There is no smallest among the small and no largest among the large; but always something still smaller and something still larger.
#5: “It’s Huge!!!”
So, after deciding I just might attempt the crazy, impossible, still impossible and downright foolish task of making an IMAX film, I had to face one simple fact – IMAX is huge. In fact, for years, the IMAX Corporation’s tagline was “Think Big!”. Well, I thought big and it was scary.
For the non-geek, non-filmmaker types reading this (i.e. normal people with lives that see the outdoors occasionally), various film and video formats come with certain size frames. We’ve all likely heard too much about “Hi-Def” and why you should get a higher resolution image. The bottom line is when screens get big, you need a lot of pixels i.e. a large frame size so the image looks nice.
Basically, it’s like when you watch some old program on your new huge LCD HiDef TV that used to look great on your old 19″ tube TV and now, it looks like crap. Well, it’s worse in IMAX. How much worse? A picture is worth a thousand words (or several thousand million pixels).
So what we have here is various video and film format frame sizes compared to the ideal IMAX resolution. You can see why I was daunted. Obviously, you could fit a bunch of the “NTSC TV” size frames (which are 720 X 480 pixels) in one IMAX frame. But look how small the “high definition” 1080p frame is. This is what is touted as state of the art resolution for the best in digital video. The new Star Wars series, Avatar and tons of other projects were shot at this size.
Ouch. And remember – a frame is a square — length times width. So when you get a frame twice the size of another, it’s four times as many pixels. Thus, while a 720 x 480 pixel image contains 345, 600 pixels, the 5600 x 4200 pixel “5.6k IMAX Max” frame contains 23, 520,000 pixels!!! That’s 68 times as many pixels!!!!!!!!!
Big Ouch. But it gets worse. It’s not just the resolution of film compared to video – it’s the color and dynamic range. Normal video is 8-bit color – which mean 8 bits of data represent each pixel. So our 720 X 480 pixel image contains 2,764,800 bits of data or around 2.7 Mb (megabytes). But 8-bit color looks not great going out to film – so you often use 12-bit, 16-bit or best, 32-bit color. So if we want great results, our 5.6k IMAX image at 32-bit will contain 752,640,000 bits of data or nearly 1 Gb (gigabyte) for each frame. That’s 272 times as much data!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Huge Ouch. And when I tried to actually start working at the “ideal” resolution bad things happened. When things worked, they were so slow as to be impossible – spending an hour waiting just to see a frame at full resolution. But mostly, things just crashed and crashed hard. I could not get anything much to work. I had to resort to very simple shots and compositions.
The best I could reliably do was the “4K film” – (see above) in 8-bit color. In 2008, I had the opportunity to do 2 minutes of test footage and 8-bit color looked, well not good. The resolution was okay (4K is still very big) but overall, dissapointing. In 2009, after some upgrades and reducing my test shots to even simpler compositions – abandoning much of what I wanted to do – I was able to render some 4K 16-bit color shots out to IMAX film. 16-bit looked much better but the shots themselves were not what I had envisioned for the film.
So I had to face facts – make the film in 8-bit color (slow, poor quality color and still fight lots of crashes) or make the film in 16-bit color (but abandon most of my original visual ideas for the film. I chose neither. I decided to wait and seek a solution that would allow me to make a film, maybe not the ideal, but at least without compromises that would really hurt the aims of the film.