By Mike Hughes
Issue: July 1, 2005


Criminy!!! We were just getting used to the notion of doing things differently in the film market with the digital intermediate (DI) craze, and now along comes the new DI - for video.

When I heard that all the different companies offering digital color grading tools were adding video I/O boards into their products, I started scratching my head. Had they not built their respective products to address the color grading issues associated to film? Were they not, at the core, designed to solve a key piece of the DI puzzle? And if so, and since DI itself has nothing to do with video, what was up?

The core of the answer to these questions revolves around semantics. While DI actually speaks to a process for eliminating the requirement for intermediate reprints for the film post production market, the reality of the matter is that DI has joined our industry vernacular and has simply come to mean digital color grading. So when someone talks about DI for the video market, he or she is actually referring to nonlinear, PC-based color correction.

Ah, color correction - no more mysterious art exists in the post house today, and no one is more important to the ultimate look and feel of any project than the colorist. His or her ability to translate the clients' desires, both spoken and otherwise, will determine the success or failure of the work at hand, and most certainly whether or not the client comes back.

Color correction is also the last bastion of linearity in the industry. From simple flat passes for dailies and offline through to full blown, client-interactive sessions, the process has been pretty set for many years.

However, the digital color grading systems are being adapted for the video world in an attempt to define DI for the segment - let's call it D2D for the fully digital video post process. It's interesting to note that D2D could also be aptly named the tapeless environment.

In the D2D world, the telecine functions much more like a scanner in the film world - it rolls the film once and deposits the scanned material in a storage repository. If we assume for a moment that nonlinear color correction tools will be employed, this part of the process becomes no different from editing and compositing. It is simply a manipulation of digital image sequences.

An added benefit of D2D is the flexibility of the post production pipeline. Different tools can be used at different stages, and the final look need no longer necessarily be defined at the front end of the project.

So, it's not difficult to see all the advantages that could be derived from a fully nonlinear capability, but there's another side to the story with gating factors that will throttle viability - two of the key areas are tools and infrastructure.

The incumbent providers of today's color correction systems have been at it for a long time. They have developed and refined deep toolsets, most offering realtime functionality (often using custom-built hardware) upon which colorists have come to rely heavily. Until such a time as the digital color grading systems can offer similar capabilities, there will be tradeoffs which have to be weighed against the upside of working nonlinearly.

The infrastructure challenges are not so different from the macro facility issues that already exist in coping with an increasingly digital world where everybody wants to share everything. Regardless of the shared storage solution that gets chosen, it must (a) interface to the traditional telecine in a familiar fashion (i.e. must have a video spigot), and (b) it must provide fast, facility-wide access to the transferred material.

It's also useful to note that if the shared storage arena can be accessed as a true Sony-protocol video device, traditional color correction systems can still play a role in the D2D world. Rather than driving the telecine on the back end, they simply access the scanned material residing in the shared storage pool as if it was video, and write the corrected results back in the same fashion.

Another interesting derivative of a move to D2D is the changing role of the telecine. A one-pass approach means that a telecine's ability to manipulate (e.g. reposition, zoom, pan, etc.) the sequences as they come off is lost - all grading is done after the transfer, not during. The responsibility to perform such operations now falls to the digital grading system.

This may seem like a problem at first, but there are already some who are probing ways to address this, and the solution seems to center around the intrinsically digital nature of such an environment.

Since digital color grading systems are inherently resolution independent (unlike their linear, video-centric cousins), why not transfer the material at higher-than-delivery resolutions and build into the process an excess of data which can be used for the operations noted above? For example, a standard definition job could be transferred at 1K. If such an approach takes hold, the video post world starts looking a whole lot more like that which already exists in film.

In summary, the trend towards an all-data post production world, independent of the resolution at which the work is done, boldly marches forward. Nonlinear is now attacking the last fortress in the realm it had yet to conquer, and video work - be it short form or long - may soon be defined only by the tape that is placed in the client's hand. Hope to see you all at NAB.