On Wednesday, Dec 10th I attended my third HPA luncheon focused on explaining tapeless workflow for post-production sales people. The “promise” of tapeless workflow is that companies should be more reliable and more productive through new efficiencies. The general consensus is that this new digital world is actually quite disintegrated - instead of systems working seamlessly, we are seeing where the seams are showing.
The post production industry now has to deal with the fundamentals of the computer business, where technology changes every few months and clients demand a lot more for a lot less. So, how do you build your business on this kind of economic “quicksand”? How do you keep up with the dramatic changes in the data-centric workflow world? Sales organizations need to understand the missing pieces in order to keep the clients happy and to invent new opportunities for revenue.
The panelists helped frame out today’s problems and promises by explaining the evolution from the simpler linear Film-centric workflow to the more dynamic but rapidly changing Data-centric workflow. They also managed to talk richly about a world of 1s and 0s without breaking into jargon-speak, a truly impressive feat. The panelist included:
Jim DeFilippis, P.E., Sr. VP Digital Television Technologies and Standards, FOX Technology Group
Ramy Katrib, CEO, DigitalFilm Tree
Mike Wong, Senior Manager, Apple
Russ Walker, Director, Business Development Production and Post Solutions, Panasonic Broadcast & Television Systems
Craig Yanagi, National Marketing Manager, JVC Professional
Leon Silverman, President of Laser Pacfic, moderated the talk with a proclamation that should be announced at every panel for every conference, that vendors onstage are not allowed to “push” their product. The panelists definitely kept the “it’s all about me” factor down, resulting in another successful panel discussion about tapeless workflow.
DEFINITIONS AND BACKGROUND
Tapeless Workflow is the new “Paperless Office”
There was general consensus that “Tapeless” as a term sucks. We don’t drive on horseless roads anymore than we work with tapeless computers. The term everyone liked better was a “File-based” workflow, but I think “Data-centric” workflow is more accurate due to the incredible importance of metadata (data about data). The other joke that came up was how MUCH a tapeless workflow still relies on tape, especially for long term archival storage.
The Evolution of Tapeless - slow and simple to rapid and complex
Film - For most of the 20th century, filmmakers all around the world shared the universal language of cellulose and worked in discreet linear steps - principle photography, lab processing, editorial and finishing. The beauty was that everyone “knew” what to do with film and knew when to get involved. Craig at JVC was very clear that film is still the ultimate high definition medium and that digital cannot exactly match film’s subtle nuances. Still, the writing is on the wall. Kodak, which has moved away completely from making consumer film has only one factory left making film stock (for features) that is down to just one shift.
Video - The 80s and 90s were all about video tape. Mike from Apple mentioned that back then, anyone who had 2 tape decks could call themselves an edit facility. Camera and deck formats served 6 to 10 years of duty before becoming obsolete. Now tape is rapidly coming to a close as there is no broad-based infrastructure any more. One unit at Fox panicked when they learned that their tape deck format of choice was no longer being made, until they were told that their tapes would still play back fine 
Estimates are that tape has another 3 to 7 years left before being completely replaced.
Data - By 2000, editorial shops really made the data-centric workflow the new reality, with smart post houses using desktop computers in “ways we weren’t supposed to.”
People figured out that the file they were working with did not need to be the same file from glass to screen, so they could have different resolutions for different workflows. As far as toolset “stability” is concerned, data-centric workflows are the opposite extreme from film, with changes in digital cameras, software tools and distribution systems occurring every few months.
The data-centric workflow offers a new killer tool, known to geeks everywhere as “metadata”. Never heard of that term? No worries, you (or your vault) have been “doing metadata” in the analog world forever, simply by writing information about tapes (like credits, format, duration, etc) on slips of paper. Not surprisingly, the digital equivalent of the tape label is much more robust and has many new “features” that can improve your business… if you know what to do with it. The best place is to ask, “What do you do today (with tape information)?” and more importantly, “Why are you keeping notes?” Many of us fall into habits or routines and capture information because “that’s the way it’s always been done.” Once you start understanding the business reasons for the current workflows, then you can make interesting modifications.
CHALLENGES
The panelists explained that the idea of a “learning curve” is almost outdated… because it implies that there is some sort of a plateau where the information levels off and becomes manageable. In reality, each year brings more change and variation faster than the prior year, so you never reach the plateau. There is a growing information gap on both a macro level (Industry) and a micro (company) level. On an industry-level, we need better standardization and unification to encourage interoperability between vendors and a better forum to distribute quality information to the community at large. The panel pointed out that the HPA (through their luncheons and outreach efforts) is a step in the right direction, taking the torch from SMPTE which used to help in these conversations. At the company level, both staff and clients have had a difficult time reacting to the rapid changes in software and hardware.
In order to illustrate the complexities at hand, the panel broke down the workflow into smaller parts, including image capture, post processing, distribution and archiving.
Image & Information Capture (i.e Cameras)
As Ramy mentioned, “It’s crazy-ass right now.” The camera manufacturing industry is in an incredible state of flux. Cameras are more varied (consumer cameras now shoot 720p to 1080p), are more proprietary (every HD manufacturer has own recording platform), change more frequently (cameras or “camputers” change with each firmware update every few months). In fact, firmware updates have to be carefully tested, as firmware problems can break existing, stable workflows. With all these changes, post guys are reporting that some filmmakers simply don’t know what camera they shot with!
The recording media itself offers up a series of issues. Are you using a single format for your shoot or using multiple formats? Does your camera shoot on proprietary data capture formats? If so, how long will those “cards” be produced? Can you pick up replacement cards in the small town you are shooting in? How reliable is the media in extreme heat or cold?
Information about the media is changing as well, as people are now relying on accurate and meaningful metadata downstream. Camera manufacturers are coping with much more metadata than ever before, like GPS information and focal lengths. Apparently, DPs today are “scared as hell” about the new roles and requirements of metadata and how it impacts their production process, especially if they have little buy-in in the process. The main problem is that it is much easier to manage the flow of metadata on set than trying to “fix it” in post. Incomplete metadata means editorial cannot find information as efficiently, so this new data burden felt in the field is quite real.
The panelists were talking about how they sell to all sides, kind of like arms merchants. Leon said that he finally understood why the RED camera looked like a machine gun :)
Post Process - Factory Workflow vs. Handcrafted Workflow
Back in the old days, everyone knew what to do and could offer the same service in a repeatable and sustainable manner. How do we establish some standard paths and expectations in the new data-centric workflow so we can be professionals again?
Managing expectations is always the name of the game. The panel offered up the new workflow analogy of a hamster “Habitrail”, a big series of tubes that connect up with each other in the middle. These pathways share some connection points, but are mostly unique for each client. Post guys deal with “shopping bag” projects which are a pile of different video formats and quality with different distribution requirements.
Fox’s TV shows are trying to figure out a unified post production workflow to gain cost and time efficiencies. Each show has a fairly similar and repeatable production pipeline, so there is a higher probability of developing a “factory-style” workflow.
DigitalFilm Tree’s clients are much more diverse, ranging from first time indy film makers to sophisticated movie producers, meaning Ramy’s teams did not reproduce a single workflow over a year of client projects. He has 2 people on staff who “sit around all day and study” workflows (how does the new firmware update on the RED Camera affect the flow of information down stream?). His challenge is to figure out how to deal with constant change from multiple inputs and multiple outputs and still make a profit.
One other great question came up, regarding how to value the cost of a workflow? When free software causes a major problem with your project, does someone pat you on the back because you saved money?
Distribution and Archiving
Of course, the internet, mobile devices and HD players have redefined how we consume our media, demanding tremendous amounts of always-on content. We want to see different versions, different resolutions and different formats, depending on where we are and what gadgets we have at our disposal. This “eternal sucking sound” of demand places a new level of importance on long term, highly available, high volume storage. We mine archives that are 100 years old and the celluloid will still play. In the tech space, planned obsolesce means that the 1s and 0s of your media may be wrapped in a proprietary playback standard and cannot be opened in 30 years.
Long term reliability is a critical issue for the industry to deal with and right now there is no solid answer. Ramy said that we need full redundancy to safeguard the entire process from production to post before tape goes away.
MOVING FORWARD
“Tapelessness is Next to Godliness”
The panel agrees that post houses that build an environment that can withstand the rate of change are going to be the winners of tomorrow. The Promise of the new data-centric workflow is higher reliability and productivity through newly developed efficiencies. Companies should be able to save money and to be more competitive. The challenge is to figure out where you are in the adoption curve and how to get on board early. We all need to be on the ball, even if it is moving. Digital workflows provide some amazing capabilities:
Better Quality Copies - Digital video makes perfect copies in unlimited quantity, so generation loss is no longer a relevant concern.
Wider Access to information - High level executives used to guard video tapes, especially in News where information hoarding was a strategic advantage. Now wider information access democratizes the work.
Near Realtime Remote Collaboration - faster pipes and better codecs allow people to collaborate creatively across distance.
Standardization - Better Mousetrap Vs. Better Environment
How can vendors build a “better mousetrap” while building a “better environment”? In order to innovate, vendors will break old confining conventions and develop solutions that “lock-in” their clients to increase revenue. At the same time, the panel agrees that widely adopted open standards benefits the industry far more than closed standards. What is the right balance between innovation and interoperability? How do we affect change?
Ultimately vendors need to listen to their customers, so Jim recommends that we as customers demand more interoperability from our equipment. For example, Lego bricks made in 1950s still mate with Legos made today. We need to work towards planned structural compatibility, where devices (hardware and software) can pass coherent metadata any other devices. By clearly asking for more openness, vendors will need to respond.
Transcoding - Glue between Workflows?
2 years ago, Fox faced a fierce internal battle trying to decide what file format to use. The question was do you want homogenous format or heterogeneous? They chose a single-file homogenous format and found that it is really hard to get one size that fits all. The lesson learned is to use transcoding tools that provide multi-format capabilities at different points in the pipeline. The resulting challenge is that transcoding becomes expensive, both of time and money resources. It’s all Insane, but somehow the post industry finds a way to keep up.
Building New Teams
File-based workflows require a convergence of people, cross disciplined teams comprised of post talent, computer folks and networking folks. Ramy’s team has found that instead of being a narrow expert in one area, they work with a wide range of outside experts with deeper understanding of the issues. Finding the resources you need is critical, especially for companies who do not have the luxury of an in-house workflow R&D team, i.e. the “guy in the backroom”. The panelists found that they would often introduce people that work within the same company separated by job roles.
CONCLUSION
The panel shares a conviction that the post industry will continue the decades old tradition of making order out of chaos, taking on a crazy project and delivering great content on time and budget. People have learned the phrase “fix it in post” for a good reason!
With all these changes, the “sky is the limit” and there is no longer a race to the bottom anymore. There is so much you can do with image quality and metadata, it just depends on how clever the team is. The main takeaway is that we need to work together and figure out how to tie together the loose ends to really create a new platform for creativity.
Regards,
Bill