The Great Debate09 Dec 2008
The question always comes up, doesn't it? Shoot on film or on HDV...
Looking at beermat budgets for a 12 minute short film, indicates that film would require an additional €2,000 over just shooting in HDV.
In terms of image quality, there are a lot of different factors at work. In simple terms, pixel count is around the same. Motion picture film can resolve to about 6um, which gives an image width of 1,700 pixels for regular 16mm and 2,080 pixels for Super-16. HDV 1080p gives us 1,920 pixels across. Where video traditionally falls down is in terms of interlacing and colour subsampling. Interlacing is a great idea for broadcast television, and sinful everywhere else. Let's not discuss it further. Put 1080i away unless you're shooting for broadcast.
The human eye responds to light logarithmically. So does motion picture film. Video responds linearly. A DPX file, which is the digital equivalent of a frame of motion picture film, uses a 10-bit logarithmic representation of the colour pixel. According to SMPTE, you need a 14-bit linear resolution to accurately capture 10-bit log. That's per colour. Unfortunately, most video uses 4:2:2 chroma subsampling to throw away half of the colour information in the belief that we the viewers won't notice. We will. Ideally then, our High Definition cameras would record 10-bit logarithmic or 14-bit linear, with full resolution for all pixels. 4:4:4 in other words.
Forget 29.97 and weird abberations therein. We'll shoot 24fps. Because that's what we're used to. We'll accept 25fps but only under duress.
But does it all matter? I uploaded some Beta-SP footage to YouTube recently and you'd never think it was shot with a Sony video camera. It looks like film. The secret was the DoP who shot it like it was film. It's 25fps, 4:2:2, 720x576. Interlaced!
So, is that the secret? We're all familiar with that butt-clenching feeling you get as producer or director, when the 16mm camera starts whirring. It's a heady mix of fear and exhilaration. Knowing that you're filming again is always a buzz. Knowing that next months rent is slithering across a metal gate is a whole other feeling. It focuses the mind in a way that few other things can.
It forces you to be prepared. To know exactly what it is you're going to shoot when you turn on the camera, and exactly how you want it to turn out. It forces your DoP to make sure that the frame is properly lit and properly exposed, at a minimum, and that the shot is going to leap off the screen when projected. It forces your actors to bring their "A" game like nothing else. So often in rehearsals, the performance seems a little flat. You produce a film camera and when it starts whirring, the performance lifts by two or three levels. Everyone has that sense of urgency and immediacy with film which video fails to produce.
Oh well, shlep another HDV tape in the camera and let's get on with take 42.
When I shoot film I tend to be conservative in terms of shoot ratio. I have often let one take suffice, if I'm confident the performance hit the mark and there's no hair in the gate. I know I have the scene from other angles and I'd prefer to save the film budget for other scenes. On video, you don't need that type of frugal thinking. Why not shoot it again?
As an editor, I find it very challenging to watch two almost-identical takes, and try and decide which to use. A high shoot ratio conspires to make the editors job much harder. So many options, and so little to differentiate them.
So, here we are, a few hundred words later, and no closer to a decision.
We've looked at the technical motivations, the production motivations and even the post-production issues. What about distribution?
On the plus side, arriving at a festival with a film can under your arm seems most appropriate. You're pretty sure they can't screw up the content as long as the audio system is working and the projector is in focus. some years back, I attended a film festival and they asked for an S-VHS copy of the film as a pre-screener. I sent them a dub which wasn't exactly ready for prime time. I arrived at the festival with my 16mm film print, and was surprised to find that they weren't at all anxious to receive it. It turns out, they were planning to screen the S-VHS dub I'd sent them. Arghhh!!!
All the other short films were in S-VHS. I was the only mug with a print. I had to twist arms, bribe projectionists with beer, and generally persuade the festival that the S-VHS I'd sent them wasn't suitable for video projection. They agreed to screen the print. Phew!
Except, as I mentioned, everyone else was on S-VHS. They fired up the film projector to show my short, and no-one had bothered to check the sound. It was bad. Really bad. Muffled, garbled, unintelligible, and so on. I doubt they threaded the film correctly either.
The real problem with film though is that it's an awful distribution mechanism for shorts. Answer print costs are colossal. You need to shoot titles which will require a rostrum camera, you'll need a negative cutter, a sound mixing and dubbing studio, and don't even think about supers. Not to mention the fact that you're limited to R-G-B colour correction. If you decide to distribute digitally, on DigiBeta or the like, and assuming you are editing on Final Cut Pro or its equivalent, you can perform primary and secondary colour corrections, travelling mattes, supered titles, do your own credits in a matter of hours rather than weeks, and of course, do your own sound mix.
In reality, it's a no-brainer. Distributing short films on 16mm prints is passé.
Some film festivals though would like to see a 35mm print. That means either a blow-up of the 16mm original, an answer print, film-out for the titles and opticals, and somehow redoing the colour correction. Or, you could push your edit suite to produce DPX files and just film-out the entire project. I'd go with the second option. The quality may suffer, depending on the interim formats (prior to DPX), but it does mean you end up with pretty much the same film as the digital copy above.
But should you shoot HDCam/HDV or should you shoot 16mm/S16?
The debate continues...