Okay, so this is way too early to be asking this, but just to get it over with...

Is it possible to manipulate game code so that one can replace existing voice clips with their own stored clips?  I know that this has officially been done with games like Time Gal, Grandia III, and Dark Cloud 2.  However, I'm not sure whether dubbing takes place after the initial game code is converted into binary, or a copy of the initial game code is stored elsewhere for dubbing.

If voice-hacking is possible, then Game Tengoku and Makeruna Makendou Z (which both have lots of voice acting) would be perfect games to dub.  I'd much prefer to hear the characters speak English and have different English-speaking voices from their original Japanese-speaking voices.*  (To be honest, I find Japanese female voices to be rather annoying because of their near-lack of variety, compared to American female voices.)

(*Actually, this wouldn't be too out of character.  Most of the time, when an anime series or game gets translated, the voice actors of the dubbing language sound nothing like those of the original voice actors.  I wouldn't even have to cite any examples; there are very, very few exceptions.)

Thanks for answering me.

So, do I need to collect voice samples with a voice recorder, store them on my hard drive, convert them into binary, and then add them into the existing binary code for whatever game I wish to dub?  Or is the process different from what I'm thinking of?

Also, as long as we're on the subject, is it possible to add voice-acting into games that have little or no VA?

There are two separate issues. One is the format the audio data must be in and the second is the game code to play it. In some cases such as a PS2 game with streaming audio, you may need to just replace a few audio or video files on the disc and that's that. Or, it could be much more complicated involving compression, proprietary formats, or other complications. When it comes to old systems that can't handle streaming uncompressed data, you'll need to break apart your sound samples into a bunch of mini samples in a format suitable for the system you're working with, and then modify the game code to play them which could range from somewhat easy to a nightmare.

As for implementing voice acting into a game that had none. It's possible yes, but would require a great deal of work. You'd have to modify the game engine and all the event data to use a newly hacked system to know when to play audio and what samples to play.  I'd imagine this would be a heck of a hacking job depending on the game. If you were lucky you could get away with adding a few new controls to the event or text engine and then writing some code to be able to find the sample needed at a given time. Not easy though.