“We made the CG based off of Otsuka-san and Ginga-san's voices.”
Kojima on the technical archaeology of game dubbing: he says that in the days of Snatcher, Policenauts and Metal Gear voice actors had to dub with nothing in front of them because there were no visuals, and that they made the CG based on Otsuka‑san and Ginga‑san’s voices — first the voice, then the image. He contrasts that with the present, where performance capture records acting, voice and expression all at once so the dubbing process becomes the same as for foreign films. He believes that change reduces the space for what voice actors can contribute and is unfortunate, because previously actors were allowed to go wild; at least now they can speak over one another.
“in the Snatcher, Policenauts and Metal Gear days, the voice actors had to dub with nothing in front of them. […] We didn't have any visuals. […] But now we have performance captures where we can shoot the actors' performance, voice and expression all at once. So the dubbing process becomes the same as what you do for foreign films. I think that takes away from all the great things voice actors can bring to the table, so it's kind of unfortunate.”