Pro Techniques from Clint Bajakian

By Randy Alberts

Clint Bajakian

The idea of a non-profit interactive audio guild that educates, protects, and promotes its members had never crossed the average game geek's mind — that is until Tommy Tallarico, George Sanger (a.k.a. "The Fatman"), Clint Bajakian, and other game audio pioneers gathered for the 2001 Project Bar-B-Q in Texas. There, Tallarico delivered a speech proposing that a new organization be formed dedicated to the betterment of interactive audio — and G.A.N.G. (Game Audio Network Guild) was born. Visit www.audiogang.org to learn more about this forward-thinking game developers' organization.

A non-profit organization centralizing a community of interactive media audio developers, G.A.N.G. has a mission much the same as DigiZine's: to provide information, instruction, resources, and enlightenment not only to its members but to content providers and listeners throughout the world. That mission is both working and at work for Tallarico, Bajakian, and Jack Wall as G.A.N.G.'s President, VP, and Senior Director, respectively. Bajakian and Julian Kwasneski, his partner at The Bay Area Sound Department, were more than gracious enough in doing the same for DigiZine readers as G.A.N.G. does for its interactive game audio members: Educate.

A Jedi Game Developer
Bajakian is an 11-year game audio and music veteran who cut his teeth developing enough Star Wars and interactive audio games at LucasArts Entertainment (Outlaws, Full Throttle, Star Wars: Tie Fighter) to automatically qualify for Jedi status. In 2002 he co-founded The Bay Area Sound Department in San Rafael, California, and along with sound designer Kwasneski has been way busy on many XBox (Star Wars: Knights of the Old Republic), Playstation (The Simpsons: Wrestling), and numerous other sound design, music editing, voice post, and original music projects ever since. It takes all of 30 seconds on the G.A.N.G. and The Bay Area Sound Department websites to get that these folks know what they're talking about.

The following is more of a game audio developer's Pro Tools seminar than a group of Pro Techniques. Rather than follow its standard Pro Techniques format, DigiZine gladly handed over the laser light pen and PowerPoint clicker to Bajakian and Kwasneski. Somewhere in between working on a new XBox game, recording a live orchestra for the new Indiana Jones: The Emperor's Tomb game for LucasArts, and organizing an industry panel at the Game Developers Conference, both Clint and Julian somehow found time to wax poetic about game audio workflow and Pro Tools' pivotal role in that process.

Linear and Interactive Audio 101
By Clint Bajakian and Julian Kwasneski

Strip Silence
A feature we use regularly in Pro Tools is Strip Silence, which allows us to change various threshold settings and automatically edit long files into smaller, game-ready files. For example, let's say we receive a 30-minute DAT tape containing the best takes of a given character's voice lines. First, we'd import the tape into Pro Tools and then select Strip Silence (Command/Control + U) to end up with individual regions for each line this way. Here's how: Once the DAT tape is imported into Pro Tools, we simply bring up the Strip Silence window and tweak the following parameters:

The Strip Threshold slider adjusts the audio level at which point a region marker gets set. We typically will set this to a very low setting, usually -48 dB; next, we set the Minimum Strip Duration, which specifies the length of the shortest region to be defined. This is great for guarding against different words of the same sentence getting split into different regions. Since the Strip Silence command shows the region placement graphically as you move the slider, you can view the changes made in real time. Of course, nothing is destructive so the process can be done and undone over and over again until you get the desired results.

Strip Silence also includes sliders for a Region Start and End Pad so we can make slight global adjustments to every region. Let's say you want to have a little handle on each end of the file or you want to make sure that the last syllable is not abruptly cut off. The Region Start and End Pads will help you accomplish this.

One final feature of the Strip Silence tool is Rename. Clicking on this button will allow you to instantly rename every region with a given word (or words) followed by unique numerical ID numbers. For example, you could rename the regions "Joe" followed by the given line numbers with any amount of padded zeros you require. So the resulting files would be something like "Joe01", "Joe02", and so forth.

To create the individual regions, simply click the Strip button and all the edits will be created automatically. Finally, with all your new regions selected, go to the Region Bin menu and select Export Selected As Files. Next, select the file type you need, choose a destination, and click the Export button. You now have multiple files, all named correctly and neatly edited.

Creating Loops for Multimedia
If you've ever had to make short, yet believable, loops for games, you undoubtedly will benefit from using Pro Tools to create them. Simply take the region you wish to loop, place the cursor wherever you'd like within the region and click Command/Control + E to cut the file at that location and split it into two regions.

Next, while in Shuffle mode, use the Hand tool to grab the region on the right and move it to the left so it becomes the first region. Now select both regions on either side of the new cut and hit Command/Control + F to place a crossfade between the two regions.

stacked on top of....

Selecting both regions and allowing Pro Tools to play in loop mode (Operations>Loop Playback) should produce a smooth loop. It takes time to get some files to loop in a believable way, but with a little practice you can create extremely smooth loops using this method.

Grouping and Bussing Tracks for a "Do-It-All" Session
One of the greatest things about Pro Tools is the TDM bussing and plug-in processing. Grouping and bussing is a major advantage to achieving high-quality results when creating anything from a single sound effect to a music or dialog track.

Take a sound effect, for example, a gunshot. It may be OK for some folks to simply grab a gunshot FX out of an off-the-shelf sound effects library, but to raise the bar of quality one should really do some creative design to best serve the product's needs. Often, layering different sounds is the foundation of a great effect. In this case, I'd start with a library of originally recorded gunshots and then mix in larger-than-life elements, such as a cannon or automobile crash on the attack that quickly fades out. Another important aspect of a gunshot in real life that would be great to achieve in a sound effect is that "thud-in-the-chest" transient. I often create that by layering in a kick drum sample that has been heavily boosted in the low-end.

So far, I've only discussed layering. The real time-saver comes in when you buss all these different elements to Aux Inputs or a Master Fader via TDM. Let's say we route everything to a Master Fader and insert EQ, compression, and gain maximization plug-ins on the Master. Now, we tweak the settings of the send levels of the source tracks, including detailed automation of both volume, pan, and any effects plug-ins we may have on that track. Then, we tune the parameters of the master fader plug-ins, keeping a close eye on our master output level. When we get the effect just right, it's done. No need later to master in a two-track editor or to bring it back into Pro Tools for mastering — the FX is done and ready for delivery, for editing in a session to picture, or for converting to a different format.

Now, let's look at a multitrack session with dialog, music, and effects elements. You can have literally dozens of source tracks, all of which contain unedited, pre-mastered elements. Each character with dialog can exist on its own track this way. The effects may, and often should, be maintained as discreet events on individual tracks. For example, a low-end component that sounds awesome alone in the effect may now be interfering with the music or dialog excessively. To have a handle on that one component may well mean the difference between a truly professional and an amateur mix. Music, incidentally, I prefer to have mixed to a stereo stem prior to bringing it into the master mix session, possibly even having mastered it as a stand-alone music file as well. Mastering can well be done in this multitrack mastering session on the fly, as we'll see in a moment.

Groups and TDM Bussing to Group Busses
So, while it's wonderful we can have all these tracks mixing and playing simultaneously and to be able to automate each track's volume, pan, and plug-in effects parameters, it's even better when you introduce Groups and TDM bussing to group busses. These in turn all flow into a single stereo Master Fader. This gradual collapsing of the audio information down through a detailed series of automation allows two important capabilities: One, a Bounce to Disk that is finished with mastered files ready for delivery; and two, a compelling professional mix that promotes your abilities to your client and ensures future work!

A Common Game Audio Development Scenario
Dialog — Let's say there are four characters, one of which is sibilant, another of which is bass heavy, the third in drastic need of compression, and the fourth just perfect. Simply insert a De-esser plug-in on the first track, an EQ on the second, a compressor on the third and no plug-in on the fourth. Then group the four tracks, naming the groups "Dialog" or something simple.

Effects — Now, let's say we have several effects tracks with separate elements broken out to different tracks as much as possible for maximum control. We work each track and mix the elements via volume, pan, and plug-in automation. Now, we create a group called "Effects" or some such thing.

Music — Ideally, we have a single stereo track, though, for more control, it is certainly ideal to have broken the music into logical stems — something like Melodic, Accompanimental, and Rhythmic elements, for example, or elements that are arranged according to frequency content. Let's say we had already created a stereo music mix in a previous Pro Tools session and it's mixed but not mastered here in the mix session. Again, easy: Just insert effect plug-ins that are typical to mastering, such as EQ, compression, and reverb directly on the music track. Do all the wild and crazy automation of that music track you need to now "duck" it for dialog, boost it at crescendos, boost the bass for a big hit, etc. If you're thorough, the volume automation view should look like a stock chart!

Mix — Now, create a Master Fader. All the individual tracks will now be routed through that fader to the speakers, or to disk upon bounce. The Groups we set up have no effect — yet. Create three stereo Aux Inputs, naming them "Dialog," "Effect," and "Music." Set the inputs of each Aux Input to its own Buss stereo pair, such as Buss 1-2, Buss 3-4, Buss 5-6; you can name these Busses in the I/O Setup window. Set the outputs of the source tracks of each Group to each Group's respective Aux Input. A shortcut to doing this is to isolate all the tracks of a particular group in the Edit Window, then assign the output while holding down the Option/Alt key, and all viewed tracks will respond to the same setting. Once this output assignment is done, set the inputs of each Aux Input accordingly.

Now the session is almost done. You'll find that when you mute any of the three Aux Inputs that that entire group will be muted. Another way to think of it is this: Say you're not hearing any of the source tracks playing through the Master Fader but you are hearing the Aux Inputs. Aux Inputs that in turn flow to a single stereo output would be pointless without the following incredibly powerful advantage.

Insert your mastering plug-ins on each group, such as EQ, compression, gain maximization, reverb, etc. Who wouldn't be interested in having a unified track for dialog with mastering plug-ins on it in the same session that each individual line was initially dragged in to its own track! And don't forget to automate the effects parameters on these Aux Inputs. If some of your dialog happens in a cave, automate the reverb mix level on the dialog Aux Input and just have it up for the cave and down for the outdoor part of the scene. Similarly, automate the [Digidesign] Maxim or [Waves] L1 level on the effects when you need that extra push over the cliff or boost in that explosion to make it soar over the mix! This method also saves on DSP power. You might have been able to L1 the explosion track itself, but that isn't so easy when that explosion is broken into several tracks.

You could also process the explosion via AudioSuite but that wouldn't be as flexible in the final mix pass. Automating plug-ins on the Aux Input groups gives that essential second tier of control over the automation work you did on all the separate tracks. But that's not all.

Mastering — The cherry on top is to insert mastering plug-ins on the Master Fader. Set them with an ear for perfection and you're ready to bounce out a final, mastered file, ready for delivery with no need to fuss over it later. Oh yeah, one last thing, and really important: The meters on your hardware interface are your best friend. Really. Graphic screen redraws can be sluggish, especially when working with a high CPU load, so on-screen metering can get a bit inaccurate compared with the lights on the hardware interface. I position my interface prominently so I can constantly look up at it while working. When you get those meters to behave exactly how you want them to, you're done, and ready to bounce, deliver, and take the rest of the day off!

The Difference Between AudioSuite and TDM Plug-In Processing
Quite simply, the difference is in the flexibility you want to reserve for later in the session work or specifically want to deprive yourself of in the interest of simplicity down the road.

Let's say a character is too bass-heavy and you're tempted to select all the regions of that character and attenuate lows in an AudioSuite plug-in. That may be all well and good but you really might be better off inserting an EQ TDM plug-in on the track to reserve the flexibility to change your mind later about the amount of attenuation. Even better yet, automate the EQ so that the EQ process can be exerted in different amounts to each line (or syllable) to taste.

On the other hand, say you have some dialog regions that are just plain too quiet. Rather than doing a whole bunch of volume automation on the track, it's much easier to simply boost the regions with a gain change in AudioSuite and get back to work mixing. It's always a subjective call as to whether to process a region, a set of regions, or an entire track with AudioSuite or TDM plug-ins. The main criteria has to do with the ability to change your mind later easily and the ability to automate the effect over the course of the track. If you're sure you won't need to automate or change your mind about the processing of something, go ahead and use AudioSuite. I'd rather take up a little more disk space and save my DSP horsepower for bigger and better things.

Multimedia: Creatively Processing Hundreds of Lines
People may not realize it, but AudioSuite in Pro Tools is one of the best batch effects processors on the market. The Region List contains the key to some very powerful batch processing. Consider this: the client needs 437 evil sorcerer lines pitch-shifted and flanged by 10 a.m. There are two great ways to do it:

AudioSuite method — Import your files into Pro Tools, select one in the region list, and bring up a pitch shift plug-in. Tweak the settings until you're happy, then save the settings. Next, select all 437 regions in the Region List and hit the Process button in the plug-in, making sure to select Region List and Process as Separate Files. It will then write 437 files to disk with an appendage on the file name indicating the type of process you exerted, in this case, a pitch shift. Those new files will also appear in the region list as new regions. Now, select one of the newly pitch-shifted regions and bring up a flange plug-in. Get the settings right and select all 437 pitch-shifted files and hit Process. You've now got 437 pitch-shifted and flanged files on disk.

TDM method — Import the files into Pro Tools and array them in sequence on one track. Insert a pitch shift plug-in on the track followed by a flanger. Tweak until everything is just right and possibly insert a compressor or gain maximizer to have the ability to bounce out finished mastered files ready for delivery. Once it's all sounding good, bounce the entire track to disk and import the resulting long file into a Pro Tools track. Use Strip Silence to break the long region into the original separate files, and finally, batch name the regions. Remember, the new regions will appear in the region list but are not yet files on disk. In the region list, they can be multi-selected and bounced to disk as a file set of 437 files using the Export Selected as Files command in the region list menu.

If you're processing many lines that require a modulating component such as fluctuating radio static or background noise, then strap the radio static or background noise on a track or two below the dialog track on which all the dialog regions are arrayed. Do random automation of the volume and effects parameters on those noise tracks under each voice region. Then, select each dialog region and Bounce to Disk. To maintain the exact same file name, double-click on the dialog region and copy the region (original file name) into the text clipboard, then hit cancel. Then select Bounce to Disk and paste the file name into the prompt. This is a great way to achieve really cool lines that make use of sound or noise elements in a way that's different for each line. From a sound design perspective, I like short wave radio tuning noise behind lines that are supposed to be suffering from radio static, especially in the sci-fi genre! An even cooler thing is to randomly automate a distortion plug-in such as Recti-Fi or Sci-Fi in combination with other static or noise elements.