David Farmer Special: Game Changer Gear

I’m always amazed at how much R&D goes into visual effects (VFX). Teams can be on for years developing new technology for a single film. It’s easy for we sound folks to feel a bit left out here. No production ever puts out the money or the interest in developing new sound technology for their film. We get hired in much smaller numbers, and are on for shorter periods of time. Where a VFX person can be on the gig for months and only do a handful of shots, the sound team is expected to cover the entire film, with many layers of sound, with only a handful of people. It’s almost assumed that we have a sound or approach handy for any possible scenario. It’s amazing that a production can tell you “We want this to sound like nothing we’ve ever heard before” (even when it looks like something we’ve seem a thousand times, or is just another version of the same old story), and then look at you sideways, and nickel and dime you when you bring up a field recording budget.

I’ve pondered this a lot over the years, and the truth is, we probably wouldn’t benefit that much from attempting new technologies on a per-project basis. (Game engines are another animal altogether though. There is more development there, but I won’t go into that here.) Entire companies have sunk loads of R&D into trying to improve things as simple as pitching, and progress has been slow. The fact remains, we can only twist sound so far before it becomes unnatural and unusable. For things we recognize, like speech for example, the ear is very unforgiving. You can’t process things very far if you want to maintain the integrity. If you’re going for sci-fi or robotic, then that’s much easier to accomplish. Most of the time we’re trying to avoid sounding robotic, and even minor processing winds up as robotic pretty quickly.

Therefore our approach to sound design is often rooted in a simpler approach. And this brings up the flip-side of the VFX vs SFX relationship. For instance in LOTR, there were dozens of people involved in the scenes with the Watcher (the big squid/octopus). There were many shots involved in that sequence, and must have taken them months. I on the other hand, spend $6 on a toilet plunger, and recorded it thrashing around in a creek by my house. And that was the primary element used for the tentacles thrashing in the water. ($6 is misleading, since my recording rig cost about $5000, but that gets re-used.) Tim Nielsen and I also recorded the rubber floor mats from our cars for the flapping, and there other elements too, but you get my point.

So I wanted to take a moment to ponder some of the advancements we HAVE seen. Some of them are new, and some are things we’ve just gotten used to but take for granted. There are more than I have time to mention, but here are some of the things that still make me go “Wow this is cool!”

Altiverb – “….. because you can’t fake THIS!!!!”, I said in the most prophetic & grandiose tone I could muster. The behind the scenes crew were taping us in the Wright’s Tunnels near Wellington. We were worldizing (playing sounds back over a speaker and recording what it sounds like in that space) sword hits, the Cave Troll & Balrog vocals, and doing some other straight up recording in there, all to get that massive reverb tail that those tunnels have. I said that before Altiverb was available. I think it actually was in development, but I didn’t start using it until The Two Towers. The truth is, now you CAN fake that.

There are other convolution reverbs, but I pick Altiverb because from the beginning, they made it possible to create your own impulse response reverbs. So naturally we went back to the tunnels, and made our own reverbs that we could use again & again, without having to haul speakers & record gear back to the tunnels. Altiverb has been instrumental in re-creating authentic environments. And from a design standpoint, you can treat other sounds as reverbs, and get some pretty crazy results. One of my favorites is to use a modified lightning strike as an impulse, and the results are pretty interesting. I usually like to record outside, as those early-order reflections are really pleasing to the ear. With Altiverb, you can record something in a dead room & then fake it as if it were recorded outside. It wouldn’t call it seamless, as exterior impulse responses have proven very difficult to get right. But you can get very close.

Izotope RX – This is worth it for the “Spectral Repair” feature alone. I use this every single day. Where I used to make edits to remove unwanted sounds, now I use Spectral Repair to “repair” unwanted sounds, and I don’t lose any length of the original sound. I usually use this in stand-alone mode as opposed to plug-in, since it’s so much easier to hear what you’re working with, and the interface is easier to work with.

I also pull this out first instead of an EQ these days. EQ’s affect the entire length of a sound, where I can use Spectral Repair on specific spots & leave the rest of the sound untouched. This has changed forever the way I look at my source sounds. Things I couldn’t use before, have become useful. I recorded some wolves for the Wargs in Rings, and there was a lot of chain-link fence movement & also chains in the recordings. I couldn’t use a lot of that material before, but now I can minimize the parts I don’t want to hear, and get a lot more source out of those recordings.

The broadband noise reduction in Izotope RX is also very very good. Using the “C” algorithm, these days I’m perfectly happy saving over the original file. Up until I started using Izotope, I’d keep a copy of the sound before it was de-noised, since there were too many artifacts in the processed sound.

Soundminer – A database that allows searching, auditioning, and VST plug-in processing. This has been a major part of my workflow ever since I discovered it. Very few of the people I know use the VST plug-in feature, but it’s a major part of my approach. Soundminer 4 has 10 slots for plug-ins, with the ability to save preset racks complete with the plug-in settings, and bypass state. Many of the major plug-ins are available in VST, including Waves & Altiverb. You can pitch sounds around & hear them through the plug-in rack, so you can find just the right pitch that’s going to sound nice processed, and send it through the plug-in rack on it’s way into Pro Tools (or Nuendo). You have to be a bit ballsy to commit a process into Pro Tools, but I’ve been working this way for nearly 10 years now. I still do some processing in Pro Tools, but easily 50%+ is done before the sound even makes it into my session.

Pocket Recorders – A sound unrecorded does me no good, and these little gizmos are just great! I have a Zoom H2, Yamaha Pocketrak 2G, and a Sony M10. The one I prefer to use is the M10, because it’s the cleanest of the three. Both the Zoom & the Yamaha, I need to do some broadband noise reduction, but the Sony can get by without it most of the time. I call these my “bird in the hand” recorders. You know the saying “A bird in the hand is worth two in the bush”? Well that’s how I feel about these. My main field recording setup is a Sound Devices 722 with a Schoeps MS mic, and I’ve had very good results with that. It’s not always practical to take out the bigger rig. I’m much more likely to record something (especially if it’s something quick) if the entire procedure is hassle-free. If there were significant audio benefits to using the 722 & Schoeps, I’d certainly do it more often, but the M10 sounds amazingly good for most things. Some of the benefits are:

• One-hand operation.
• USB Bus-powered for downloading the files.
• Small – I take it more places.
• Inconspicuous design (less people interrupting “hey what is that?”).
• Price – they’re pretty much disposable recorders. I’m not afraid of damaging them.
• Off-the-shelf batteries with exceptional battery life.

I simply wind up with more sounds because of these. I was recently on a trip, and had the Yamaha 2G in my pocket (so small I can barely notice it’s there), when I found myself at a fireworks show. I recorded quite a bit of the show, and the file was incredible. I was standing in a great spot, but even compared to a fireworks show Tim Nielsen and I recorded, where we both had our Schoeps rigs & 722’s linked together for a quad recording, this pocket recording was better! The explos were simply massive, and I even had it set to record in MP3 mode. The MP3 mode was accidental, but the recording came out amazing and I’ve had no problem using those sounds.

Many of the pocket recorders record at 96k 24-bit, but IMO that’s less of a benefit than being tapeless. Of course this also applies to a 722 or 744 or Cantor, etc.. Ever since I started field recording, I was hankering for a pre-record buffer, which DAT never had. Let me tell you, countless hours were wasted recording on DAT for Rings. I spent day after day after day at the SF Zoo trying to get animal sounds. And I had to be IN RECORD, the entire time, just in case something happened. Then I’d have to come back, and digitize those tapes in real-time, THEN go back & hunt for sounds. Let me tell you I don’t miss DAT one bit – good riddance to tape! How many hours would I have gotten back if I’d had the 722 then. It would have been weeks of time – no doubt about it. The first show I had the 722 for was King Kong and it was amazing, even if only for the time saved not having to load tapes in real-time.

But back to the pre-record buffer. Many of you know what this is, but for non-soundies, this is a bit of memory that sound is always feeding into, but isn’t being saved until you press record. For example, the Sony M10 has 5 seconds of pre-record buffer. So lets say I’m sitting at a bus station, and there’s nothing interesting happening. I’m in record-pause. Nothing is being recorded yet, BUT once I press record, the previous 5 seconds gets saved at the start of the file and recording simply continues. So I can wait until something like an air brake release happens, THEN press record, and I have still captured that air release. So all those hours at the zoo, I could have waited until after I heard something, pressed record, and still captured the sound, instead of recording hours & hours of nothing between events. Its worth pointing out that the little Sony M10 ($200) has a 5 second pre-record buffer even at 96k, where the Sound Devices 722 ($2500) only has 2 seconds. 2 seconds is not enough for me.

To get the files off the recorder, I just pop it onto USB, then use Synchronize Pro to MOVE the files to a “Files to Sort” folder on my library drive. “MOVING” means it deletes the file from the recorders memory once it’s copied, so I don’t even need to delete the files from the recorder. Pretty slick.

Back when I was a student, I paid $700 for the first line of DAT Walkmen (I think it was the D3), and a box of DAT tapes that were $17 each. An M10 cost me $200 plus about $30 for a 16GB Micro SD card. I’d call that pretty major progress.

Elastic Audio – I’ve been using this a lot the past year or so. I plan to do a video capture at some point this month on how I use it. Music people use it a lot, but it hasn’t caught on to most SFX people yet. It’s far from perfect, but I feel it’s really going places. And it’s very usable in it’s current state.

Pro Tools – I know this seems obvious right? I feel it’s worth a mention because we’ve come to take it for granted. Not only because of it’s features and price point, but it’s market saturation. This has made it possible for people to work & be compatible with nearly everyone else. It’s no small feat that I can be working at my house, and do stage fixes for a project that is mixing in LA. I really should include internet speed in that equation, but having a consistent workstation platform is major.

Digital Video – Here’s another one we take for granted. I arrived on the scene as editing on film was on it’s way out, so never had that pleasure. But I certainly don’t miss rewinding 3/4″ video to build SFX to picture, or printing to 3324’s or 3348’s.

Data Storage – Hard drive space has reached disposable prices. Everyone has hard drive price stories, and I won’t even go all the way back to my beginning to where it was REALLY insane. Even as recently as LOTR though, everyone was impressed with my 3-75G Club Mac Firewire drives. Yep, a whopping 225GB in 3 cases! And firewire! I was really pushing the envelope there actually. Firewire was untested and was failing a lot on many systems with Pro Tools. I got lucky that the drivers that came with those drives actually worked with Pro Tools.

That 225GB set me back $2100, and now you can get 2TB drives for $100. For a while there, storage was increasing at a decent rate, and we could keep more & more of our libraries online. Now, it’s outpaced me. Even recording at 96k 24-bit, I can’t fill up the space I have available. And this is a very good thing, especially for keeping backups. You do back up your work right? RIGHT????!?

Written by David Farmer for Designing Sound