We haven’t been out to the theatre for a while and we really wanted to get something in before year end. The cinema was really the easiest choice so we decided to catch a film which has been making a fair bit of noise lately: “Avatar” by James Cameron staring Sam Worthington, Zoe Saldana and Sigourney Weaver.
This question was posed on LinkedIn recently and I responded in my usual disruptive style. Arguments were made against film because 4k sensors potentially don’t have enough dynamic range because of the photon capture well sizes. I think I covered that in my response reproduced below:
While I agree that the dynamic range issue could be a problem, I think that the argument can get a little confused because you can also shutter more quickly and thus you don’t have to fill the electron well. You can shoot at 50Hz progressive frames instead of the archaic 24Hz progressive frames. You then achieve much better motion rendering and make it much easier to shoot motion than with film.
With film the intermediate process immediately reduces quality, even with the finest sensor you can never do a 1:1 transfer of analogue media without loss of ‘signal’ or introduction of noise. Celluloid noise is a nature of the structure of the molecules deposited on the base, as well as the base itself, so while this noise is ‘analogue’ and thus more acceptable to us than digital noise, it is still present. CCDs are not without noise, but it can be more predictable and controllable as many TK’s demonstrate. I think digital origination has more advantages than resolution in presenting an image that is potentially clearer and more reliable than film.
Furthermore I have walked out of films because of the poor state of the print and so I welcome digital cinema for reliable theatrical presentation as well.
If you want to see a picture that can meet of exceed that of a 1950s cinema epic then take a look at SHV from NHK in Japan (demonstrated at IBC in Amsterdam last year with a real live link from London). Their prototype camera uses four 2.5 inch (64 mm) CCDs each with a resolution of only 3840 × 2048. Using two CCDs for green and one each for red and blue, they then used a spatial pixel offset method to bring it to 7680 × 4320. Aptina/Micron are now making a CMOS sensor for NHK’s next camera which is 4112 (H) x 2168 (V) and that can capture at up to 60fps. I dare say four of those will do the job!
OK, SHV will be another 5 years away before it is practical, but it is leading the charge towards very high resolution capture at decent frame rates and someone has to do the bleeding edge development to set the bar.
D-Cinema is here, it has many advantages which balance out and for me the most important issue is doing away with people saying they need to shoot a “filmic look”, Cinema is my favourite medium but film is a legacy and we need to move on. I have no doubt offended someone with this post, but I think it needs to be said 24p should be put behind us as soon as possible.
This is an article I offered for publication to the BKSTS, but after a couple of revisions it kind of fell by the way-side and has laid neglected on my storage since. So here it is, an analysis of satellite technology from the viewpoint of its use in D-cinema distribution. I hope it proves useful to someone and that some insight can be gained into the various technologies available.(copyright Bob Hannent, do not reproduce without permission, yada, yada, yada…)