Fallout 2 mod RPU - Talking Heads Addon

Hey Goat Boy, Dogmeat is meant to be a blue-heeler, not a German Shepard. The inspiration was from Mad Max 2. Bethesda screwed up with their version!

tumblr_nv9yl5HrEb1qm932ko8_250.gifv


https://fallout.fandom.com/wiki/Dogmeat_(Fallout)

https://en.wikipedia.org/wiki/Dogmeat_(Fallout)

https://madmax.fandom.com/wiki/"Dog"#:~:text="Dog" was Max Rockatansky',heeler in Mad Max 2.
 
@Goat_Boy is there long term plans to eventually add voice animations and find voice actors?

Would love to help out on the project but still trying to get my feet wet with modding and I'm not the best animator / artist. But if you eventually need voice actors I'd be up to volunteering for a couple.

EDIT: Also, have you done the enclave npc's yet? I feel like they would be easy enough to do by simply copying one of the many already present animatioms available in the game. Wouldn't add a lot of uniqueness to each NPC but would knock a lot of them out
 
Last edited:
Right, well aware it's a big task to complete which is why I figured if anything it would be a long term plan. The work so far has been great though, looking forward to replaying the game with the latest updates.
 
@.Pixote.

I appreciate the feedback, maybe I'll change Dogmeat in the future

@Black_Electric

Yeah as mentioned above the Lip syncing program is pretty bad, to my knowledge it's the only one out there. It's really tedious making anything with it.

Once I have more motivation I'll give it a try though cause I was pleasantly surprised by all the folk who wanted to voice act on this mod :)
 
@.Pixote.

I appreciate the feedback, maybe I'll change Dogmeat in the future

@Black_Electric

Yeah as mentioned above the Lip syncing program is pretty bad, to my knowledge it's the only one out there. It's really tedious making anything with it.

Once I have more motivation I'll give it a try though cause I was pleasantly surprised by all the folk who wanted to voice act on this mod :)

If I ever find time outside of my other projects maybe I'll give take a crack at it. Love the work you've done so far!
 
Once I have more motivation I'll give it a try though cause I was pleasantly surprised by all the folk who wanted to voice act on this mod :)

It's not worth it IMO, the effort involved in lip syncing all the of heads you released would take years. Having them just react good, bad, neutral is fine.
 
It's not worth it IMO, the effort involved in lip syncing all the of heads you released would take years. Having them just react good, bad, neutral is fine.

I know it's totally not worth it hahaha

If I would though I'd probably just do the companions and MAYBE some major NPCs
 
It's not worth it IMO, the effort involved in lip syncing all the of heads you released would take years. Having them just react good, bad, neutral is fine.

Would it be difficult to just code a tool to perform the lipsync extraction? Would just need to analyze the audio recording vowels and automatically generate the sequence.

Speaking from ignorance here because I don't know how the speech recordings are formatted yet for the game.
 
Unless you are the one who is going to make such a tool, it's likely not going to happen.
 
Unless you are the one who is going to make such a tool, it's likely not going to happen.

Guess I know where to start... just wish the engine were updated to use a more modern audio format. Will have to generate the face sequences based off the .wav file before it's converted to .acm.

EDIT: @Goat_Boy are you maintaining all of the artwork files in case I could get this working? The head frames for the nine or so phonems would need to be created. (Not going to happen anytime soon, but don't want to start if you won't be able to add the additional frames to the current artwork you have for the bad/neutral/good faces in the future).
 
Last edited:
Guess I know where to start... just wish the engine were updated to use a more modern audio format. Will have to generate the face sequences based off the .wav file before it's converted to .acm.
EDIT: @Goat_Boy are you maintaining all of the artwork files in case I could get this working? The head frames for the nine or so phonems would need to be created. (Not going to happen anytime soon, but don't want to start if you won't be able to add the additional frames to the current artwork you have for the bad/neutral/good faces in the future).

Almost all characters have the needed talking animations :) If you ever need them I'd be happy to send you the sprites, you could also open the animations (.FRM files) in the mod with Fallout Frame Animator. The talking animations have either the NP, GP or BP prefix at the end of their name.
 
Almost all characters have the needed talking animations :) If you ever need them I'd be happy to send you the sprites, you could also open the animations (.FRM files) in the mod with Fallout Frame Animator. The talking animations have either the NP, GP or BP prefix at the end of their name.

@Goat_Boy that's great if you already have the .frm's competed. Including the *p.frm's?

I don't think I should need the sprites, I'm just going to try and do a proof of concept using Klint outside of the temple of trials, with Sulik's frm's and some crappy voice recording by me. The idea would be to make it seemless for you or anyone to gather the voice acting files and the frm's and then run the generator to automatically create the .LIP files, convert the .wav's to acm, etc.

If I can get that far then I will need to work with you on integrating the .lip's with your mod so they'll talk during speech.

I'm looking at the bit-wise for the .LIP's now to see if I can discern how to generate them based off this helpful description:
https://falloutmods.fandom.com/wiki/LIP_File_Format
 
Update: I've managed to fully document the list of phoneme codes on the wikia page. The previous author only worked out the FRM frames but with the Phonemes now documented, it makes it possible / easier to generate the LIP files:
https://falloutmods.fandom.com/wiki/LIP_File_Format

Next step is to understand how the frame timing for each phoneme is encoded into the file, then hopefully I can find some open source speech synthesizer to extract the phonemes from a .wav file (voice acted speech), otherwise this will be much more tedious.
 
Update: I've managed to fully document the list of phoneme codes on the wikia page. The previous author only worked out the FRM frames but with the Phonemes now documented, it makes it possible / easier to generate the LIP files:
https://falloutmods.fandom.com/wiki/LIP_File_Format

Next step is to understand how the frame timing for each phoneme is encoded into the file, then hopefully I can find some open source speech synthesizer to extract the phonemes from a .wav file (voice acted speech), otherwise this will be much more tedious.

Oh wow that's really cool! I usually went by this order:

1 - Closed (in-between frame)
2 - A
3 - E
4 - I
5 - F
6 - L
7 - M
8 - O
9 - U

If you manage this you'd be a hero my friend ;)
 
Last edited:
I'm curious if Black Isle used a speech synth or hand did these. I don't imagine speech synthesization was very good back then, but some of the phoneme sequences were a bit wierd.

Tired of listening to Frank Horrigan's sound files at this point to identify sound codes, lol.

Marcus comes in as the real MVP with some of the more clear audio recordings.

UPDATE: Looks like python has API support for speech to text transcription of .wav files using one of several online speech to text engines. Should be easy enough to batch process tons of .wav files at once. Then I'll just need to use a speech to phoneme converter to get the list of sounds and add time stamps to each. Writing it into a fallout-compatible .LIP should be easy at that point.
 
Last edited:
I'm curious if Black Isle used a speech synth or hand did these. I don't imagine speech synthesization was very good back then, but some of the phoneme sequences were a bit wierd.

Basically on the lip program I mentioned you do everything by hand, every single mouth movement. I guess the developers did it by hand as well?

UPDATE: Looks like python has API support for speech to text transcription of .wav files using one of several online speech to text engines. Should be easy enough to batch process tons of .wav files at once. Then I'll just need to use a speech to phoneme converter to get the list of sounds and add time stamps to each. Writing it into a fallout-compatible .LIP should be easy at that point.

Wow is it really that easy? Let me know how it goes, I could start working on the voice acting straight away.
 
Wow is it really that easy? Let me know how it goes, I could start working on the voice acting straight away.

Yeah I'll need some time to get the proof of concept working. I found an API that supposedly marks time stamps for each of the phoneme's in an audio file. Will need to evaluate how well it ID's the sounds and timing, that will be the hardest part.

Plan for tonight is just to evaluate the .acm and .lip files for how the timing is encoded in the binary, should be straight forward. Looks like some one has already done a write up on the fallout wikia which is nice. The perks of being in a community over 20 years old.

https://falloutmods.fandom.com/wiki/ACM_File_Format
 
Yeah I'll need some time to get the proof of concept working. I found an API that supposedly marks time stamps for each of the phoneme's in an audio file. Will need to evaluate how well it ID's the sounds and timing, that will be the hardest part.

Plan for tonight is just to evaluate the .acm and .lip files for how the timing is encoded in the binary, should be straight forward. Looks like some one has already done a write up on the fallout wikia which is nice. The perks of being in a community over 20 years old.

https://falloutmods.fandom.com/wiki/ACM_File_Format

Coding is fascinating, looks like magic to untrained eyes like mine hahaha
 
Back
Top