Tim Nielsen Special: On the Art of Economy
[Written by Tim Nielsen]
I want to write a series of relatively small ‘thought for the day’ type articles on a variety of topics. In the first, I want to expand on something that came up in the introductory interview, when I said that my main advice to people entering into their careers should learn when to stop.
One of the things that I love and admire, not only in sound, but in filmmaking and art in general, is economy. And I do not economics. but by economy, I mean simply:
“To achieve the maximum effect for the minimum effort.”
My favorite example of that statement is found in the movie Harold and Maude. I’m going to spoil something, so if you haven’t seen the film, you might want to skip the rest of this paragraph. There is a shot in that movie, I haven’t counted the frames, but I’d be shocked if it was longer than 20 or 30 frames total, that is the best example I’ve ever found. Harold and Maude are sitting near a garbage dump, and she’s describing how glorious seagulls are. There is an insert shot, so short that most people miss it, to Maude’s arm, where you can make out what appears to be a tattoo. A number. And when you realize what the shot is, a concentration camp tattoo, and you understand that Maude survived the concentration camps, the entire movie changes. What was a wacky story of an eccentric old weirdo becomes something a whole lot more powerful. Suddenly Maude makes sense. In one shot, she goes from crazy old lady to concentration camp survivor, and her actions, her very being, suddenly explained.
But for something so powerful, something so important, Hal Ashby made the decision to keep the shot on frame for such a short duration that many people miss it. I can’t think of a director today who would have taken one of the most important pieces of information for truly understanding the film, and letting most viewers miss it. Hal Ashby was an editor before he became a director. And he must have somehow known the exact length of the insert that a percentage of the people would get it, and a percentage wouldn’t. Regardless, the insert itself is such a great reminder in general of how much can be done with so little. One little shot, a second or so in length, can change your entire experience watching this film. I’ve seen the film at least a dozen times. It’s certainly in my top ten of favorite films, and Hal Ashby one of my favorite directors.
One of my favorite books of all time is The Little Prince, by Antoine de Saint-Exupéry. Seemingly a children’s book written by an adult, it’s really a book written by a child for adults who have lost their way. I had never read it as a child, a good friend gave me a copy while at USC film school, along with the Graham Greene novel The Power and the Glory, and Graham Greene was also a master of economy, and quickly became one of my favorite authors. But Saint-Exupéry also wrote one of the most beautiful books every written, Wind Sand and Stars, about his time spent in the desert after his plane crashed. And in addition to those two brilliant books, he’s also the author of one of my favorite quotes, and really the idea behind this post:
“Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away.”
I wish I had found that quote, and understood it, long ago in my career. So to all of you starting out, memorize those words.
In sound, what I’ve found after years of editing, is that after I’ve completely cut a scene, after I believe I’ve added everything that’s needed, I’m able now to go back and delete about half of what I’ve cut. In every case, the result is a much more defined track.
I remember cutting the backgrounds for The Fellowship of the Ring. I didn’t actually have a lot of editing credits before being invited to go down to New Zealand, so I certainly felt a lot of pressure on myself to try and do exceptional work. I mentioned that I love cutting rain. So I was delighted when I saw the scene of the Hobbits entering Bree, and I saw so much rain. Real rain, down-pouring rain, rain on fences, lanterns, thick cotton cloaks. Rain into mud puddles and rain blowing sideways.
So I cut. And then I cut some more. And then a bit more, convinced that every piece I was adding, was adding ‘something’ to the layering of rain. That each new piece was filling in some piece of the puzzle that had been missing until I added it. And soon I had well over 100 tracks of rain, divided into three predubs. The first was for background steady rain, the far-field rain, the stuff that was to provide the bed. The second predub held all the midfield type rain, sweeteners designed to help sell the perspective cuts between all the shots. The third predub the individual rain sweeteners, the rain on the lantern, the rain on cloth, the rain drips from the gate. The problem was that each of these premixes themselves were also over-cut. The background rains had way too many layers, and although I couldn’t hear it at the time, were already muddy and thick, definition gone. The midfield rains were each 10 or 12 tracks layered together too, and I can still picture Chris Boyes, who mixed the FX, staring blankly as I tried to talk him through it all on the mix stage.
Right then and there on the mix stage we started to delete things, there was no other choice. All the material we needed was there luckily, but it became a game to find the pieces that actually were needed, and weed out all the stuff that was making it sound bad. Of course, as I had been editing that scene, I had convinced myself that ever piece I was adding was adding something. Why?
Here we find one of the problem with our brains. Our brains lie to us! All that work, my brain wouldn’t let me know the truth. That most of it should have been deleted before I took it to the mix stage. There were of course other reasons. I had wanted so hard to do a good job. We had a long schedule, and I had a lot of time to cut. I was more worried about having something missing, that I couldn’t see that I could be bringing too much material. But first and foremost, my brain lied to me, and told me that everything I had put there was necessary.
In the end I’m sure about 20% of the rain I cut was all we needed for that scene. As my library has grown, and my rain collection has grown, with better and more articulate pieces I could now probably cut that scene again with half as many tracks again as we ended up using back then.
So here’s an exercise that every editor should do with his or her tracks when they’re done cutting. Go through, track by track, effect by effect, and mute, one at a time, each one. Then play over that section that contained it. If, in the mix of your work, in the overall layering of all your material, you can’t tell when something is muted or not, delete it. Learn early on not to be precious about anything that you have cut, so that you have no emotional attachment to it. So that with an objective ear, you can simply start taking away.
In the case of something like an explosion, my first pass will usually layer about 10 or 12 elements, to build the overall sonic event. When I then go through, muting and playing, I usually find that at least 3 or 4 of those elements, in the overall layering, simply are no longer ‘doing’ anything. So they go. Same for backgrounds, same for everything. If I can no longer hear it ‘missing’ when I mute it, it doesn’t belong.
I find this test especially useful in creature design. Creatures are the thing that I find get muddy the fastest. On John Carter, I recently had to make some sounds for a creature. We had used some of my sounds from another movie in our first temp, but as we started to see animation come in, it was clear these were totally the wrong pace anyway. So I decided to start over from scratch. Our amazing assistant and recordist, Nia Hansen, had recorded for me a lot of new animals, and with those and some of my own, I went about making a new set. But in the end, they didn’t quite work either. What had happened was that once again, I had overdone it, layering too many things, each time convincing myself that I was adding something, when in fact, all I was doing was killing any character, and any dynamics, by layering and layering over and over different animals sounds, I ended up with ‘Generic Screamy Monster Sound’. The difficulty is that when you’re working on it, that lousy brain of yours lies again and again. It took a playback for the director to realize myself that it wasn’t right, and it clearly wasn’t working for him yet as well.
And so one last time, I threw out most of what I had done, going back to the new recordings, but this time, really working with them on a much smaller scale, a much simpler one, finding the right ‘pieces’ instead of the right ‘layerings’, and created an entirely new pallet of vocal pieces. And then I went into editing mode, and trying to limit myself to this new pallet, recut the vocals for this scene once again. And this time it worked. I was happy with the result, but more importantly, the director and the clients were happy. The scene played clean and powerful, it was articulated, and in my opinion, the vocals have a lot of character and sound unique.
The same logic now applies for me to plugins and processing. I know many people, and I certainly do this too, tend to chain plugin after plugin to create elaborate processing chains. When I find myself doing the same, I start bypassing the plugins, and usually I find fairly quickly that the sound becomes better, not worse. Too many plugins tend to result in the same ‘processed’ sound that tends to rob sounds of their uniqueness. We tend to think that each plugin should be changing the sound, making it more unique. But I often find quite the opposite. Each plugin in the chain often reduces dynamic range, often frequency range, and in the end, you’re left with something now called ‘The generically processed version of something that used to resemble a sound effect’.
So that’s my advice. Learn where the delete/mute/bypass keys are, and make friends with them.
And always remember what Michelangelo said:
“Every block of stone has a statue inside it and it is the task of the sculptor to discover it.”
Or this one, which he also said:
“I saw the Angel in the marble and I carved until I set him free.”
When you are done building all your tracks, treat them as a block of stone, and carve away!
That may seem a strange way to think about your own material, but that’s how I’ve come to think of the sounds that I record, that I cut, that I design, and that I mix. They all have to be treated with the same irreverence and detachment. They’re all just blocks of stone, of sound. On Prince of Persia, Ken Fisher and I decided that after he cut and we premixed the big finale reel, that we would take some time, going through the premixes and mute things. We muted a lot. We didn’t want to mute everything before we premixed because possibility some of it would be needed in the final. But we were pretty sure what we would need and not need, so we took the liberty of muting I would say about 20% of our predub material, being pretty confident that once music was in, once everything was playing, that you wouldn’t miss it. It was never heard again. Not once did anyone on the stage call for it, and it was never missed.
Now to go back and delete about 20% of this article before it get posted…