Aaaand another thing: all that evident chopping and changing with the plot has led to a complete mess of different-sounding vocal takes from the various actors. There are so many edits with different-quality audio it gets really annoying and jarring.
I noticed this a lot in Act 3.. even companions where their voice sounds completely different between two audio clips.
I think it's common for audio to be redone.. but they clearly hired a buncha amateurs to voice most of this and they had almost no quality control. Sometimes a character just randomly sounds like they have a cold or they just woke up in a random dialogue node then goes back to their normal voice.
"Clearly the tadpole fucking with your hearing, we planned this from the start"
As I said to volkore, it's not the quality of the performances in and of themselves that's the problem, it's the difference in audio quality. As an audio guy, I think I understand the problem (which I'm kind of explaining here), but it's pretty bad in places (as you say, sometimes they sound like they've got the cold or whatever, that's down to the difference in audio quality).
Something went very right with the
direction of the voice acting, in terms of drawing the best out of performers and getting generally good context-specific feeling-tones to the performances, but something went quite wonky in the audio chain side of things.
*****************
Expanding on this further, it sounds like what they
should have done but didn't do, was record the audio through a "Larian audio recording laptop and mic combo" as described above, feed the digital out to the studio's software, let the studio people do any on-the-spot editing, and process it however they liked
for the monitoring only, but give back the edited but
unprocessed versions of the audio files to Larian, so Larian's own audio team would be able to process all the different audio batches uniformly. But what it
sounds like they did (maybe because it's such a huge project that they were pushed for time) was maybe specify the same mic/compressor for all sessions (some industry standard like the U87/Urei 1176 combo that every pro studio will have), but not exert enough control over the different types of monitoring processing that each studio used, and let the various studios "bake in" their various idiosyncratic processings.
On third though, it actually could be resolvable, because if they did it that way because they were pushed for time, then because everything will have been meticulously ordered in a database, they ought to be able to request the raw audio files from the various studios, so they can process them uniformly for an Enhanced Edition. That way, if they specified a certain input chain (a particular mic/compressor combo - which would still have more variability than the "Larian laptop" idea, because mics and hardware boxes can sound a bit different from studio to studio - but less variability than the variability we've got in the game now) that's the same in all recordings, then there will be more uniformity of tone if they can do what I've just described (request the raw files from the studios and process them all uniformly).
