One of the interns asked me about my response to question 30 in the application, which was: Watch the (silent) video below and briefly explain how you would approach creating the sound effects for this shot. What elements would you record?
In terms of the intern application & a practical exercise this one was a real world test since, as one of the applicants guessed, this video is from World’s Fastest Indian and it was a sequence I had to cut sound for. So I had a clear memory of how I approached it & which elements worked. Whenever we’re premixing or mixing & the director is present I always hold my breath as any subjective moments play down, waiting for a reaction & hoping for a positive one. And although this moment is reality based, it is slowed down… but thankfully first play down with the director & he loved it. So what was my approach? And don’t get me wrong; that doesn’t mean its the only approach, but its one that worked for me…
First of all, and I appreciated the few people who raised this aspect, I would spot the scene with the director. Of course this shot is part of the film so it would happen in that context, but getting some idea of the general feel (do you want this to be played quite real? stylised? emphasise the slow mo aspects? etc) as well as specific (do you want the wheel to sound broken? intact?)
Secondly, I also appreciate the people who commented about the desert ambiences, the off screen action, point of view and the back story (ie where did the wheel come from? what are those sparks flying? etc) and they are valid considerations but I was primarily asking about the sound effects.
So my usual approach with a scene is to start off by dropping markers in ProTools on each hit point; the first one is about 2 frames into the shot as the wheel bounces for the first time and then the two other bounces. I often use temporary markers to build a non-linear grid based on the sync of the action in the scene – it means I can identify sync for each element and then focus on the sounds for each event as a seperate issue.
1. Ok so first on my list of FX would be the wheel bounces. If you have ever taken a wheel off a bicycle & dropped it from waist height you will know it has a definite characteristic too it, and theres no need to fake this, its relatively easy to record. I would record a variety of bicycle wheel drops… but the wheel looks heavier than a push bike, more like a motorcycle or trailer (it is in fact from a trailer) so I would also try to locate a heavier spoked wheel & record drops with it. Now some people would fuss around taking it out to try & find a similar surface to drop it on, but that wouldn’t be my main concern when first recording the wheel drops, for two reasons: firstly I intend to create the ‘wheel hit surface’ sounds as seperate elements, so I can control their balance in the mix. Secondly, the primary character I want to capture of the wheel drop is its weight & resonance, so I would first try dropping it on concrete, then tarseal then gravel & see what provides the best character ie what makes the entire wheel & spokes resonate the most.
2. Next I would record a good solid thumpy hit. Maybe muffle the spokes & again slam the tyre onto concrete, but as opposed to part 1 where I was after the entire wheels resonance this time all I want is a short bassy thump, almost like a body hit.
3. Next is the wheel impacts surface sound. Notice how on the first two hits down dust rises, so that says its not just bouncing on gravel – its actually making gravel fly. So I would take my bike wheel & motorbike wheel & I would start them spinning & I would slowly lower them onto gravel so that they not only hit the gravel but the stones fly. I would record maybe a dozen of these, varying the intensity since we want each of the three hits to be unique. i would also do some hard fast abrupt ones, almost like short gravel skids.
4. Since I’m playing with the gravel I would record another element which is a VERY archetypal bike wheel sound; throwing stones at the spokes & making them ping. I’d record individual ones but also a handful. What use is this? Well it sonically reminds us of movement; the shot may be slowed down but that means we have more time to hear the details. Secondly it reminds us of what the object is generating the sounds & if a stone was to flick up & be hit by a spoke, that spoke would be fair motoring in speed!
5. Now the second bounce looks like it could have hit a small bit of bush, and in the third bounce it fair crashes through bush. This is obviously the desert so we need to record some dry brittle branches. I would record some individual branch snaps & breaks, but the most interesting sound would be the wheel passing through the bush. I would try this both ways ie try swiping the wheel on the bush at various speeds, but I would also try grabbing a branch & wiping it past a stationery wheel. As with most recordings it pays to vary your peformance from ‘very subtle’ to ‘completely over the top’ because you just never know how it will play in context.
6. next I would use the same approach on the spokes as I did with the stones, except using bits of the bush. So I would get the wheel spinning as fast as possible & I would stick small branches into the spokes, so they ‘ping’ the spokes & reinforce the feeling of speed. I would record quite a range of intensities of this; for example as the wheel rises up from the third bounce it is fair crashing through the bush, so that requires more complex dense sounds. But you can also see in the last 11 frames of the shot a small piece of debris goes flying off to the right, which requires a singular branch snap & spoke ping…
7. Next I would think about the wheel travelling – what sounds could it make while its flying through the air & not touching anything? I’d try recording air rushing past the spokes ie spinning the wheel as fast as possible & then moving the whole wheel past a stationery microphone. I would also try lightly rubbing small objects (eg a piece of cardboard or a twig) against the rubber on the tyre & also in the spokes… I’m trying to capture movement so I would try to vary the performance of this, remembering I only need little pieces up to 2 seconds in length between each bounce.
8. Lastly for the recording I would think about what motivated the wheel to bounce so high on its third hit. it must be reacting to having hit something; a rock or a sudden rise in the ground. So i might try recording some specific elements to help it eg 2 rocks bumping together and maybe a dirt thump.
Ok so heres my FX Record List:
1. resonant wheel hits
2. thuddy wheel hits
3. spinning wheel hits on gravel
4. stones, gravel hit spinning spokes
5. bush branches hits & wipes, swooshes
6. bush branch hits spinning spokes
7. perform dopplers of wheel spinning, try with friction
8. rock hits and dirt thumps
Now to construct it!
I would work across at least two and maybe three predubs for this, so that in the final mix we don’t end up with all the elements locked together. I would make a set of tracks for the percussive hits, another set for the movement & a third set for subjective elements.
So on the first set of tracks I would sync up the resonant wheel hits, the wheel thuds & the spinning wheel hits on gravel. The main thing is variety – we dont want each hit to sound the same; to my eye the first & the third hits seem the heaviest. The first one throws the most gravel & the last one causes the direction change so they should be heavier sounding than the middle hit.
On the second set of tracks I would work with the movement sounds, so on the first wheel bounce I would use a bit of the gravel spoke pings, not much on the second & lots on the start of the third, but the third hit is mostly about the brach/wheel interaction so I would layer the branch wipes & branch spoke pings…
Now the wheel flying through the air sounds I would probably split from these other two predubs. Maybe the flying elements wont be needed much at all – if we are keeping it real then these sounds would be very subtle, after all your ear would need to be fairly close to hear them in reality…
In terms of processing sounds there are some I would likely try pitch shifting – it might worth well to vary the pitch between each of the hits as an overall approach, but I could also imagine pitching the bicycle wheel resonant hits down a bit too add more weight. I would also imagine trying to doppler some of the elements eg the wheel flying through the air elements & also maybe some short fast dopplers on friction elements.
Other ideas I would explore would be to try a short swishy woosh leading into each bounce, especially the first one. I would also try some higher pitched sounds for the wheel flying through the air, almost evocative of the propellor on a plane….
I tend to use a lot of volume automation when editing, so that when you present the material to the premixes, if the mixer puts his faders flat he gets what I consider to be my balanced version of the sound effects… I dont tend to print EQs unless its for a fault (eg a mic bump) simply because working with mid or near field monitors is not the same as working on a film dubbing stage & I dont want to discover that my sounds seem thin when heard at the predub because I’ve EQ’d them. As this is exterior I wouldnt mess around with printing reverbs or slap delays or anything; again a mix dub stage has better sounding reverbs than I can ever afford AND its part of a mixers aesthetics & skills to dial in appropriate verbs (unless its a temp mix & there may not be time)
I guess thats it… I’ll update this if I think of other aspects & when I get some time I’ll re-read all the applicants answers & see what other ideas were suggested. Some elements are pretty obvious, but thinking about what the unique sounds that can be added to provide character is what often makes the difference between a plain, generic sound effect and an interesting one. But the character needs to feel true & not an affectation. In this case the elements I used which added character, were motivated & reinforced the action were the spoke pings and it is one of the joys of working on sound post for films vs TV. With TV its all about the deadline – its not how good are you, but how good can you be in the time available. With film we get enough time to try the obvious approaches, live with it for a while, develop it further & be objective about it. Sometimes that means going full circle & arriving back at the first sounds cut for a scene, but its an evolution.