Phenomenon was ever real, but Apple allowed the rumors, and with them users' hopes, to continue.
After promising a 64-bit version of the Carbon API to developers, and after shipping 64-bit Carbon betas, Apple suddenly reversed course. They dropped 64-bit Carbon in favor of the 64-bit Cocoa API. Developers like Adobe, who had been relying on 64-bit Carbon for apps like Photoshop CS4, were sent back to the drawing board. Photoshop CS4 was 64-bit on Windows but 32-bit on Mac OS X because of Apple's abrupt roadmap change.
Apple neglected the Xserve RAID, sticking with parallel ATA architecture despite the easy performance gains to be had from serial ATA. Subsequently, they discontinued the Xserve RAID altogether, instead selling Promise RAID hardware to complement the Xserve.
Apple ended the Xserve, making enterprises and facilities built on all-Apple infrastructures question their investments.
Meanwhile, gamma issues plagued the platform. An editor may see the image appear one way in FCP, another in Color, a third way in the QuickTime Player, and then a fourth way after compression for delivery. The issue wasn't limited to display; poor color and gamma handling in QuickTime meant that two different applications may interpret the color in the same file differently, making round-tripping across applications a challenge.
Graphics card performance continued to suffer on the Mac platform. While Apple has long had a reputation for excellence with graphics, it has nothing to do with the hardware anymore. PC graphics performance nearly always outstrips Mac graphics performance with the same card, and the selection of graphics cards available on the Mac platform is small.
Applications like Adobe Premiere Pro, Autodesk Smoke, and Blackmagic Design DaVinci Resolve all preferred or required high-performance NVIDIA graphics, but there are only a handful of Mac-compatible NVIDIA cards. The cards that are available suffer from poor driver quality.
Development on FCP failed to keep pace with the changing industry. You could buy multi-processor computers from Apple with well beyond 4 GB of RAM, but FCP remained single-threaded and 32-bit. Tapeless acquisition swept the industry, but FCP's media engine couldn't accommodate native editorial for many formats, forcing editors to transcode to ProRes first.
All the while, Apple promised that the next release of Final Cut Pro would be "awesome." Adobe had already introduced a 64-bit rewrite of Premiere Pro, featuring GPU acceleration and native format editing. The rumor mill began speculating that these features and more would arrive in FCP8, allowing Apple to once again leapfrog the competition, and setting expectations for the new release very high.
When FCPX was previewed at the SuperMeet at NAB, and when it was released a few months later, it was equally surprising what new features were included, and what old features were conspicuously absent.
In Part 2, I'll outline how I see these trends developing and what I think it means for my own business.