For Crysis 3, the team at Crytek took a different approach from their previous games for capturing performances.
For this production, Side was commissioned to cast actors who could provide the likeness, movement and voice for the lead characters in Crytek’s sci-fi shooter, and these were captured, together, during performance capture shoots.
An important part of this process is ensuring a high percentage of the dialogue recorded during shoots can be used in the final game. However, recording final audio on the motion capture soundstage brings with it some challenges.
Firstly, a soundstage without set or audience can sound quite ‘live’, which is not ideal for recording game dialogue. And secondly, motion capture suits, head-cams, and props can create many unwanted noises. But compared to recording actors one by one in the booth, recording the vocal performances while the cast is performing physically together in the same space adds greater realism to the final scene.
On such projects, it’s vital the sound team are involved at an early stage. At Side, if we haven’t used the soundstage before, we carry out a pre-shoot visit, taking test recordings to evaluate acoustics. If the soundstage is okay, we may still use heavy drapes or matting to improve sound further. And we’ll liaise with the motion capture team to ensure they install their servers in quiet boxes or off set.
With Crysis 3, our location sound team consisted of a mixer and boom op. The primary audio source was a lavalier radio microphone. The actors wear these on their foreheads if the setup allows, or the arm of a motion capture helmet if not.
And having the boom audio offers a different perspective as well as an alternative source if the lavalier audio is compromised. With the boom op in the volume together with the actors, they can make dynamic adjustments to mic positioning to improve quality, or even create desired effects.
The mixer records actors to separate tracks on Pro Tools, synchronised to time code from the motion capture system, and backed up to a separate hard drive.
From there, we output a number of mixes, including one of all actors for the director, writer and producer; boom audio for the boom op; and talkback for those who need it. Plus, if there’s facial capture, while the head-cam footage is captured, we send the individual actor’s audio to one channel and the mix of all actors to another. We can also send audio to a monitor wedge for footage playback or just to remind an actor of a read.
IN THE MIX
It’s important to note that although the mixer may want a retake for sound, with so many different parties involved, if the take was perfect for everyone else, then the line may be marked for ADR later in the studio. But in these cases, recording to Pro Tools has advantages over recording to a simple location recorder.
If the mixer does have a problem with a line, it’s easy to grab another take on the soundstage by looping the line while the actor wears headphones, reducing the requirement for post-shoot ADR.
Naturally, there was some ADR to do on Crysis 3. For these recordings we prepare a Pro Tools session using the reference video footage of the actors, and then use a dubbing system that displays the script on the actor’s screen, along with the reference video and a ‘wipe’ to help the actor with the timing of the line.
And we try to get the best match to the audio from the soundstage by using the same lavalier mics in the ADR studio. If ADR is being done due to a line change, we’ll also capture new headcam footage in the studio, which will replace the soundstage footage.
For Crysis 3, over 90 per cent of the dialogue recorded during performance capture was used in the final game. When you consider many movies need up to 75 per cent of their dialogue ADR’d; that’s not a bad result.
To see other articles in our Audio Special series, visit our archive