Intelligent Voiced Dirty Talk SE

[y_sengaku]

Really thank you for updating the mod. 

 

>The biggest difficulty in all this really is just the time it takes to go through the dialogue lines and find the bits that would work (description) 

 

I wonder whether er can make use of Lazy Voice Finder - FInd voice assets just you want, though I certainly have difficulty in running this tool with my latest (win11) laptop as is also with the case of the same author's Yakitori Audio Converter. 

 

It might especially be useful to find the parallel line like "come on....." among various voice types.

[Guest]

18 hours ago, foxdogcat said:

Thanks for taking my request and adding support for gay scenes! I appreciate it!

You're welcome! Thanks for your interest!

[Guest]

17 hours ago, y_sengaku said:

Really thank you for updating the mod. 

 

>The biggest difficulty in all this really is just the time it takes to go through the dialogue lines and find the bits that would work (description) 

 

I wonder whether er can make use of Lazy Voice Finder - FInd voice assets just you want, though I certainly have difficulty in running this tool with my latest (win11) laptop as is also with the case of the same author's Yakitori Audio Converter. 

 

It might especially be useful to find the parallel line like "come on....." among various voice types.

I had not heard of that, interesting. If/when I make more voices, I'll probably check this out in addition to the AI voice generators. If it sounds like what it is, it could be a huge help. Thanks for the tip.

[Guest]

20 hours ago, BetLV said:

Heya! I'm testing this out, really liking the added support. So I've noticed that my new werewolf voice, with custom voice files and all, isn't playing in non aggressive scenes. With Yamete Kudasai, I get a scene with aggressive dialogue. But when I make a regular scene with Matchmaker, I get no voice from the male werewolf. I am not sure whether this is intentional... but I would have liked to have dialog playing for both occasions! I have tried to test various settings out ingame with your MCM if it had any effect, btw. Is there any fix I could try? I'm not extremely versed in papyrus, but if you have some suggestion that would help!

Is there also any support for werewolves only scenes? It may not have been anything you'd think about, but HCOS has straight and gay werewolf duo animations.

Thanks for letting me know. I didn't fully test the creature voices, so this is good feedback. I'll check it out.

[z3nkai]

17 hours ago, y_sengaku said:

Really thank you for updating the mod. 

 

>The biggest difficulty in all this really is just the time it takes to go through the dialogue lines and find the bits that would work (description) 

 

I wonder whether er can make use of Lazy Voice Finder - FInd voice assets just you want, though I certainly have difficulty in running this tool with my latest (win11) laptop as is also with the case of the same author's Yakitori Audio Converter. 

 

It might especially be useful to find the parallel line like "come on....." among various voice types.

 

Neither of those programs worked for me ended up using BSA browser and Skyrim Audio converter. (I'm the guy trying to use the Ai voice Generator.)

[y_sengaku]

10 minutes ago, z3nkai said:

 

Neither of those programs worked for me ended up using BSA browser and Skyrim Audio converter. (I'm the guy trying to use the Ai voice Generator.)

 

I assume we have to install some (not only the latest 2015+) virtual studio redistributable for using their audio function, though win11 might have a potential compatibility problem with some of them like VS 10.  

 

I at least confirm that at least the search function by word/ phrase of Lazy voice finder from the vanilla voice files (that has been problematic to find the phrase in question) also work with my latest win11 machine, however.

Edited March 4, 2023 by y_sengaku

[Guest]

On 3/3/2023 at 3:28 PM, BetLV said:

Heya! I'm testing this out, really liking the added support. So I've noticed that my new werewolf voice, with custom voice files and all, isn't playing in non aggressive scenes. With Yamete Kudasai, I get a scene with aggressive dialogue. But when I make a regular scene with Matchmaker, I get no voice from the male werewolf. I am not sure whether this is intentional... but I would have liked to have dialog playing for both occasions! I have tried to test various settings out ingame with your MCM if it had any effect, btw. Is there any fix I could try? I'm not extremely versed in papyrus, but if you have some suggestion that would help!

Is there also any support for werewolves only scenes? It may not have been anything you'd think about, but HCOS has straight and gay werewolf duo animations.

A voice assigned to the MaleWerewolf voice type should always play when a male actor is a werewolf as long as its the only male voice for the scene (or its one of two voices in a gay scene). At least, that was the intent. I find it odd that it changes based on whether the scene was aggressive because the logic is the same regardless--there has to be something else going on. The system knows if an actor's a werewolf by checking if the actor race is WerewolfBeastRace, and as far as I know all werewolves have this race. It then determines whether to use MaleWerewolf or FemaleWerewolf based on the actor's SexLab gender, where beast race genders are taken into account (beast genders is an option in the SexLab MCM). Is there any other differences you know of between when it worked and when it didn't? Was the actor's race, actor's gender, or number/gender of actors in the scene different?

 

If not, is there something you could point me to reproduce the issue? Like what mods should I get to test the werewolves?

[MattTheRagdoll]

Hey currently working on putting together my own voice pack, for anyone doing the same it worth noting that the files need to be 44100 Hz sampling and 16bit not 32bit, learned that the hard way! I was just wondering how many sound files can we put in each folder (for each gender)? I'm using the provided one as a guide, but could I add more sound files? If so would I need to edit the ESP?

 

Nevermind figured it out (was looking at the wrong part of the ESP)! Great work btw!

Edited March 6, 2023 by MattTheRagdoll

[BetLV]

On 3/5/2023 at 12:14 AM, L0rd0fT0rm3nt_ said:

A voice assigned to the MaleWerewolf voice type should always play when a male actor is a werewolf as long as its the only male voice for the scene (or its one of two voices in a gay scene). At least, that was the intent. I find it odd that it changes based on whether the scene was aggressive because the logic is the same regardless--there has to be something else going on. The system knows if an actor's a werewolf by checking if the actor race is WerewolfBeastRace, and as far as I know all werewolves have this race. It then determines whether to use MaleWerewolf or FemaleWerewolf based on the actor's SexLab gender, where beast race genders are taken into account (beast genders is an option in the SexLab MCM). Is there any other differences you know of between when it worked and when it didn't? Was the actor's race, actor's gender, or number/gender of actors in the scene different?

 

If not, is there something you could point me to reproduce the issue? Like what mods should I get to test the werewolves?

I used the m1 folder, and is the only voice so it's default for both male and malecreature. I was playing with SL Matchmaker as well as SL Aroused Creatures and all normal, non aggressive scenes play no male voice. The aggressive scenes with Yamete Kudasai played aggressive dialogue. This has been with 1 female humanoid and 1 male werewolf in all scenes. The MCM didn't seem to have much effect by fiddling with the settings.

[Guest]

12 hours ago, MattTheRagdoll said:

Hey currently working on putting together my own voice pack, for anyone doing the same it worth noting that the files need to be 44100 Hz sampling and 16bit not 32bit, learned that the hard way! I was just wondering how many sound files can we put in each folder (for each gender)? I'm using the provided one as a guide, but could I add more sound files? If so would I need to edit the ESP?

 

Nevermind figured it out (was looking at the wrong part of the ESP)! Great work btw!

Thank you sir! Are you sure the bit depth has to be 16 bit, though? Because I thought my audio was 32 bit. I might be confused, I'm actually just learning these terms for the first time. Maybe its a difference between computers, operating systems, or Skyrim versions?

[ShimizuModding]

how many voice clips can each folder support?

[Guest]

22 minutes ago, ShimizuModding said:

how many voice clips can each folder support?

Its different per folder. If you look at one of the voices already made, its the exact number they have. The filenames have to be the same too.

[ShimizuModding]

does the mod wait for the currently playing voice clips to finish before the next clip, so i can add long clips without overlapping

Edited March 7, 2023 by ShimizuModding

[Guest]

9 hours ago, ShimizuModding said:

does the mod wait for the currently playing voice clips to finish before the next clip, so i can add long clips without overlapping

99% of the time, yes. The exception is that the MaleOrgasmOral and MaleOrgasmNonOral clips for the female voice do not wait for completion because they play at the same instant in time as the male voice's Orgasm sound. In other words, the clips for the female reacting to the male orgasm in real-time should be the same length or shorter than the male orgasm clip. There is one other exception, but it is just a glitch that I will add a patch for soon.

 

So, in general long clips are OK. The only other things you should be aware of are that the audio clips will always play to completion regardless of what happens and that disk-space could theoretically be a concern if you're adding a lot of long clips since wav files are pretty big, but it would probably be OK.

[Guest]

On 3/6/2023 at 1:10 PM, BetLV said:

I used the m1 folder, and is the only voice so it's default for both male and malecreature. I was playing with SL Matchmaker as well as SL Aroused Creatures and all normal, non aggressive scenes play no male voice. The aggressive scenes with Yamete Kudasai played aggressive dialogue. This has been with 1 female humanoid and 1 male werewolf in all scenes. The MCM didn't seem to have much effect by fiddling with the settings.

I'm sorry man, I couldn't figure it out. I spent a few hours trying to debug it but with no success. I've had the worst time with things being finicky with no apparent pattern of behavior, despite my best attempts. I was able to get some werewolves to play their voices but I had just as many times that they didn't play any voice. There was no consistent pattern of behavior. Then, I started to have the same problem with scenes that didn't even have any creatures. Its not your fault, but I'm totally burnt out.

[y_sengaku]

Hello,

 

So far, I have little problem on using your mod (additional way of volume adjustment might be preferable, though).

 

BTW, I'd appreciate to have some clarifications especially on voice slot assignment from MCM when you have time to spare. 

 

1. Question #01: On what criteria the individual actor assignment function mainly distinguishes various actors? BaseID, RefID, Full-Name or shortened name?  I have this question both A) because I'm not sure how your mod distinguish individual player characters across different save datas, and B) the game sometimes has multiple "double" actors with different refID like Ulfric (main, civil war, and Sovngarde version) - I know SLFramework  itself and other mods tend to treat them as almost totally different by character base (different sexual experience, vanilla voice assignment, and also schlong size as well). Can your mod (individual actor assignment) treat them together as the same actor in the slot assignment?
2. Question #02 (main): Is it somehow possible to assign more than two voice types/ actors with the single voice file slot of your mod? To give an example, can we assign both (and only) MaleArgonian and MaleKhajiit with M2 (MaleArgonian)? In such a case, I'm now duplicating M2 and M3 with the same voice files and assigning these voice type actors separately (M2 for MaleArgonian and M3 for Male Khajiit actors, but the files themselves are the same).  
3. Question #03: Is there also any way of resetting the assignment from MCM? Now I either imput "Empty" or something in MCM or just delete the additional tweaks from .ini file directly, but I'm afraid that my ways of this kind of editing maybe too crude.

Edited March 10, 2023 by y_sengaku

[NoppaiKohai]

If you are considering doing AI voices, there's a thread going where the OP has a pretty hefty amount of tokens to spare to help voice stuff, and there's likely others who would gladly assist

 

 

 

We need to organize a landing place for AI voice line request from mod authors, I bet quite a few here would pay the few bucks a month for Elevenlabs and contribute to revoicing some of the mods that have no voices or janky voices

[Guest]

4 hours ago, y_sengaku said:

Hello,

 

So far, I have little problem on using your mod (additional way of volume adjustment might be preferable, though).

 

BTW, I'd appreciate to have some clarifications especially on voice slot assignment from MCM when you have time to spare. 

 

1. Question #01: On what criteria the individual actor assignment function mainly distinguishes various actors? BaseID, RefID, Full-Name or shortened name?  I have this question both A) because I'm not sure how your mod distinguish individual player characters across different save datas, and B) the game sometimes has multiple "double" actors with different refID like Ulfric (main, civil war, and Sovngarde version) - I know SLFramework  itself and other mods tend to treat them as almost totally different by character base (different sexual experience, vanilla voice assignment, and also schlong size as well). Can your mod (individual actor assignment) treat them together as the same actor in the slot assignment?
2. Question #02 (main): Is it somehow possible to assign more than two voice types/ actors with the single voice file slot of your mod? To give an example, can we assign both (and only) MaleArgonian and MaleKhajiit with M2 (MaleArgonian)? In such a case, I'm now duplicating M2 and M3 with the same voice files and assigning these voice type actors separately (M2 for MaleArgonian and M3 for Male Khajiit actors, but the files themselves are the same).  
3. Question #03: Is there also any way of resetting the assignment from MCM? Now I either imput "Empty" or something in MCM or just delete the additional tweaks from .ini file directly, but I'm afraid that my ways of this kind of editing maybe too crude.

 

I'll consider adding a volume adjustment slider, but it would only be a master slider for all the mod's voices and its already at the max volume as is.

1. If you're familiar with Papyrus, it gets the actor name by actorToVoice.GetLeveledActorBase().GetName(). I did a quick test and this returns the full name.
2. Not currently. The duplication trick you mentioned is the way to go for now.
3. Manually inputting "Empty" is fine. You could really insert any nonsense in there and it would be fine. As long as its not anyone's name.

 

[Guest]

8 minutes ago, NoppaiKohai said:

If you are considering doing AI voices, there's a thread going where the OP has a pretty hefty amount of tokens to spare to help voice stuff, and there's likely others who would gladly assist

 

 

 

We need to organize a landing place for AI voice line request from mod authors, I bet quite a few here would pay the few bucks a month for Elevenlabs and contribute to revoicing some of the mods that have no voices or janky voices

Thanks for the heads up. I may *occasionally* do a voice, but for the most part I'm trying to not get too pulled into this project because I have some other stuff going on. I respect other people's interest in doing AI voices, but I'm a bit old school with manual mixing. So that said, I think it should be someone else besides me to take up the helm for creating voices, at least from the AI side.

[z3nkai]

Just a heads up. Still working on the AI stuff just two things:

1) Coming up with dirty talk that's either lord accurate or lore neutral while at the same time not sounding cheesy is harder than I thought it would be (I literally have a pornhub open in another tab in hopes a line pops out that gives me inspiration.) I think the problem is one of too many options as I'm not tied down with having to use existing dialog.

 

2) Eleven Labs doesn't seem to have a system in place to force the AI to emphasize certain words over others which add a bit more trail and error to get the right tone. Its a "I didn't say we should kill him" problem as what words are empathized drastically changes what's being said.

 

That being said, once I have a script that I like for each of the mods set topics should be somewhat smoother sailing.

Edited March 12, 2023 by z3nkai
Highlights and lowlights

[y_sengaku]

Just a brief note for those who try to make the  voice pack (including myself):  

I'd personally recommend also to check some vanilla voice files especially for grunting/ with very strong tone even if you try to build the voice pack with help of AI.

 

The majority of the vanilla voice types in fact share these files with corresponding ID especially for grunting when attacking/ damaged.
Some of them also have out of breathing sound by the same voice actor, and they might especially be useful since AI seems not so be good to generate this kind of sound.

 

These are some examples of such lines: 

• dialoguege__0003526d_1.fuz: attack with the sound of out of breathing
• dialoguege__0003526f_1.fuz: (the same as above)
• dialoguege__00035270_1.fuz: damaged
• dialoguege__00035271_1.fuz: damaged (strong)
• dialoguege__00035272_1.fuz: damaged plus a breathing (out of breath)
• dialoguege__00035273_1.fuz: damaged (strong) plus a breathing
• (dialoguege__00047615_1.fuz): these 0004761(x) are voices for attacking when fatigued (attacking grunt plus out of breathing sound)
• dialoguege__00047616_1.fuz:
• dialoguege__00047617_1.fuz:
• dialoguege__00047618_1.fuz:
• dialoguege__00047619_1.fuz:
• dialoguege__0004761a_1.fuz:
• dialoguegeneric__00013ee5_1.fuz: dialoguegeneric__00013ee(x)/ 00013ef(x) are primary voice for attacking.
• dialoguegeneric__00013ef0_1.fuz
• dialoguegeneric__0001717d_1.fuz: dialoguegeneric__0001717d,0001717e, 1717f: short attacking sound.
• dialoguegeneric__00017702_1.fuz: badly damaged, often almost scream
• dialoguegeneric__00017703_1.fuz

 

They must also be important as a mix base of the orgasm sound (its grunt part).

 

Some unique actors and dlc added voice time does not certainly share these ID in file names, however, but they belong rather to the irregular and I suppose relatively people will try to build their voice pack for just one or a few actors.

Some other lines that might be of interest: 

• dialoguege_dialoguegeneric_000e0cc6_1.fuz: "I'll teach you to talk to me that way!"
• dialoguegeneric__0002a467_1.fuz: "By the gods, no!"
• dialoguegeneric__00039b97_1.fuz: "No. Not like this......"
• dialoguegeneric__000494ad_1.fuz: "This what you want? Huh?"
• dialoguegeneric__000964ff_1.fuz: "that's your best?"
• dialoguegeneric__00096848_1.fuz: "Gods... damn you!"
• dialoguegeneric__0009687f_1.fuz: "This what you want?"
• duneldergl_duneldergleamt0_0001fb79_1.fuz: source of "who would ever want to harm such beauty?" in various voice types (in spite of the comment on Eldergleam Sanctuary (as a follower), even some unique voice have this voice line file in common).
• widragonki_widragonkilleds_00050cd5_1.fuz: "By the gods... I don't even know what to say..."
• wisharedin_wisharedinfosto_0006d1de_1.fuz: "By the gods!"

[dragonged12]

Just a regular user here... I've gotten the mod set up so that it shows the MCM and, I can make all the choices there. I just can't figure out how to make them sound in-game. Extracted the file folders, (male and female even-toned0 into the mod's sound folder, where I can see two other folders... (gape and smack.). I renamed each download folder M1, and F1. I see inside the other folders that they are wav files when you open them, but that the download folders have more folders in them, which contain the wav files them selves. What am I missing? thanks

[y_sengaku]

2 hours ago, dragonged12 said:

Just a regular user here... I've gotten the mod set up so that it shows the MCM and, I can make all the choices there. I just can't figure out how to make them sound in-game. Extracted the file folders, (male and female even-toned0 into the mod's sound folder, where I can see two other folders... (gape and smack.).

 

You actually have to put your custom voice folder (F1, F2, M1, M2, M3.....) in Data/Sound/fx/IVDT (thus one level above on these folders), not Data/Sound/fx/IVDT/Sounds (one level below) where gape and smack sound folders are located.

 

 

[Zukzukarr]

Love how you keep developing this! Hopefully some AI voice packs are gonna be released by others, can't wait.

[drygg49]

On 3/12/2023 at 1:53 AM, z3nkai said:

Just a heads up. Still working on the AI stuff just two things:

1) Coming up with dirty talk that's either lord accurate or lore neutral while at the same time not sounding cheesy is harder than I thought it would be (I literally have a pornhub open in another tab in hopes a line pops out that gives me inspiration.) I think the problem is one of too many options as I'm not tied down with having to use existing dialog.

 

2) Eleven Labs doesn't seem to have a system in place to force the AI to emphasize certain words over others which add a bit more trail and error to get the right tone. Its a "I didn't say we should kill him" problem as what words are empathized drastically changes what's being said.

 

That being said, once I have a script that I like for each of the mods set topics should be somewhat smoother sailing.

 

Since I don't know anything about this stuff, I could help what I can; that is to come up with lore friendly dirtytalk. 

 

I could write u some lines maybe, if u need inspiration.

 

nice work btw