David Farmer Special: Perception of Sound

I’d like to talk a little on the perception of sound. We’ve mentioned how intangible it is, and therefore our perception of it is wide open to our own interpretation. I like this definition of Perception, as found on Wikipedia:

“The processes of perception routinely alter what humans see. When people view something with a preconceived concept about it, they tend to take those concepts and see them whether or not they are there. This problem stems from the fact that humans are unable to understand new information, without the inherent bias of their previous knowledge. A person’s knowledge creates his or her reality as much as the truth, because the human mind can only contemplate that to which it has been exposed. When objects are viewed without understanding, the mind will try to reach for something that it already recognizes, in order to process what it is viewing. That which most closely relates to the unfamiliar from our past experiences, makes up what we see when we look at things that we don’t comprehend.”

And this very accurately describes one of the most profound career-changing moments I’ve ever had. On one of the earlier shows I was doing at Skywalker Ranch, I was perusing the sound library (which was then on audio CD). I can’t remember the show, but it might have been Con-Air. I was looking for some air releases. I had some pretty good ones in my library, but for obvious reasons I wanted to explore the Skywalker library, being a major fan of the sound work coming out of there. I found an ambient track of a garbage truck. You know one of the ones that has the lifts that pick up and dump the contents into the back of the truck? It started in the distance, and had these fantastic squeaks & wronks. It was just an amazing sequence of sounds that wove in & out of each other. The truck pulled forward to its next stop, slightly closer. Air brake release, hydraulic arm reaches out, grabs the bin, air release, raise the bin, clunks several times, engine revs, hydraulic arm lowers, bin clunks onto the ground, hydraulic arm retract, air brake release & the truck pulled forward even closer to repeat……. Now only a sound geek like myself could sit there in awe of the sounds I was hearing. You’d think I was listening to Mozart or something, but no, just a garbage truck. But I remember being mesmerized by the uniqueness of what I was hearing. The wonderful ambience & distance of the air releases on a quiet morning. I remember saying to myself, “How the hell do these guys find these things?” I was amazed, but at the same time very frustrated. I’d done a lot of recording myself, how come they repeatedly find things to record that have such interesting qualities? These were not just ordinary air releases, they had a voice, a tone, a nature that changed each time. Why was it that these guys could find these things to record but I could not? I kept listening, it sounded like the truck was approaching the recordist position, yep it was. Cool! I was going to hear these things close up! And then it happened. “Wait a minute… this sounds like my garbage truck… yep that sounds familiar… this sounds just like mine… this can’t be mine… what the @#$%?” I backed the CD up a couple of times – yep that was my garbage truck alright. At some earlier point, I had sent up a DAT of some train recordings I’d done using a DAT walkman and an Audio-Technica AT822. I had forgotten, but I’d included this recording of our neighborhood garbage truck.

Now before you get the wrong idea here, this is not intended to put down any Skywalker work, in fact I remain a loyal fan. This incident changed none of that. In fact I was flattered that one of my sounds wound up in their library. But what did forever change was how I perceived sounds, and not just my own. Why did my own sound, sound at least 10 times better to me, when I thought it was theirs? I’d heard and used portions of that sound many many times before. How did I fail to recognize it as my own for the first minute or so, when I always paid so much attention to detail? It was like an alternate reality, and I’ll never forget how that reality morphed when I realized I was listening to my own sound. I remember just sitting there for a good long while, like someone had clubbed me over the head. Clearly I felt a bit bothered that I had been so duped, but what I took from it was what was possible. I realized that I HAD in fact been able to get amazing things recorded, even with off-the shelf consumer-grade gear. Those moments hadn’t eluded me like I’d thought. They weren’t reserved for just a select few. Doing the level of work I had always aspired to do was suddenly within reach. I felt more free to embrace my own style and stop trying to copy-cat.

There’s a book titled “The Holographic Universe” by Michael Talbot. In it, he describes reading a street sign. He recalls reading it as one thing, based on the color of the sign and what he expected it to read, then literally watching the letters re-arrange themselves while he was looking at it to say something else. This was exactly how I felt when I recognized that garbage truck. Our predisposed notions provided a stimulus and an observation that was as real as any other. And so what we expect something to be, is what it is, unless something occurs to change it.

It’s in the details – or not

How many times have you worked on something, and thought it was the best thing since sliced bread, only to listen the next day and think its garbage? I’m guessing this has happened to pretty much everyone. Fortunately, the reverse is also true. We can work on something we think is crap, finally give up, then come in the next day and say – “hey that’s actually really quite good!”. Now detailed work is very important. It’s very difficult to get an inspired track without it, but it’s easy to find ourselves operating on an area of detail that only we the creator will ever notice, and is of little or no consequence to the final product. I’ve seen people work for days on a scene, and really hard I must add, to make only minor changes or improvements. I’m certainly guilty of this too, and it’s because of this that I’ve devised my workflow to minimize this.

In a linear scene (film style), everything is in motion. Things go by fast and only once, at least to the audience. When we’re working, we stop & start, go back & forth, and go over scenes & shots again & again & again. We notice details that the audience will never pick up on as the scene whizzes past them. It’s import IMO to not get bogged down on the details that are of no consequence, or worse – distracting, to the scene. What is this scene trying to do? What do we want the audience to know, or not know, at this place in the film? I like to watch the scene and build according to what my eye is drawn to first, then go back & watch again. Again I build according to what my eye is drawn to. I do this until I feel nothing is missing where my eye is drawn. This is the base, the part that must be covered, the part that will be distracting if not there. Then comes the more abstract pass. Is there something the scene needs to play on the emotion? Are there details that can help put us on screen? Is there anything I’ve done that is already distracting? These are purely subjective, and it is here where the art can change direction the most. There is a scene in FOTR, in the mines of Moria, where Gandalf is reading the book to the group. I wanted to make sure we heard those pages crackle & sound dirty. When he runs his hand over the pages, they needed to sound gritty. When he blew the dust off, flecks of dirt & grit needed to land on the floor. All these things help convey that the book was old & not a prop. These are the worthy details IMO, and they DO take time and a great deal of thought & review. Design is a lot more than just the big loud hero moments, the subtler details are just as important, as tricky, and satisfying.

When deciding what elements are helping, particularly when creating a single event, I use a very quick test. It’s pretty easy to get caught up in adding too many elements to create something. So to test an element’s “worthiness”, I’ll play everything together, then mute that one element. If I have to play the sound more than 3 times, muting & unmuting, to decide if it’s making an improvement, it’s not significant so I can lose it. If I can’t hear it at all – it’s a waste of VU energy and it goes bye-bye.

Sonic Memory

How many times have we had this happen? A director listens to something, and buys off on it. Then s/he hears the exact same thing later or another day, and says “Hey we need to take a look at that section”. I’ve come to the conclusion that this is not because they can’t make up their mind (OK sometimes that is the case), but because things are perceived as sounding different when heard on different occasions. Whatever they were focusing on the first time, affects how they hear it. If they change their focus to something else the next time, that can easily change the perception of what they hear. And I’m not talking about focusing just on sound. It can be anything. They might be thinking about a VFX shot, or an actor’s performance, or did they hang on that shot too long…. It can be anything.

Once I was in a playback of an entire AB reel, for final fix notes (project & names withheld). Director, Producers, Mixers, Supervisors, we were all there. I noticed pretty quickly that something was wrong so I went to the machine room. It took a few minutes to figure out that the FX stem had been patched incorrectly so wasn’t coming up on the console. The first 9 minutes of the reel had no sound effects, only ambiences & foley. This was not a walky-talky either, there was big action. So after the reel was done, the lights came up, and we proceeded to notes. I didn’t want to call anyone out on this faux pas, and while I was searching for a tactful way to bring it up the director went right into the notes. We went thru the entire reel, and not one person noticed that there were NO SFX for the first 9 minutes. In fact the music supervisor made the comment that it was “the best reel yet”. I’d say this is a pretty far-out scenario, but still it was fascinating to me that noone else noticed. I can imagine a scenario where the foley had the SFX covered and that would be why, but that was not the case this time. So how great do the distractions need to be to not notice 9 minutes of missing SFX? Were the images so compelling that they thought they heard something that wasn’t there? Did the memory of going over & over the mix fill in the gaps? I can’t say, but it was certainly a very interesting phenomena.

Is louder better?

Hell no. But your chance of getting something approved increases drastically if it is played louder. And not just to clients, this goes for to sound people too. Louder just makes things appear more “there”. I certainly do NOT think louder is better, nor do I know anyone who would say that. However I prefer to make sure something is heard clearly before doing a fix, if at all possible. I can’t tell you how many times I’ve been asked to replace something in a mix, only to wind up right where I started from, just turned up so it could be heard. This is not a slam at SFX mixers either. Frankly I’m amazed more things aren’t lost in a mix with all the tracks of SFX hitting the desk. It’s a lot to keep up with.

But in design, and also mixing, this phenomena of “louder is better” is a dangerous perception trap. Lots of times people go for the compressor or limiter and squash the snot out of something, thinking it’s helping, when all they’re noticing is its getting louder. Typically when something get compressed, the output goes up. The dynamics go down, but the output goes up. It sounds louder, and therefore we hear more detail, but we haven’t noticed what else happened during the processing. Some plug-ins have an auto-match function (SoundToys “Decapitator” is one), that can automatically adjust the output to as gain is applied. If you’ve never used a plug-in or device like this, it would pay off to give one a try. The first time you try one you’ll be astounded by what you hear. When you apply the compression or drive, the output level will not increase, but instead you’ll hear what you’re actually doing to the sound. You might not like what you hear.

A mic is not a camera

This past weekend, I went to the Reno Air Races to record. It was a very last-minute decision. For the past 10 years or so I’ve wanted to go, but it slips my mind and I don’t get a press pass in time, or am too busy to go. This year I happened to check just last Thursday and saw the races were going on through the weekend. So I sent out a few inquiries where I might be able to get some good recordings. My main question was “Where can I go to get away from the PA?” One response was to get a pit pass as the PA is not so noticeable there. I had been to an airfield years ago where a P51, Corsair, and Harrier had started up & taxied right by me, so I thought maybe that was a great place to be. I also got another tip for a place outside the event. So I convinced myself that I’d get good material, and decided to go. I’d never been, so didn’t know what to expect, but in the worst worst case I figured it would be a great scouting mission for another year.

So off I went. Getting into the pits was easy & not expensive. And the tip was correct, I was able get away from the PA. However, I wasn’t able to get next to anything I wanted either. They don’t start the engines in the pits. They tow the planes about 75 yards out onto the tarmac & start them there. It’s too far away for the recording I’d hoped to get, not to mention all the other sounds. Typically there are other events going on, like acrobatic shows, etc.. So even with my trying to record a 75 yard away idling P51, there were far more other things that stomped all over it. There were gas trucks, tons of golf carts, other planes, photographers, voices……. And also the planes that were racing never got any closer than about 200 yards to me. No good.

Now I should’ve known better than to expect to get anything as part of the general crowd, but I tried anyway. It was just another reminder that people that don’t work in sound have no idea what it takes. They think what they’ve seen in movies is how it works, or that you can record something from 200 yards away and make it sound like it was 5 feet away. They see a microphone, and think it works like a camera. If it’s not pointed AT them, it’s not recording them – WRONG!!! There are 3 major misconceptions:

1- “Don’t you have some kind of directional mic?”

Yes but it doesn’t work like that. It’s not a camera. With a camera if its out of frame it doesn’t exist. Not so with a mic – we can minimize it to a degree but for full range recording there’s only so much off-axis rejection we can do (or want to do).

2 – “Can’t you “just get that out” of the recording?”

No, you’ve watched “The Fugitive” too many times. We don’t move a slider & magically remove unwanted sounds without screwing something else up. Sometimes we can help it out, but it’s by far best to have an uninterrupted recording. (Izotope RX’s Spectral Repair is the closest thing to this kind of magic I’ve seen – but I still don’t want to base a field recording session on being able to “get things out”)

3 – “Oh yeah it’s quiet there.”

Really? From the other tip at the races, I also tried to record at this “quiet” location. There were lots of people at this spot, and there was no getting away from them. In many ways this was worse than being in the pit, except the planes did pass closer there. At one point I wrote down everything I was hearing.

• Walla out the wazzoo – where there are people, there is yapping
• Car radio blasting AC/DC’s “TNT”
• 4 guys revving their motocross bikes
• Tents flapping (it was pretty windy too)
• A tied up trash bag flapping like crazy in the wind.
• About 4 cars right around me, all playing the announcer from the races over the radio – it was worse then the PA inside the event.
• Radio controlled helicopter
• During a race, a guy standing in front of me kept taking his camera in & out of a velcro holster.
• Same guy – twisting his foot on bone-dry crackling tumbleweed brush as he tried to get a better shot.
• Another race – a guy next to me decided it was time to clean his cereal bowl, during the ENTIRE race. Clank clank clank, then rinse the bowl with a crackly plastic water jug, pour water out on the ground.

Not one of these people did I perceive as intentionally ruining my takes. They simply have no clue what it takes for us to get a good recording. They see the mic pointed at the planes, and to them, that’s all I’m getting. Not one of them asked or offered to be quiet. However one guy that walked in front of me ducked under the mic, as if it were a video camera and couldn’t “hear” him if he went under it.

Naturally this was a shot in the dark that had little chance of turning out great recordings, but I did wind up with some good pieces in spite of all this. Actually it was another example where finding the source without headphones was IMO hands down better. The speed & distance of the planes were all different, and the sound lagged behind the planes by varying distances every time. The only way to find the source was to use both ears in the space (you need both to locate), and point a mic there. If someone were recording & using headphones, in all likelihood they’d be pointing the mic at the plane, but that’s not where the sound was coming from at all.

OK there’s one more major misconception – 5 tweeters (essentially) & a subwoofer do NOT make a good sound system.

I need my mids man!!! I need my mids!