Interesting idea. Kind of like VBR encoding for MPEG2--use the bits where they're most needed. Still, it might create some very major headaches (figuratively and literally) to implement with display technology.

I think Microsoft has a demo on their site showing how they are able to fabricate intermediate frames to eliminate motion artifacts. One day in the not-too-distant future I think this will be common. After all, why shoot 60p when you can shoot perhaps 10p and generate realistic intermediate frames to give the illusion of 60p?

By the way, as for which format should be used to ensure future playability, I'd have to say that Hollywood instinctively knows that answer--24f/sec film. Guaranteed compatibility with whatever comes down the pike, simply because the vaults are full of material created in this format. Horrible motion artifacts, but any DP knows how to minimize them with proper shooting technique, and as just mentioned above some sort of intermediate frame generator will one day be able to help.

..I think Microsoft has a demo on their site showing how they are able to fabricate intermediate frames to eliminate motion artifacts. One day in the not-too-distant future I think this will be common. After all, why shoot 60p when you can shoot perhaps 10p and generate realistic intermediate frames to give the illusion of 60p?..

These Microsoft generated intermediate frames would actually be a completely generated image, i.e. a representation of actuality but not actually an image of the real world through a lens. To whom would intellectual copyright of these images belong? 'The camera never lies' would become a moot point, as the capabilities of the software start to have an impact on resulting recording.

The next step would be for the camera to perform all sorts of other image enhancements, using the optical information received throught the lens as just one of many streams of information used to create the resulting recording. For instance, knowledge of the topography and spatial characteristics of the set, information on the physical properties of features on the set (e.g. what textures are to be seen and how they might be represented from different angles), combined with images from multiple cameras, could result in the creation of a complete virtual recording of the subject, able to be digitally recreated from multiple angles or with simulated camera motion or pans and zooms.

I think I'll crawl back into my box now....

Hi Spot.

The road which brought me to my current career, took me through a period as an audio engineer and studio operator, during which I developed a serious interest in motion picture sound. I hung around some motion people sound people, bought a few books and attended a few SMTPE events. I thought some of this info might be in a 1980s issuance of "The Audio Cyclopedia" that I had, but I looked and it wasn't there.

Some of my information came from hanging around Scottsound, and discussions with the late, and greatly missed, Bruce Scott, who had a lot of good stories about early motion picture audio history. I also heard stories from some of the other folks who visited there.

As I recall, the 24 fps was chosen not only for the number of feet-per-second necessary to produce 9kHz without excessive noise, but also because it needed to be some sort rate easily divisible so that an 1800 rpm synchronous motor could be used in a projector. I still haven't figured out how the number 24 works there, but that's what I was told. Apparently, there were concerns that 30 fps would be too hard on the projector's sprocket and claw mechanisms, and 20 was not considered adequate when film resolution, optics, film printing techniques, etc. were taken into consideration.

Considering that most of this came from some old film-audio guys, it is probably suspect -- But it sounded good at the time.

-Travis (Voiceover Guy and Entertainment Technology hobbyist.)