xVASynth Voice Packs

[kadsend]

Hello, is it possible to add dremora voices and even maybe the dremora lines expansion voices to the voicepack?

[Lustgartenknecht]

Thank you for your work!

[Herowynne]

@Mister X Thanks so much for these voice packs! ❤️

 

The new voice pack for Laura's Bondage Shop 3.43 is great! ? It's really nice to hear the new first-time conversation between Laura and Katerina.

 

Would love to see an updated voice pack for Simple Slavery 6.3.21. There are quite a few changed voice lines since version 6.3.18. Thanks!

[Tseverin]

I'm sure I'm doing something wrong here. So far, Laura's voice pack doesn't seem to be working. Doesn't look like anything is overwriting it in MO2. Obviously can't check xedit.

 

Should I move it to the bottom of my load order? I just have it right after Laura's bondage shop.

 

Is there something I'm missing? 

[Gudulba]

@Mister X Thanks you so much for providing these voice packs!

 

And yes, Simple Slavery 6.3.21 would be great ?

 

And maybe Treasure Hunter Whore? ?

Edited January 4 by Gudulba

[Jasmine92]

On 12/31/2023 at 8:47 PM, Herowynne said:

@Mister X Thanks so much for these voice packs! ❤️

 

The new voice pack for Laura's Bondage Shop 3.43 is great! ? It's really nice to hear the new first-time conversation between Laura and Katerina.

 

Would love to see an updated voice pack for Simple Slavery 6.3.21. There are quite a few changed voice lines since version 6.3.18. Thanks!

 

On 1/2/2024 at 2:50 AM, Tseverin said:

I'm sure I'm doing something wrong here. So far, Laura's voice pack doesn't seem to be working. Doesn't look like anything is overwriting it in MO2. Obviously can't check xedit.

 

Should I move it to the bottom of my load order? I just have it right after Laura's bondage shop.

 

Is there something I'm missing? 

 

I ran into the same error, when trying to create my own voicepack for LBA 3.43 (only when I looked for the problem I stumbled upon this Version and realized @Mister X had already done all the work). The problem is, that the SSE_Edit Export Script, that generates the batch file for XVASysnth, only exports around 9k voice-lines, while the voicepacks for older LBA versions have around 40k voice-lines. Hence the lines that are there are better, but a lot are missing. I'm not versed enough in SSE_Edit to look for the problem, but maybe some will find it, now that one knows where to look.

 

@Mister X as a fellow XVASynth user I'm interested in your workflow. Since with my setup, the "all in one go synthesize -> ffmpeg -> lip -> fuz" process in Xvasynth is way to slow (less than 0.5 line/sec) and often fails and does not use parallel processing in xvasynth 3.0.3, I only generate the voice files via xvasynth and do all the post processing, lip generating and fuzzing with external helper scripts. Do you do everything in XVASynth one go? Have you managed to speed up the process?

[Lestat_Rhodes]

can I make a  request to have this mod voiced?

 

[Mister X]

12 hours ago, Jasmine92 said:

 

 

I ran into the same error, when trying to create my own voicepack for LBA 3.43 (only when I looked for the problem I stumbled upon this Version and realized @Mister X had already done all the work). The problem is, that the SSE_Edit Export Script, that generates the batch file for XVASysnth, only exports around 9k voice-lines, while the voicepacks for older LBA versions have around 40k voice-lines. Hence the lines that are there are better, but a lot are missing. I'm not versed enough in SSE_Edit to look for the problem, but maybe some will find it, now that one knows where to look.

 

@Mister X as a fellow XVASynth user I'm interested in your workflow. Since with my setup, the "all in one go synthesize -> ffmpeg -> lip -> fuz" process in Xvasynth is way to slow (less than 0.5 line/sec) and often fails and does not use parallel processing in xvasynth 3.0.3, I only generate the voice files via xvasynth and do all the post processing, lip generating and fuzzing with external helper scripts. Do you do everything in XVASynth one go? Have you managed to speed up the process?

 

I'm not sure about the script, my knowledge about Lua and xEdit scripting are very limited. What I do know is, that the script respects the lines of certain actors and their voicetypes. Meaning that it exports lines exclusively said by Laura only once for her voicetype and not for all female voices. In earlier versions the actors didn't have specific voices set, once that was done it reduced the amount of total voicelines drastically.

 

For my workflow: it's time consuming, but I imagine that it increases the over-all quality.

1) exporting the voice lines with my altered version of the extraction script

2) Refining some phrasings, spelling mistakes and known problems for the T2S feature (xVASynth doesn't like things like "..." or "?!?") - batch processing all files at once with VS Code

3) Batch creating voice lines with xVASynth features "super-resolution" and "Clean up" enabled (that's the most time consuming part)

4) Refining the output with SynthUtil to eliminate even more "robotic" sounds

5) Fire up CK (with fixes installed) to create LIP files

6) Unfuzer packs the .wav and .lip files into .fuz files ready to distribute (works for SE/AE, too)

7) OPTIONAL pack all the .fuz files into a .bsa archive with CAO

 

9 hours ago, Lestat_Rhodes said:

can I make a  request to have this mod voiced?

 

 

I never downloaded this mod, but I can have a look at it 

[Jasmine92]

On 1/12/2024 at 11:50 AM, Mister X said:

I'm not sure about the script, my knowledge about Lua and xEdit scripting are very limited. What I do know is, that the script respects the lines of certain actors and their voicetypes. Meaning that it exports lines exclusively said by Laura only once for her voicetype and not for all female voices. In earlier versions the actors didn't have specific voices set, once that was done it reduced the amount of total voicelines drastically.

Hm, that might although be a reason, However when I did a quick test I definitely hat missing lines when trying to talk to Laura and the other guard/girl in the shop. Looks like I need to test a bit more, maybe try your edited export script.

 

On 1/12/2024 at 11:50 AM, Mister X said:

For my workflow: it's time consuming, but I imagine that it increases the over-all quality.

1) exporting the voice lines with my altered version of the extraction script

2) Refining some phrasings, spelling mistakes and known problems for the T2S feature (xVASynth doesn't like things like "..." or "?!?") - batch processing all files at once with VS Code

3) Batch creating voice lines with xVASynth features "super-resolution" and "Clean up" enabled (that's the most time consuming part)

4) Refining the output with SynthUtil to eliminate even more "robotic" sounds

5) Fire up CK (with fixes installed) to create LIP files

6) Unfuzer packs the .wav and .lip files into .fuz files ready to distribute (works for SE/AE, too)

7) OPTIONAL pack all the .fuz files into a .bsa archive with CAO

1.) I was using the original script from @DoubleCheeseburger, will try your altered version.

2.) although doing that with a python script written in vscode

3.) I only create the voice files with no ffmpeg post processing in xva synth, because multiprocessing part of the post-processing seems to be broken in V 3.0.x and with ffmpeg in xvasynth activated i end up with ~ 1.4 lines/sec, while my GPU can churn out about ~30 lines/sec (RTX 3070) when omitting post processing.

3.5) To mend that I use another script that uses ffmpeg directly to up-sample,de-noise and de-ess the sound files, I messed with the filters a little bit and ended up with a decent sounding one. (ffmpeg -y -hide_banner -i "{src_wav}" -filter_complex aresample=resampler=soxr:precision=33,deesser=i=0.3,afftdn=nr=8:nf=-40 -acodec pcm_s16le -ar 44100 "{trgt_wav}")

4.-5.) I only use the Synthutils output for the lip generation (which takes the longest time out of all the processes) and I don't use the CK anymore, but found this little tool on github which speeds up the process nicely.

6.) same (took me 1 year to realize you can configure multiprocessing in the options...)

7.) using bsarch for that

Edited January 13 by Jasmine92

[Mister X]

On 1/13/2024 at 8:18 PM, Jasmine92 said:

I only create the voice files with no ffmpeg post processing in xva synth, because multiprocessing part of the post-processing seems to be broken in V 3.0.x and with ffmpeg in xvasynth activated i end up with ~ 1.4 lines/sec, while my GPU can churn out about ~30 lines/sec (RTX 3070) when omitting post processing.

 

Hmm, never tried that. For me it's mainly the extra options during generation that slow things down. FFMPeg past processing only takes a fraction of the generation time for one batch

[Jasmine92]

12 hours ago, Mister X said:

Hmm, never tried that. For me it's mainly the extra options during generation that slow things down. FFMPeg past processing only takes a fraction of the generation time for one batch

Was the same for me in XVASynth 2, v3 however broke something and now it's only one ffmpeg task at a time that takes around 1 second each. Reinstalling xvasynth and deleting all configs didn't help, even completely new windows install has the problem. (I'm using the steam version btw.)

[steamst]

The Download-link does not work. 😕

[Mister X]

On 1/20/2024 at 4:16 PM, steamst said:

The Download-link does not work. 😕

 

For me it sends you correctly to the MEGA folder

[steamst]

7 hours ago, Mister X said:

 

For me it sends you correctly to the MEGA folder

Now its working again. I got a Mega-Error message about deleted folder.

[UltraSeven]

Hi! Thanks a lot for your work, but could I request something?

 

Would it be possible for you to do voices for BaboDialogue? The latest version available currently is 5.21 (with 5.4 soon to be released for the public, currently only available for subscribestar members), and the most recent voicepack made by somebody on LL is for version 5.0 made in february 2023 by BadJoke.

 

I know its a huge undertaking, since theres a lot of lines to cover, but I wanted to ask you, since I'm pretty sure there's a bunch of people who would be happy to have an up to date voicepack.

 

Thanks!

 

 

[Mister X]

13 hours ago, UltraSeven said:

Hi! Thanks a lot for your work, but could I request something?

 

Would it be possible for you to do voices for BaboDialogue? The latest version available currently is 5.21 (with 5.4 soon to be released for the public, currently only available for subscribestar members), and the most recent voicepack made by somebody on LL is for version 5.0 made in february 2023 by BadJoke.

 

I know its a huge undertaking, since theres a lot of lines to cover, but I wanted to ask you, since I'm pretty sure there's a bunch of people who would be happy to have an up to date voicepack.

 

Thanks!

 

 

 

I can have a look at it once v5.4 is out to public, but no promises...

[ThatOneSpiderPhoenix]

For the Sexist Guards voice pack, how would I choose which voice type is used for player dialogue? It seems like it defaulted to female commoner (I think?) but I would prefer to use female young eager if possible.

[Mister X]

13 hours ago, ThatOneSpiderPhoenix said:

For the Sexist Guards voice pack, how would I choose which voice type is used for player dialogue? It seems like it defaulted to female commoner (I think?) but I would prefer to use female young eager if possible.

 

For that you have to change the voice type of your character for that play through. I recommend using PC Head Tracking and Voice Type SE or something similar for that.

[krafs]

On 2/5/2024 at 10:00 AM, Mister X said:

 

I can have a look at it once v5.4 is out to public, but no promises...

It's out as a public release on his SubscribeStar page. The releases found on LL are always lagging behind nowadays.

 

https://subscribestar.adult/posts/1098430

 

 

On a different note. Would you consider making an updated voice pack for Public Whore Comments? The current one is quite old at this point and far from complete.  High quality, just a bit old.

 

[mr wolf]

does the sluts resume voice work with LE version ? IF it does when I try to run the game with it install it doesnt even load with it install

[Mister X]

On 2/9/2024 at 7:46 PM, krafs said:

It's out as a public release on his SubscribeStar page. The releases found on LL are always lagging behind nowadays.

 

https://subscribestar.adult/posts/1098430

 

 

On a different note. Would you consider making an updated voice pack for Public Whore Comments? The current one is quite old at this point and far from complete.  High quality, just a bit old.

 

 

I have a few packs where I need / want to do updates. It just takes veeery long to create those voices, so it's something I do when I think of it and nothing else demanding needs to be done.

 

18 hours ago, mr wolf said:

does the sluts resume voice work with LE version ? IF it does when I try to run the game with it install it doesnt even load with it install

 

Read the description, namely the paragraph "How to use them with Legendary Edition". Unpack the bsa file and optionally repack it for LE

[krafs]

On 2/11/2024 at 2:50 PM, Mister X said:

 

I have a few packs where I need / want to do updates. It just takes veeery long to create those voices, so it's something I do when I think of it and nothing else demanding needs to be done.

 

Fair enough.