THE NUMBERS GAME
Ok, just a couple of more things about UHD, compression, streaming, benchmark people, and how numbers dictate everything and don’t mean anything.
Way back in the day, when the motion picture industry decided to give this new fangled talking pictures thing a try, they had to make some changes to the process. The biggest change, (to the pocketbook of the Producers), came when it was decided that the jerky 18 frames per second that silent film was recorded and exhibited at would not be simpatico with this new synchronized sound format. So what did they do? Did they gather all of the greatest minds together and ask for a consensus on what would bring the greatest quality and therefore the most enjoyable viewer experience? Possible, but if they did, they immediately ignored the findings and went with: “What’s the least frame rate we can get away with to make the action look smooth and synchronize the sound?” That answer was 24 frames per second recorded with each frame within that 24 being projected twice within the given second for a total of 48 views a second.
Why’d they do that? Money. By moving to a 24 fps rate they had increased their elemental costs by 33%. That was a lot of money in the day. They didn’t want to go any higher, remember, there was a depression starting when all of that transpired. Add on to this the extra size and weight that 33% more film added into the distribution methods of the day, plus, to mention the fact that projectionists, back in this time, made more money than plumbers and had very strong unions that didn’t look kindly on extra workloads for their members and you come up with the businessman’s algorithm: Q / $ = LCD or Quality devided by Cost equals Lowest Common Denominator.
What does that have to do with media today? Now we are focused on file size and bandwidth not film rates and union work rules.
Well, it’s exactly the same thing. If you look at film as a storage system, which it is – you can collect and store data, which can then be retrieved – the essence of WORM – plus, the old school distribution and projection is identical to bandwidth and bit streams. I mention this colorful exposé as analogy to show that human nature, like water, will always take the path of least resistance, especially when it comes to spending money. Remember the formula: Q / $ = LCD? Well, that’s going to ultimately dictate what the numbers will be in the compression/distribution scheme for UHD.
Now, what are those numbers?
We don’t really have any practical application numbers yet to work with, as the technology is in its infancy but we can draw from some parallels. Blu-ray DVD is the HD standard disc distribution system and H.264 has been the de facto codec for compression of media to that format. A 90-minute HD film encoded to Blu-ray could have a file size of 20GB. A UHD version of that same film, encoded with H.264 could easily have a file size of 250GB or more. That same film encoded with H.265, based on a compression scheme that runs about 60% of H.264, you have a file size of around 150GB or more. These numbers are in any way hard and fast, merely used as reference points to an ever changing and completely unsettled, at this point, playing field.
To the question of bandwidth and streaming, again, Q / $ = LCD. I’m sure there are idealized numbers out there in some utopian media world where everything streams at 120fps and full bandwidth without any buffer under runs, but here in the real world it’s going to come down to the guy in the projection booth dialing up the speed of the projector all the while asking the guys out front, though the squawk box: How’s it look now? At present there are neither comparative numbers nor scenarios to go by in this area. The numbers currently being bandied about in regards to UHD streaming are 10 to 15 Mbps. Again, the problem lay not in our stars, but it the newness of the infrastructure – to paraphrase Shakespeare badly.
NETFLIX AND BENCHMARK PEOPLE
Is 15Mbps Good Enough?
Netflix recently started test streaming some of its product in UHD or 4K to some very good results. And they did it with one of their top shelf shows, House of Cards. Whether or not they realized it, this was the perfect show to use because the success of these tests gives us a very definitive answer to that question:
Yes, 15Mbps is, not only good enough, but should exceed everyone’s expectations as a broadcast rate.
To clarify, one need look at the creator/show runner of this series –
David Fincher began his career as a cinematographer who moved into directing commercials (a standard move for cinematographers) and finally into feature filmmaking and now, whatever strikes his fancy. His profile rose when he made some running shoe commercial and he was given his first opportunity at big-time filmmaking with one of the “Alien” sequels – which he infamously almost bankrupted by going way over budget on production trying to get the “perfect shots”. He was quoted as saying at the time, whilst in a war with the producers, one of whom was a friend of mine, that I am trying to create shots here that maybe 10 or 11 people in the world will understand what I am doing – to me that’s worth it.” Of course the film was finished and it was beautiful and Fincher went on to create a great portfolio of other beautiful films. I relate all of this to show that, in Hollywood, David Fincher would be considered a Benchmark Person. Someone who has set a standard for quality that will not be compromised. In his case, especially when it comes to the question of quality images. So, if he signed off on 15Mbps then you know, with a certainty, that the image going out was more than acceptable.
We should start an awards program for BenchMark People! I can think of several right off the top of my head. And you’re not one of them – unless, of course, you are.