Worldizing is not a new concept. The term was first coined by Walter Murch as the process of taking recorded and edited sound or music into the field, playing it back then recording the result. The process was developed because of the limitations of the technology back in the day. Nowadays we have high quality convolution reverbs and processors such as Altiverb, TL Space, Waves IR, Space Designer, Speakerphone etc etc which can do a pretty damn good job of replicating a space or piece of equipment. However this stuff isn't perfect and more importantly (in my case) isn't cheap. So for me wordizing is still a technique I turn to when I want a particular sound or effect that I can't quite get with my meagre digital means.
Nice example from Lord Of The Rings: The Fellowship Of The Ring;
So where does this all fit into The Game Audio Tutorial? Well for one of the exercises in the book we needed some dialogue to guide the player around the map, this bit of dialogue was going to be delivered by a small radio. Now at this point I could have started adding eq and distortion to the dialogue lines trying to approximate the sound of a small speaker but I already had something a little different in mind.
A couple of years ago I bought one of these (see pic) its a Smokey Amp by Zinky Electronics it has a really thin raspy distorted tone which would be great for a walkie-talkie esque sound.
So finally processing steps;
Original dialogue recorded @ 96k 24bit into Digidesign 003 using a SE Electronics 2200a
For obvious reasons the Smokie Amp doesn't really like line level signals and seemed to produce the best results when run with audio from my iPhone
Worldized dialogue recorded back into Digidesign 003 @ 96k 24bit using a Rode NTG-3
The robot in GAT is a small homage to the famous singing robot played by Ellen McLain in Portal; GlaDos. This post will take you through the process of creating the robot voice in the Game Audio Tutorial using Melodyne.
(Warning Contains Spoiler)
The dialogue line has been performed in a very monotone staccato manner as to try replicate a robot type voice. It was recorded with a Neumann u87 into an Audient ASP8024 and MOTU 24 I/O.
Charlotte Fotheringham - robot voice actor
The first stage of creating the robot voice is to correct the pitch in Melodyne this snaps the word or part of a word to the nearest note. Listen to the corrected sample
The next stage is to constrain the pitch modulation of the dialogue line this stops the pitch of the dialogue modulating giving it a very synthesised feel.
The penultimate stage is to adjust the dialogue line's formants up two tones. Formants are groups of harmonics present in the dialogue by shifting these up the fundamental pitch remains the same but the harmonics become stretched out, again making the line sound more synthesised.
The last process is where I wanted to differentiate the character from GlaDos slightly. I wanted the character to sound slightly comedic rather than menacing so rather than leaving the pitch constant I pitched individual words up and down by a couple of semi-tones to give the character a jerky musical quality.
The owls & night birds in The Game Audio Tutorial have a bit of a strange back story to them and id thought i'd make this post about them and the problems they incurred.
As you can glean from the title of the post I needed to record some owl and night bird vocals for The Game Audio Tutorial and for various reasons to do with the book I couldn't use any library recordings.
Phase1: Record birds in their natural habitat
So I went on a local nature website to try find out where these critters live. After picking a couple of locations that I could visit all in one evening I set out with my usual combo of an FR-2 and a Rode NTG-3 to try record some of these birds.
Unfortunately none of the birds seemed to be in, or if they were in they definitely didn't come out to play. All I managed to record whilst sat in a scary-ass wood on my own in the dark in the middle of the Lake District was the rustle of trees and possibly the quiet approach of the local axe murderer. (although that could have been a rabbit or a fox for all I know it could have been an elephant)
Phase 2: Record reserve birds
Luckily there is a falconry centre close to me which houses a couple of owls and other birds of prey which would be great to record. So I gave them a ring to see if they'd be interested in taking pity on a lowly penniless sound designer and let me come and record some of their birds. They agreed to letting me record the birds but unfortunately they wanted a quite substantial fee for doing this, which is fair enough really as I'd be taking up an afternoon of the centre's time. However the budget I have for the Game Audio Tutorial doesn't quite stretch to this (it's zero, nill, nadda,) so I had to go back to the drawing board. Again.
(on a side note when I'm not so penniless I'm going to go back and take them up on this offer)
Phase 3: Desperation
The last idea I had (apart from going out and sitting in a forest all evening again)was to get some recordings of night birds like owls and the such and to try to mimic them with my voice then pitch the recording to a frequency that sounds realistic enough to be an owl. So I recorded myself mimicking owl noises and pitching them to create bird like sounds.
It sorta works. You could almost say that the human and slightly unrealistic quality to the vocalisations gives them an eerie and unworldly tone. If I claimed that this was intentional I'd be lying.
And this neatly brings me on to the point or moral or theme or end bit (running out if ideas now) of this post.
The human voice is pretty versatile, it can be used to produce a range of vocalisations for all sorts of applications. For instance if I'm trying to describe a sound to someone and struggling to do it often i'l vocalise it to try and get my point across. Or if I'm struggling to create or record a sound I'll have a go at vocalising it to see if that works, I've lost count of the times that someone has come across me making noises to myself then wandered off again totally bemused. Some may argue that a sound designer's voice is one of their best tools and I'd agree. So now I've made this bold claim here's a couple of videos and examples of the human voice being used in real-life sound design by real-life sound designers!
Dragon Age Origins: 0:50
Jordan Ivey uses the sound of a human imitating a cat in the Deepstalker creature.
Star Wars: R2D2
Star Wars contains tons of vocal performances by both Ben Burt personally and other sound designers arguably one of the most iconic being the voice of R2D2.
This excerpt was taken from the book The Sound Of Star Wars it is excellent and I suggest that everyone vaguely interested in sound to go buy it.
Over a period of months, R2's voice became a fifty-fifty meticulous blend of electronic and human sound. Eventually Burtt built a circuit using the ARP 2600 that enabled him to play notes on a synthesiser and at the same time record human sounds into a microphone; For example, if Burtt raised his voice in pitch, the electronic sound would shape itself to conform. Burtt would make the sounds as if in slow motion and then speed up the result, which created the rapid high-pitched sound of the droid's speech.
"Artoo hs a scream, which is just me screaming" say Burtt. "I did the scream up at Park Way in the basement where I worked. I remember I was lying on the floor under the workbench table because it was the quietest place in the room; it would insulate me somewhat, because it had a filling cabinet on either side of it. Later, I sped up my scream a little bit, so it's higher in pitch. But it's funny-I've tried to repeat that scream over the years, and I've never been able to hit that note again without coughing or something."
To finish I'll leave you with this as a rather extreme example of what can be done. A complete replacement of all the Half-Life 2 audio with vocal samples (I didn't create this btw). Enjoy
Since it's nearly Halloween I've decided that this latest post is going to be about some spooky metal sounds in the gat sound library so here it is;
The metal murmurs featured in the gat sound library a classic example of spooky metal sound effects these are designed to be used as one shot stinger type sounds randomly throughout a spooky game level. This effect can be created relatively easily. The base sound for the effect is a bowed 14" cymbal, this doesn't have to be a cymbal but it is a good example.
The cymbal I have used is actually broken, it has a 12cm crack in the metal. This is good for two reasons. Firstly it gives it a slightly more abrasive tone when bowed, it also causes changes in the tone and resonant frequencies of the cymbal depending on the distance between the bow and the crack. The second reason is that it makes it the cymbal cheap, sites like ebay have a whole host of broken or damaged stuff like this that can be bought relatively cheaply.
Bowing the cymbal requires two things a bow and rosin. In this case I have used a cello bow as they are a bit tougher than a violin bow. Rosin is a solid resin produced from pine trees that is rubbed onto the bow hairs to create friction between the bow and the surface you are trying to bow.
The cymbal was recorded at 24bit 96k using a shotgun microphone and a contact microphone.
The shotgun microphone picks up the high frequencies nicely but lacks bottom end. The contact microphone is pretty much the opposite, it picks up the lower frequencies better than high frequencies. When mixed together this produces a much thicker sound.
To give the murmur some space a reverb is added, in the example a large church convolution reverb is used.
The next step is to add an echo effect with a large intensity this gives the murmur the "larger than life quality"
The murmur is then pitched down 25 semitones, this both not only lowers the pitch of the murmur but also stretches the length of the murmur greatly giving it an eerie drone like quality.
Today i spent some time recording various shells and bullet casings impacting a rock for the casing physics impacts found in the game i am currently working on.
Casings and bits right to left ;
12 Guage Shotgun Shell, Big Fat Brass Slide, Really Thin Brass Slide, WW2 German Mauser Replica Casing, WW1 German Mauser Casing (Actually fired in anger and picked up from a feild in Verdun, Belgium), WW1 German Mauser Bullet (agian from Verdun), Bullet from replica casing and a WW1 allied bullet collected from the Somme
The impacts on the rock from the various casings was recorded using a RODE NTG-3 and a sE Electronics 2200a.
So finally processing steps;
