Diary Of Mine for Paradise Halls

[Amazonian Ara]

I've had one slave stall at 32% for the longest as well.  I finally placed him in the house slave camp (HSH) for a few while I worked with another slave.  When I took him out, he quickly got angry and gave me an excuse for a good whipping and restraint session.  That promptly jumped him into the high 40s and then up to 62.

 

Now for something slightly unrelated, is there another place in Falkreath that wants a sex slave other than the inn?  And are worker slaves ones for farms or mines?

 

Edit: is there any chance of some backwards compatibility of DoM with HSH so that resignation is also raised somewhat in tune with respect during HSP training?

Edited April 9, 2022 by Amazonian Ara

[DocClox]

21 minutes ago, TrollAutokill said:

I meant the "Enjoying your slave life?" status

 

I posed three images in the spoiler in my last post. The last one should have been the "Enjoying" one.

 

 

 

[Gukahn]

47 minutes ago, Amazonian Ara said:

Now for something slightly unrelated, is there another place in Falkreath that wants a sex slave other than the inn?  And are worker slaves ones for farms or mines?

That's more of a And you get a slave question not a Dom Question (unless i miss some updates since i still wait to do it.) but to answer that, you can sell your slave to almost everyone, depending on the pahe states they have they will be either that or this, inkeeper buy them sometimes as wenches or hookers, similiar to other npc who buy them as worker or lovers.

 

@DocClox

Weird thing happen with your resignation states.. Mine rises pretty fast sometimes even faster then submission ? but i am on 2.7.5

Edited April 9, 2022 by Gukahn

[DonQuiWho]

7 hours ago, DocClox said:

 

Easy enough. The number is three separate tats, for hundred, tens and units. You need to make the tats and position them so they all line up to make a number in whatever place you wish. Making the tats is the time consuming part of the process.

 

You need a number for the slaves. The Spellbook's Brand spell adds the slave into a list and uses the index of that list as t he number. I could have used the PAHE list index, but I wanted something that wouldn't change if slaves were sold or otherwise removed from that list.

 

Once you've got the number, you need a script that breaks it down into hundreds, tens and units and applies the corresponding number tat for each position. The Brand spell does that, too.

 

Feel free to lift the code from the spellbook if you want.

 

 

Love you too, Antiope. As has been pointed out, there is a json file you can edit to apply whatever combination of tats you wish. As it happens, you can also use the slave number tats from the base SlaveTats package and have the number auto-applied rather than spending the time mucking around in slavetats MCM

 

 

Or, you know, magic. But like I say, you don't need to use my tats.

 

I'd love to do my own, but I read your guide and decided the science of advanced pocket rocket calligraphy may be beyond me ...  ?

 

Is there a really simple ABC guide covering the application of nice little pic tattoos for those like me who are still in our infancy. and struggling to get the cat to sit on the mat?

[DocClox]

18 minutes ago, DonQuiWho said:

Is there a really simple ABC guide covering the application of nice little pic tattoos for those like me who are still in our infancy. and struggling to get the cat to sit on the mat?

 

You mean making new tats, or customising the json file so you can make custom livery out of existing tats?

 

I'm happy to have a go at either.

[DonQuiWho]

3 hours ago, DocClox said:

 

You mean making new tats, or customising the json file so you can make custom livery out of existing tats?

 

I'm happy to have a go at either.

 

I can probably work out how to create tats - or at least nick my favourite ones off the various Tat Packs available out there.

 

It's the .json structure I didn't understand, insofar as how does one add a simple tat through that?  I did read what you wrote, and I really do appreeciate that you did that in the first place, but writing from knowledge is always different from reading from ignorance  ?

[Antiope_Apollonia]

So, I have finally finished creating the custom personalities for all the NPCs in Whiterun, as I promised was coming months ago.

 

I've tried to make them as lore-friendly as possible, relying on what can be learnt from their dialogues, lifestyle, profession, relationships to other NPCs, etc.—you don't want to know how much time I've spent staring at the wiki pages of the more obscure NPCs, digging for insights, lol.  I've tried both to ensure a realistic distribution of traits across the population as a whole, and to make the individual personalities themselves as realistic as possible for the most part.  I do have reasonable expertise in psychometrics, as my professional life touches on this academic area extensively, so I'm reasonably confident in my work here.  I've taken a little bit of liberty to give "better" personalities to fan-favourite NPCs and "worse" ones to reviled characters, but only a little bit.  I've prioritised lore-friendliness and realism first and foremost.  

 

Note that I haven't done anything with the children for now; I may still do in the future.

 

But...

They're all still in my spreadsheet I've been working from.  I still need to get them into the .json format that DoM reads.  Which is pretty daunting in terms of the tedium and time involved.  So I'm wondering if anyone here with some programming skills would be able to work up a simple utility to do that relatively easily?  My ignorance of coding is pretty high, but I wouldn't think that'd be hard to do for someone with the relevant expertise.  If I had such a utility available, it'd also probably help encourage me to keep working on this; I'm hoping to eventually do all vanilla NPCs, but that's going to be a while.  If anyone is able to create such a tool, I'd be grateful for the data entry pain you'd be saving me, and I'll have more time available to work on the NPC enslavement quests.

 

I just need to be able to convert rows of a spreadsheet with:

Name, Location, Gender, Honest, Calm, Lively, Kind, Thorough, Open, Wilful, Tough, Sensual, Submissive, Bold, Smart, BaseID

... into the .json format found in \Data\SKSE\Plugins\StorageUtilData\Diary Of Mine

 

When they're done, I'll pack them up in an archive that can be installed with a mod manager, and Troll can stick them on the download page if he wishes.

Edited April 9, 2022 by Antiope_Apollonia

[karellen]

Hello there! I am curious about the slaver npcs and the training process. Do they train the slave in the sensual arts? Do they play animations or do they just modify the values? Do they use ZAZ furniture for training the slaves? 

[DocClox]

10 hours ago, DonQuiWho said:

 

I can probably work out how to create tats - or at least nick my favourite ones off the various Tat Packs available out there.

 

It's the .json structure I didn't understand, insofar as how does one add a simple tat through that?  I did read what you wrote, and I really do appreeciate that you did that in the first place, but writing from knowledge is always different from reading from ignorance  ?

 

Fair enough.  Let's give this a go.

 

The file you want to modify is in data\docclox\slavers_spellbook\json\branding.json. It's basically a text file, so you can open it with notepad or any text editor.

 

This is my current banding file:

 

[
	{ "name": "Black", "section": "Stockings", "area": "Body" },
	{ "name": "Black", "section": "Stockings", "area": "Feet" },
	{ "name": "Slave (Pubic)", "section": "Slave Marks", "area": "Body" },
	{ "name": "Running Mascara (right)", "section": "Makeup", "area": "Face" },
	{ "name": "Running Mascara (left)", "section": "Makeup", "area": "Face" },
	{ "name": "Tears 3", "section": "Tears", "area": "Face" }
]

 

Json is basically javascript, if that helps at all. Square brackets indicate a list and curly brackets mean grouped data. So each "{ ... }" entry is a separate tattoo.

 

The fields inside a tattoo entry are basically the SlaveTats menu entries. So if you look at the "Slave (Pubic)" entry in the list, you can see it matches up with the area, section and make from the MCM:

 

 

And it's really as easy as that. Just list the MCM values for the tats you want and they'll be applied by the spell.

 

Serial numbers need a bit extra care.Let's add the slave numbers from the base MCM to this list. "area" is "Body" as before, and "section" is "Slave Number", but we don't know the tat name since it varies with the number. So instead of giving it a name, we use the special entry "is_serial" to indicate that this tat is a serial number, and we give it the value one which means "true" in computerland.

 

So this is my amended branding file

 

[
    { "name": "Black", "section": "Stockings", "area": "Body" },
    { "name": "Black", "section": "Stockings", "area": "Feet" },
    { "name": "Slave (Pubic)", "section": "Slave Marks", "area": "Body" },
    { "name": "Running Mascara (right)", "section": "Makeup", "area": "Face" },
    { "name": "Running Mascara (left)", "section": "Makeup", "area": "Face" },
    { "name": "Tears 3", "section": "Tears", "area": "Face" },
    { "is_serial" : 1, "section": "Slave Number", "area": "Body" }
]

 

To apply the tats, cast the Brand spell from the Spellbook. If you don't like the result you can remove the tats with the Clear Branding spell.

 

... and that ought to work. And doesn't. In fact the numbers don't appear when I apply them through the MCM, so something is clearly wrong with my SlaveTats install. Let me get back to you.

 

And maybe I should move this to a blog post or something before I take over Troll's thread.

 

[Antiope_Apollonia]

14 minutes ago, DocClox said:

And maybe I should move this to a blog post or something before I take over Troll's thread.

A blog post for this would be great.  Not only does it not derail this thread, but it's much easier to find in the future.  I have questions, but maybe better to hold my tongue until you're ready to post said blog. ? 

Edited April 10, 2022 by Antiope_Apollonia

[TrollAutokill]

10 hours ago, Antiope_Apollonia said:

So, I have finally finished creating the custom personalities for all the NPCs in Whiterun, as I promised was coming months ago.

 

I've tried to make them as lore-friendly as possible, relying on what can be learnt from their dialogues, lifestyle, profession, relationships to other NPCs, etc.—you don't want to know how much time I've spent staring at the wiki pages of the more obscure NPCs, digging for insights, lol.  I've tried both to ensure a realistic distribution of traits across the population as a whole, and to make the individual personalities themselves as realistic as possible for the most part.  I do have reasonable expertise in psychometrics, as my professional life touches on this academic area extensively, so I'm reasonably confident in my work here.  I've taken a little bit of liberty to give "better" personalities to fan-favourite NPCs and "worse" ones to reviled characters, but only a little bit.  I've prioritised lore-friendliness and realism first and foremost.  

 

Note that I haven't done anything with the children for now; I may still do in the future.

 

But...

They're all still in my spreadsheet I've been working from.  I still need to get them into the .json format that DoM reads.  Which is pretty daunting in terms of the tedium and time involved.  So I'm wondering if anyone here with some programming skills would be able to work up a simple utility to do that relatively easily?  My ignorance of coding is pretty high, but I wouldn't think that'd be hard to do for someone with the relevant expertise.  If I had such a utility available, it'd also probably help encourage me to keep working on this; I'm hoping to eventually do all vanilla NPCs, but that's going to be a while.  If anyone is able to create such a tool, I'd be grateful for the data entry pain you'd be saving me, and I'll have more time available to work on the NPC enslavement quests.

 

I just need to be able to convert rows of a spreadsheet with:

Name, Location, Gender, Honest, Calm, Lively, Kind, Thorough, Open, Wilful, Tough, Sensual, Submissive, Bold, Smart, BaseID

... into the .json format found in \Data\SKSE\Plugins\StorageUtilData\Diary Of Mine

 

When they're done, I'll pack them up in an archive that can be installed with a mod manager, and Troll can stick them on the download page if he wishes.

Something like that, https://www.excel-easy.com/vba/examples/write-data-to-text-file.html

[DonQuiWho]

6 hours ago, DocClox said:

Fair enough.  Let's give this a go.

 

And maybe I should move this to a blog post or something before I take over Troll's thread.

 

Thanks very much.  Very helpful.  Look forward to the blog post

[TrollAutokill]

On 4/9/2022 at 3:38 PM, Amazonian Ara said:

 

Edit: is there any chance of some backwards compatibility of DoM with HSH so that resignation is also raised somewhat in tune with respect during HSP training?

There is an HSH patch. So resignation should go up if you installed the patch and put DoM below in your MO2 load order.

[sidfu1]

so troll what do you think we shoul do for uniques that when enslaved do not have their armor in their iventory to remove?

[Jasmine92]

 

20 hours ago, Antiope_Apollonia said:

So, I have finally finished creating the custom personalities for all the NPCs in Whiterun, as I promised was coming months ago.

 

I've tried to make them as lore-friendly as possible, relying on what can be learnt from their dialogues, lifestyle, profession, relationships to other NPCs, etc.—you don't want to know how much time I've spent staring at the wiki pages of the more obscure NPCs, digging for insights, lol.  I've tried both to ensure a realistic distribution of traits across the population as a whole, and to make the individual personalities themselves as realistic as possible for the most part.  I do have reasonable expertise in psychometrics, as my professional life touches on this academic area extensively, so I'm reasonably confident in my work here.  I've taken a little bit of liberty to give "better" personalities to fan-favourite NPCs and "worse" ones to reviled characters, but only a little bit.  I've prioritised lore-friendliness and realism first and foremost.  

 

Note that I haven't done anything with the children for now; I may still do in the future.

 

But...

They're all still in my spreadsheet I've been working from.  I still need to get them into the .json format that DoM reads.  Which is pretty daunting in terms of the tedium and time involved.  So I'm wondering if anyone here with some programming skills would be able to work up a simple utility to do that relatively easily?  My ignorance of coding is pretty high, but I wouldn't think that'd be hard to do for someone with the relevant expertise.  If I had such a utility available, it'd also probably help encourage me to keep working on this; I'm hoping to eventually do all vanilla NPCs, but that's going to be a while.  If anyone is able to create such a tool, I'd be grateful for the data entry pain you'd be saving me, and I'll have more time available to work on the NPC enslavement quests.

 

I just need to be able to convert rows of a spreadsheet with:

Name, Location, Gender, Honest, Calm, Lively, Kind, Thorough, Open, Wilful, Tough, Sensual, Submissive, Bold, Smart, BaseID

... into the .json format found in \Data\SKSE\Plugins\StorageUtilData\Diary Of Mine

 

When they're done, I'll pack them up in an archive that can be installed with a mod manager, and Troll can stick them on the download page if he wishes.

 

should be quite easily done with python, which can read csv and write json files, tricky part would be the json structure, depending on how many dimensions it has. Do you have an example of the csv and the json? I'm no python crack either, but i could try.

 

You timing however is quite intriguing:

right now I'm overhauling my complete modlist for a new playthrough, since many things are new, or got updates. (DoM is one of them). So i tested the 2.7.5 build and as mentioned before, got hooked on that mod. (damn you @TrollAutokill for stealing so much of my time! ) Testplaying and reading this thread lead me to @reedeonry's voicepack over here, which boosted my immersion even further. Which lead to me trying out XVAsynth to redo the voices, because I use a never Version of DoM and i didn't want to miss out on spoken lines. I soon realized that, as good as it is and really fits well in some cases, there are others, that will let you think "hu? What did you say?" ("Thank you! I live to please you, mistress." being one of those lines, where the word live ends up being pronounced as in "the broadcast was live" not like "i live in Boston"). And before I new it i was sifting through a couple hundred unique lines of DoM, pasting them into xvasynth and changing the sentence to make them sound better. (I can't count anymore how often I heard Serana ask me, if she is a good pet, or tried to find a word that resembles what "EEEEE!" should sound like.) And since I heard that the new major release of DoM is around the corner, and I, in no way, am going to do all that again, I wrote a little python program, that sifts through the exportet csv file of exported voice lines, and changes them based on regex rules, so I can redo all that again against a new version of DoM when it comes and easily add or update the rulesets.

However, I am still in the early stages of testing and since my upload speed is not very fast, uploading the current version of the finished voiceset might take a while (and since TrollAutokill said it would be better just to wait for the next major release, I might just do that). I add the python script in the very rudimentary early version it is right now to the post, (with the ruleset i created for DoM) maybe someone else will want to test it or add to it.

 

 

voice_line_replacer_v0.2.py

Edited April 10, 2022 by Jasmine92

[Antiope_Apollonia]

 

26 minutes ago, Jasmine92 said:

should be quite easily done with python, which can read csv and write json files, tricky part would be the json structure, depending on how many dimensions it has. Do you have an example of the csv and the json? I'm no python crack either, but i could try.

There are example .json files at:

\Data\SKSE\Plugins\StorageUtilData\Diary Of Mine

Here's the spreadsheet:

DoM Custom Personalities.ods

 

23 minutes ago, Jasmine92 said:

Which lead to me trying out XVAsynth to redo the voices, because I use a never Version of DoM and i didn't want to miss out on spoken lines. I soon realized that, as good as it is and really fits well in some cases, there are others, that will let you think "hu? What did you say?" ("Thank you! I live to please you, mistress." being one of those lines, where the word live ends up being pronounced as in "the broadcast was live" not like "i live in Boston").

I've been meaning for a while now to look into whether xVASynth can read IPA.  Is the only option to sort of brute-force it into pronouncing things the way you want by mangling English spelling?  Or can you use IPA transcriptions to feed it precise pronunciations?  Because I can transcribe IPA fairly rapidly.

[TrollAutokill]

1 hour ago, sidfu1 said:

so troll what do you think we shoul do for uniques that when enslaved do not have their armor in their iventory to remove?

Use rip-off clothes, it will remove their outfit too. Then since they are unique you can use the naked outfit from PAHE safely.

[TrollAutokill]

1 hour ago, Jasmine92 said:

 

 

should be quite easily done with python, which can read csv and write json files, tricky part would be the json structure, depending on how many dimensions it has. Do you have an example of the csv and the json? I'm no python crack either, but i could try.

 

You timing however is quite intriguing:

right now I'm overhauling my complete modlist for a new playthrough, since many things are new, or got updates. (DoM is one of them). So i tested the 2.7.5 build and as mentioned before, got hooked on that mod. (damn you @TrollAutokill for stealing so much of my time! ) Testplaying and reading this thread lead me to @reedeonry's voicepack over here, which boosted my immersion even further. Which lead to me trying out XVAsynth to redo the voices, because I use a never Version of DoM and i didn't want to miss out on spoken lines. I soon realized that, as good as it is and really fits well in some cases, there are others, that will let you think "hu? What did you say?" ("Thank you! I live to please you, mistress." being one of those lines, where the word live ends up being pronounced as in "the broadcast was live" not like "i live in Boston"). And before I new it i was sifting through a couple hundred unique lines of DoM, pasting them into xvasynth and changing the sentence to make them sound better. (I can't count anymore how often I heard Serana ask me, if she is a good pet, or tried to find a word that resembles what "EEEEE!" should sound like.) And since I heard that the new major release of DoM is around the corner, and I, in no way, am going to do all that again, I wrote a little python program, that sifts through the exportet csv file of exported voice lines, and changes them based on regex rules, so I can redo all that again against a new version of DoM when it comes and easily add or update the rulesets.

However, I am still in the early stages of testing and since my upload speed is not very fast, uploading the current version of the finished voiceset might take a while (and since TrollAutokill said it would be better just to wait for the next major release, I might just do that). I add the python script in the very rudimentary early version it is right now to the post, (with the ruleset i created for DoM) maybe someone else will want to test it or add to it.

 

 

voice_line_replacer_v0.2.py 11.06 kB · 1 download

Another solution would be to adjust the lines in the plugin. I will look at your suggestions in the python, file. I will never stoop to code python, but I can read it...

Edited April 10, 2022 by TrollAutokill

[TrollAutokill]

9 minutes ago, Antiope_Apollonia said:

 

There are example .json files at:

\Data\SKSE\Plugins\StorageUtilData\Diary Of Mine

Here's the spreadsheet:

DoM Custom Personalities.ods 29.35 kB · 0 downloads

 

I've been meaning for a while now to look into whether xVASynth can read IPA.  Is the only option to sort of brute-force it into pronouncing things the way you want by mangling English spelling?  Or can you use IPA transcriptions to feed it precise pronunciations?  Because I can transcribe IPA fairly rapidly.

xVAsynth has a dictionary on how to pronounce certain words. But not all voices can use it. So some words have to be changed to the correct pronunciation. That was the case for Kimli when she talks about her wages, she said "waiggs", so I had to change it to "wadjes" in xvasynth.

[Jasmine92]

1 minute ago, Antiope_Apollonia said:

 

There are example .json files at:

\Data\SKSE\Plugins\StorageUtilData\Diary Of Mine

Here's the spreadsheet:

DoM Custom Personalities.ods 29.35 kB · 0 downloads

 

I've been meaning for a while now to look into whether xVASynth can read IPA.  Is the only option to sort of brute-force it into pronouncing things the way you want by mangling English spelling?  Or can you use IPA transcriptions to feed it precise pronunciations?  Because I can transcribe IPA fairly rapidly.

I will take a look.

 

It understands the ARPAbet and there are already dictionarys available (at least for english). There is also one for Elderscrolls proper names. However, not all voicepacks work with the new AI Model that understands ARPAbet (they are get constantly updated though). There are converters from IPA to ARPAbet, don't know how good they are though.

 

The bigger problem however is the melody of the sentence and which words get "highlighted" by a rising volume or pitch. "THIS is my house?" vs "This is MY house!" etc. One can control it to a certain degree by adding spacing, commas, or exclamation marks which quite often really improve upon the flow. For example just synthesizing the line "You're a piece of crap! Sorry! No please! Sorry! I didn't mean it!" does not really sound right, because it will all sound very even toned with no pausing, I ended up changing it to "You're a piece of crap! ? ? Sorry! No please! Sorry! ? ? I didn't mean it!" which looks funny, but the added silence and pacing that come from the questions marks helps a lot in conveying that the NPC did something out of reflex and then realizes "i fucked up"

[Jasmine92]

26 minutes ago, TrollAutokill said:

Use rip-off clothes, it will remove their outfit too. Then since they are unique you can use the naked outfit from PAHE safely.

Reading this bring me to the question: what is the current best practice when it comes to avoiding the naked bandid bug? (short of not enslabing bandids...) Since a saw a few new options in the PAH menu and now you are mentioning something that sounds like DoM could also play a role in it

 

16 minutes ago, TrollAutokill said:

xVAsynth has a dictionary on how to pronounce certain words. But not all voices can use it. So some words have to be changed to the correct pronunciation. That was the case for Kimli when she talks about her wages, she said "waiggs", so I had to change it to "wadjes" in xvasynth.

That's one of those cases I thought a set of rules would be good to have. Should add that to my ruleset. btw. Where do I have to go to start the Kimli questline?

 

 

21 minutes ago, TrollAutokill said:

Another solution would be to adjust the lines in the plugin. I will look at your suggestions in the python, file. I will never stoop to code python, but I can read it...

As I said, some sentences would have to be changed to something that looks just wrong as a subtitle. However a help would be to to unify noises like "EEEEE! Ahhhh! IIIEE and so on, so that regex rules can easily change them into something that will make xvasynth produce a sound that sounds like them. And if @Antiope_Apollonia might be able to create some IPA or ARPABet words that sounds even better, we are set to go! (Or you now, you could just moan and scream into a mike and give me the audio files )

Edited April 10, 2022 by Jasmine92

[Jasmine92]

23 hours ago, Antiope_Apollonia said:

So, I have finally finished creating the custom personalities for all the NPCs in Whiterun, as I promised was coming months ago.

 

I've tried to make them as lore-friendly as possible, relying on what can be learnt from their dialogues, lifestyle, profession, relationships to other NPCs, etc.—you don't want to know how much time I've spent staring at the wiki pages of the more obscure NPCs, digging for insights, lol.  I've tried both to ensure a realistic distribution of traits across the population as a whole, and to make the individual personalities themselves as realistic as possible for the most part.  I do have reasonable expertise in psychometrics, as my professional life touches on this academic area extensively, so I'm reasonably confident in my work here.  I've taken a little bit of liberty to give "better" personalities to fan-favourite NPCs and "worse" ones to reviled characters, but only a little bit.  I've prioritised lore-friendliness and realism first and foremost.  

 

Note that I haven't done anything with the children for now; I may still do in the future.

 

But...

They're all still in my spreadsheet I've been working from.  I still need to get them into the .json format that DoM reads.  Which is pretty daunting in terms of the tedium and time involved.  So I'm wondering if anyone here with some programming skills would be able to work up a simple utility to do that relatively easily?  My ignorance of coding is pretty high, but I wouldn't think that'd be hard to do for someone with the relevant expertise.  If I had such a utility available, it'd also probably help encourage me to keep working on this; I'm hoping to eventually do all vanilla NPCs, but that's going to be a while.  If anyone is able to create such a tool, I'd be grateful for the data entry pain you'd be saving me, and I'll have more time available to work on the NPC enslavement quests.

 

I just need to be able to convert rows of a spreadsheet with:

Name, Location, Gender, Honest, Calm, Lively, Kind, Thorough, Open, Wilful, Tough, Sensual, Submissive, Bold, Smart, BaseID

... into the .json format found in \Data\SKSE\Plugins\StorageUtilData\Diary Of Mine

 

When they're done, I'll pack them up in an archive that can be installed with a mod manager, and Troll can stick them on the download page if he wishes.

 

I wrote a little python script, called it "The Personalitizer" and look what it created: (see attached File), however only 69 out of the 77 entry's survived because I cleansed everything with a NULL Value. The JSON also has values for trauma, virgin(oral, anal, vagina, same-sex). I took the liberty and made them all virgins for the time being. You could add some rows to your spreadsheet for those too (only use int values for them). And since I didn't now what trauma does I also set it to 1 atm.

 

 

3 hours ago, Jasmine92 said:

 

 

should be quite easily done with python, which can read csv and write json files, tricky part would be the json structure, depending on how many dimensions it has. Do you have an example of the csv and the json? I'm no python crack either, but i could try.

 

You timing however is quite intriguing:

right now I'm overhauling my complete modlist for a new playthrough, since many things are new, or got updates. (DoM is one of them). So i tested the 2.7.5 build and as mentioned before, got hooked on that mod. (damn you @TrollAutokill for stealing so much of my time! ) Testplaying and reading this thread lead me to @reedeonry's voicepack over here, which boosted my immersion even further. Which lead to me trying out XVAsynth to redo the voices, because I use a never Version of DoM and i didn't want to miss out on spoken lines. I soon realized that, as good as it is and really fits well in some cases, there are others, that will let you think "hu? What did you say?" ("Thank you! I live to please you, mistress." being one of those lines, where the word live ends up being pronounced as in "the broadcast was live" not like "i live in Boston"). And before I new it i was sifting through a couple hundred unique lines of DoM, pasting them into xvasynth and changing the sentence to make them sound better. (I can't count anymore how often I heard Serana ask me, if she is a good pet, or tried to find a word that resembles what "EEEEE!" should sound like.) And since I heard that the new major release of DoM is around the corner, and I, in no way, am going to do all that again, I wrote a little python program, that sifts through the exportet csv file of exported voice lines, and changes them based on regex rules, so I can redo all that again against a new version of DoM when it comes and easily add or update the rulesets.

However, I am still in the early stages of testing and since my upload speed is not very fast, uploading the current version of the finished voiceset might take a while (and since TrollAutokill said it would be better just to wait for the next major release, I might just do that). I add the python script in the very rudimentary early version it is right now to the post, (with the ruleset i created for DoM) maybe someone else will want to test it or add to it.

 

 

voice_line_replacer_v0.2.py 11.06 kB · 1 download

 

Well... there it is. If anyone has suggestions for better xvasynth patterns for different lines, I will add them to the rules so they will be implemented in all feature versions. Don't expect to much of version one, I already realized after creating it, that I have to update the rules so that only voices which support ARPAbet will get the "live not live" fix for example.

 

PAH - Diary of Mine v2.7.6 - Voicepack v1.zip

DoM - Personalities.zip

Edited April 10, 2022 by Jasmine92
removed Mega link, embeddet file instead

[Antiope_Apollonia]

1 hour ago, TrollAutokill said:

xVAsynth has a dictionary on how to pronounce certain words. But not all voices can use it. So some words have to be changed to the correct pronunciation. That was the case for Kimli when she talks about her wages, she said "waiggs", so I had to change it to "wadjes" in xvasynth.

So it doesn't have any ability to read this, then?

/ˈwɛɪdʒəz/

 

That's a shame.  The xVASynth people should really look at changing that.  The IPA is an infinitely more precise way to transcribe pronunciation; there's a reason it's one of the first things you learn in Linguistics studies, and you'll see it used in any good dictionary.

[Antiope_Apollonia]

32 minutes ago, Jasmine92 said:

DoM - Personalities.zip 31.22 kB · 0 downloads

Perfect!  Looks like you've managed it without a hitch—great work.  Is the tool something you can share so I can keep working on this project? 

 

@TrollAutokill Do you want to stick this on the download page?

[Antiope_Apollonia]

2 hours ago, Jasmine92 said:

It understands the ARPAbet and there are already dictionarys available (at least for english). There is also one for Elderscrolls proper names. However, not all voicepacks work with the new AI Model that understands ARPAbet (they are get constantly updated though). There are converters from IPA to ARPAbet, don't know how good they are though.

I have no familiarity with ARPAbet, but from looking at its Wikipedia page, it looks like just a one-to-one mapping of IPA into characters on a standard American keyboard, so converters will probably do just fine as long as they can read the IPA diacritics without issue.  It looks like ARPAbet handles stress in a rather bizarre fashion, so I'm less confident that converters will handle that well, but that wouldn't be that awful to convert by hand.  

 

I'm not looking to learn a whole new alphabet—I already use IPA in my day-to-day—but if IPA–ARPAbet conversion works well enough, then I can do IPA transcriptions if you want to send me specific strings, or you can pull IPA off of Wiktionary, and I can proofread your work (pronouncing a word in isolation isn't the same as pronouncing it in context in a stress-timed language like English).