Session Opening Remarks
Speaker Leon Silverman
President, LaserPacific Media Corporation

I've had a recent very personal example of the challenges of archive in the digital era. After more than 21 years in the same office - the same physical space, I was scheduled to move into a new office. The LaserPacific management team, whose offices had been spread out in a sprawl of 5 interconnecting warehouses and buildings, were, after 21 years, going to be moving to a common office space. As I packed up my office to make the move, I had a number of interesting and profound realizations. The most profound was the realization as I was packing and alternately tossing the detritus of 21 years of inhabiting the same space, that I had spent more time in this office than I had spent living at my parent's house. For some reason, that really affected me. And as I was crouching down, reaching into the bottom of my desk drawer, contemplating my own mortality, I found it. That large plastic enclosure, which thankfully had that lame little key still in it, lest I would have to take a hammer to open it, which contained my first digital archive. Now I always thought that the words "digital archive" was an oxymoron - a contradiction in terms. For those of us in the film industry we have all come to learn that film was the only true archival medium, but here I was at the dawn of the digital era and I was determined to make sure I had a plan. I was a very early computer user who had learned the value of back-ups almost from the start. I say almost because I was also a very early computer user who learned that computers crash and all of their data can be lost in a blink of an eye. After I blinked out more than a few early computer user crash and gone tears - I decided to start my archive.

Relatively religiously I copied those documents so as to make them safe from the pernicious computer gods. And now in my hands from so long ago, was the evidence of my forethought.

And of course, I had to smile. In my well-meaning intentions to protect my documents, I had secured for the ages, those quaint looking 8 inch floppy discs - you know the ones that really were floppy. In my mind, I rapidly stepped through the evidence of progress which would further await me in another desk drawer - 5 and a quarter inch floppies, 3 and a half inch floppies both the big capacity ones and the little capacity ones that came before, Jaz drives - both the big capacity ones and the little capacity ones that came before, zip discs - yeah big and little, CD-Rs and of course my current BIG favorite - DVD-rs. Now I could have actively managed my archive and migrated my back-ups, but I didn't… and now for my early files, there was this other little matter. My 8-inch floppy contained WordStar files, and since my archive predated DOS, Windows or Mac OS, was a proud product of the CPM operating system.

Now fortunately for me, in 21 years, I had never once had the inclination or the need to restore this archive for any purpose, nor would I ever have the opportunity, as I quite literally tossed that plastic container into the trash bin along with the other detritus and long forgotten mementos of my working life.

Of course, my rambling tale is hardly lost or new to many of you who deal with this issue every day of your professional career. But your issue is different - very different. You want to ensure that the elements that might be digital can not only be stored, restored and actually used - you want to make sure that the industry understands its role in this transition that is already well underway.

Once it was easy…or maybe I should say easier. There was film… and without minimizing the complexity and challenges of dealing with the relatively significant variations in film formats, aspect ratios and element diversity, because of the work of many of you here, our industry has begun to get its hands around its film history and recent past. The awareness of the need for elements that can be archived and restored has resulted in the upsurge in film separation masters and protection elements and the care in which many films are now stored.

Then the multi-format head of the tape hydra arrived and the challenges of keeping track of the progression of video elements entered a whole new realm. For in a relatively short 25-year time span, our industry has undergone a number of significant changes in the nature and type of video form factors, encoding schemes, frame rates and production/post production infrastructure. The analog 2" era which started in the 1950s, begat the analog composite 1" era in the late 1970s, and as analogue component such as Betacam and then Betacam SP (and its less popular, but nonetheless viable Panasonic cousin M2) in the early 80's, was replaced by the short-lived Digital Composite D2 (and its less popular, but nonetheless viable Panasonic cousin D3) in the late 1980's which co-existed with the Digital Component D1 mastering format and ended up duking it out for standard definition Digital Component supremacy with the slightly more compressed Digital Betacam (and its less popular, but nonetheless viable, for a time, until they stopped making it - Ampex DCT) in the early 90's. Of course in the past 10 years, we have added the HD variants as well, starting with Analog 1035 line HD, 1080 HDCam 60i, 24P and the finally, increasingly popular and very viable Panasonic cousin D5 with its 720 60P form as well. Oh yes, there is that D6 HD format, the one with the 32 rotating heads that never really did catch on here in the US, but is in fact being used as a mastering format today in England and France. Now of course, this does not include all the analog viewing formats such as VHS and 3/4" or the other "D" format such as D7, D9, or Betacam SX, IMX and DV whether mini or not such as DVCam, DVCPro, DVCPro 50, 100 and DVCPro HD. I think you are getting the picture. Although in the future will anyone be able to play these pictures. In the future, the media archeologists will dig through the layers and find a very rich and thick film bedrock and increasingly thin and sparse layers of the tape format de jour.

This past 25 years has been marked by filmmaker's discovery of the power at first, of electronic and then, of digital video tools. We have seen how film and video have merged and now coexist as the preferred production/post production workflow for television, music videos, commercials and other media content. Media creators have embraced this workflow without regard to issues such as tape format, color encoding scheme, bit depth or the ability to retain or even replay these tapes in the future.

But that's actually not the problem or the whole story of why we should be worried. While the video world is a mess of formats and choices, it's a well-defined mess. It for the most part, leaves us with the migration challenge that we pretty much have already signed on for when one starts thinking about how to handle the video format proliferation.

The real challenge is at our doorstep - and its name is data. For as the world of film and video, even digital video, becomes data-centric, the task of storing and restoring becomes increasingly even more complex.

In 1992, with Kodak's introduction of new scanners, laser film recorders and Cineon software, the industry was able to scan film as digital data, manipulate it and record the data back out to film. These tools enabled images that had gone through a digital process to be seamlessly inter-cut with film that had not been digitized. The introduction of these tools heralded the beginning of an explosion in digital visual effects. While many effects houses found ways to back up and archive some of these effects elements, to most, the film element itself was what was considered archival. Tape back up formats became victims to the steady march of progress in the computer world as formats become obsolete every few years when faster, higher density tape schemes would inevitably arrive. Some of the early popular storage formats such as Metrum and Syquest gave way to DLT and then DTF. New formats such as LTO, AIT, and SAIT begin to make the prospect of effects backup and archive formats as daunting as videotape. But our data centric image world is no longer just about visual effects.

As filmmakers such as Gary Ross on Pleasantville and then cinematographer Roger Deakons and the Coen Brothers on Oh Brother, Where Art Thou? discovered how scanners and digital "photo finishing" tools could be used to enable a new color palette and creative freedom in film timing, the interest in what is now known as digital intermediate was born. And motion pictures will never be the same again.

It is estimated that within the next 3 to 4 years, nearly every major motion picture will be finished digitally. The current so-called DI workflow, which has grown from a visual effects model continues to evolve.

Though Cineon and the closely related SMPTE-developed DPX file format became relatively ubiquitous in the visual effects world, the notion of a data file that looked the same on different platforms or displays proved illusive at best. This is especially true of the digital intermediate process as well, where custom workflow and custom look up tables form the basis of competitive differentiation. As feature film post production becomes a digital film finishing process, the negative created from the digital files will of course become an archival element. Separations from this element, made either optically or from digital files directly to separations will form a common strategy for archive. But what about the data itself that creates this negative? How should this data be stored and on what? What about the files that are part of the work in progress that creates the final elements? What about the information applied to the raw scans of the original film that contains the color correction and image optimization information? Should this metadata be retained? Can it be retained? Is it or will it be of any value? For surely there are, at this point, no cross industry TDLs (timing decision lists) that can describe color attributes that are platform or device independent. And surely there is no agreed upon way to characterize different scanners, color timing devices and film recorders so that the information regarding how film was scanned, timed and filmed out could be retained for future or cross platform use. There are no recommended practices for retaining this data, and metadata. Should there be? Can there be a way for the industry to create some expectations so that the digital process of film post production becomes more predictable and still leave room for the "secret sauce" that currently differentiates competitors? Can we describe a data file with baked in rendering intent for the archive and leave at the same time an instruction set that allows us to unwind the ball of string for future restoration? The current workflow in digital intermediate is for the most part a 2K workflow. As higher resolution scanners become more practical, surely we will want to create elements at the very highest quality level possible. Will there be a time when we can be assured that a very high resolution scan and workflow can reproduce every nuance that it is the film and thus the digital data itself be considered archival? And when this is practical, as I believe it will be, will we not need a way to consider the digital data, the digital elements themselves archival as we consider film today - after all the data is the master record?

I have recently been involved with a project that was to be shot both on film and digitally. The goal was for Michael Mann to shoot the movie - Collateral with Tom Cruise for DreamWorks mostly on film with about 20% of the film shot digitally on the Thompson Viper camera. The plan was to record the digital data on hard drives so as to keep the data "pure" and uncompressed and not in what was perceived as a compromised "video" form. Elaborate plans were made to back up the anticipated 2 1/2 Terabytes of digital data shot each day from the portable hard drives on the set to another set of hard drives and SAIT tape as a back up. As the hard drives were to sit nearline for more than 6 months until they were needed to conform the film, it was determined that it would be appropriate to "spin" the drives every month to check the veracity of the media. A workflow that was cumbersome, but doable was developed, even though to everyone it felt a whole lot like standing up in a hammock. But this was a way to give the filmmakers the "pure" data path that they thought they needed. Budgets were created to outline the cost of this "digital lab" process and service, and this is where the plan fell apart. It turned out to be more than 10 times more expensive then a Sony HDCam approach and more than 3 times more expensive then a traditional lab and telecine approach. In the end, videotape saved the day. Sony's new HDCam SR ironically enabled the Thompson Viper camera to shoot in 4:4:4 and record to a mildly compressed tape format. The pure data model was discarded in favor of a data on Sony tape model. And it probably was a good thing. Because instead of shooting 80% of this film on film, 80% was shot digitally - the equivalent of well over a 100 terabytes of data.

But this is not a tale of how everything worked out in the end and they all lived happily ever after. Because the day is quickly approaching that a digital data camera WILL be practical. That we WILL record digital data to something? And then what? On what will we store the equivalent of the original camera negative? And in what form? How will we know that we are making good choices? How will we know that our digital movies will not be stored on the future equivalent of Word Star files on 8" floppies in the CPM operating system? How can we, with our eyes open begin to find ways to understand how the film world which met the video world will morph and transition to this uncertain digital future?

It is at events such as this that we can begin to illuminate and educate ourselves so that we can help and inform those that want to promote technology and those that want and demand new creative tools. Perhaps in concert with industry standards and recommended practice efforts inside organizations such as SMPTE and in concert with like minded concerned professionals such as within AMIA, the Academy Technology Council and the ASC Technology Committee that we can help create a future that is not only worthy of its past, but will also allow our images to be stored and restored so that one day we may truly say the words - digital archive and know that there is such a thing.



SPEAKER BIO

Leon Silverman, President of Hollywood-based LaserPacific Media Corporation, joined Eastman Kodak when the company was acquired in 2003. In addition to his role at LaserPacific, he serves Kodak as the Director of Strategic Business Development, Entertainment Imaging Services and a Vice President of Entertainment Imaging. For the past 26 years, Silverman has helped introduce new technology to Hollywood. He played a key role in the establishment of LaserPacific's Electronic Laboratory, which pioneered many of the tools and techniques that are now the standard for the electronic post production of film. In 1989, he was recognized for his contributions to the creation of the Electronic Laboratory when LaserPacific received an Emmy Award for Outstanding Achievement in Engineering Development.

He was also recognized in 1996 for his contributions to the development of LaserPacific's SuperComputer assembly process, which was again the recipient of an Emmy Award for Outstanding Achievement in Engineering Development. Silverman spearheaded LaserPacific's entry into MPEG digital compression in 1992 and created LaserPacific's DVD Authoring and Encoding facility. He has also helped to establish LaserPacific's pioneering efforts in 24P High Definition.

Silverman is currently focused on new digital post production methods and technology which will expand LaserPacific's role in the motion picture community as well as Kodak's role in the digital services business.

Silverman is the current President and a founder of the Hollywood Post Alliance. He is a Manager of the Hollywood section of The Society of Television and Motion Picture Engineers (SMPTE) where he also chairs its Education Committee. In November of 2003, he was named a Fellow of the Society.

He serves on the Board of the Entertainment Technology Center at the University of Southern California and is also a member of its Technical Advisory Board. He is an Associate Member of the American Society of Cinematographers, where he sits on their Technology Committee. Silverman has served as an Instructor at UCLA Extension where he taught Post Production for Film and Television as well as an instructor at the USC School of Cinema-Television.