Cheyron Posted June 4 Share Posted June 4 On 6/2/2024 at 1:18 PM, DirebearCoat said: I get you, but just for your information Microsoft bought Bethesda.... They aren't nice like independent Bethesda was. Just saying... EDIT: And no, I don't work for Microsoft, either. don't be afraid to live your life don't let billionaires intimidate you Link to comment Share on other sites More sharing options...
PeterMartyr Posted June 8 Share Posted June 8 @Cheyron the nexus will never give you a cease & decease or a take down notice either, they just delete the mod with out warning, either with no explanation or a formal professional message stating why, with a potential ban strike depending on the site terms and conditions or the user agreement... which I have never read. but pretty sure it cover actual real-life copyright infringement, nevertheless, that could harm their bottom line. It is them you gotta worry about.. mate, and you cannot hide or avoid that if the mod is still there, you are OK, and it is not an issue, for the moment @ READERS EDIT there a mod that uses an actor voice, they used all the movies the actor appeared in and splice it all together to create new dialogue, it is still here, so AI just a smarter way of doing that , I do not think it is ethical on a personal level, but that just my opinion 1 Link to comment Share on other sites More sharing options...
cloudedtruth Posted June 13 Share Posted June 13 Cutting, splicing, and reusing dialogue does take quite a long time to do. I agree that the usage of tools can be extremely handy to create new dialogue in a much more simplified manner. The only issue I have with AI generated voices is that it's often still possible to tell it's not a real human doing the voicing. They've come a long way but human voice inflections mid sentence or even word is harder to recreate, as well as the noises people make naturally when they talk. Lastly, it's hard to accurately recreate emotion in a voice especially when it changes quickly or fluctuates some mid-sentence. Link to comment Share on other sites More sharing options...
scorrp10 Posted June 13 Share Posted June 13 12 hours ago, cloudedtruth said: Cutting, splicing, and reusing dialogue does take quite a long time to do. I agree that the usage of tools can be extremely handy to create new dialogue in a much more simplified manner. The only issue I have with AI generated voices is that it's often still possible to tell it's not a real human doing the voicing. They've come a long way but human voice inflections mid sentence or even word is harder to recreate, as well as the noises people make naturally when they talk. Lastly, it's hard to accurately recreate emotion in a voice especially when it changes quickly or fluctuates some mid-sentence. Actually, creating a more or less believable AI-generated voice can be quite labor-intensive as well. First, you write the text with a lot of additional prompting. I.e. text: "And if you try anything funny - she added menacingly - I'll fry you to a crisp!" After the voice is generated, the 'she added menacingly' part get cut out. Also, I had to do several generations of each body of text with varying parameters, and then mix and match. Link to comment Share on other sites More sharing options...
anjenthedog Posted June 13 Share Posted June 13 16 hours ago, cloudedtruth said: Cutting, splicing, and reusing dialogue does take quite a long time to do. I agree that the usage of tools can be extremely handy to create new dialogue in a much more simplified manner. The only issue I have with AI generated voices is that it's often still possible to tell it's not a real human doing the voicing. They've come a long way but human voice inflections mid sentence or even word is harder to recreate, as well as the noises people make naturally when they talk. Lastly, it's hard to accurately recreate emotion in a voice especially when it changes quickly or fluctuates some mid-sentence. Yup. Damn hard. Spent probably on the order of 400-500 hours over the years fiddling with various TTS programs, and even when I use every trick in the book, it still comes out stilted. That lack of "humanity" has been evident in all consumer grade text-to-speech (TTS), no matter what it's renamed as, since its inception in the 1990s. There are tricks that can be performed with the source text, (eg purr-puss-fully misspelling words and adding various ctrl characters so they sound closer to how they should, altering both pronunciation and syllable inflection, but it's freaking tedious work and leaves the source text looking a bit like swiss cheese written by a monkey. The base problem is that the software can't interpret intent. Link to comment Share on other sites More sharing options...
Cheyron Posted June 13 Share Posted June 13 On 6/8/2024 at 4:41 AM, PeterMartyr said: @Cheyron the nexus will never give you a cease & decease or a take down notice either, they just delete the mod with out warning, either with no explanation or a formal professional message stating why, with a potential ban strike depending on the site terms and conditions or the user agreement... which I have never read. but pretty sure it cover actual real-life copyright infringement, nevertheless, that could harm their bottom line. It is them you gotta worry about.. mate, and you cannot hide or avoid that if the mod is still there, you are OK, and it is not an issue, for the moment @ READERS EDIT there a mod that uses an actor voice, they used all the movies the actor appeared in and splice it all together to create new dialogue, it is still here, so AI just a smarter way of doing that , I do not think it is ethical on a personal level, but that just my opinion I am specifically talking about using AI to mimic the voice actor. I have used xVASynth and i have seen mods using something called Elvenlabs (name?). If you need your mod to have voiced dialogue, then just do it. How is that unethical? Like if I want MaleEvenToned to say something in my mod, it's kinda lame if half the guys lines are subtitle only. If someone has subtitles disabled, will they even know he is talking? Definitely not going to be my voice lol. Why is that unethical? Do they own the sound of their voice? I am genuinely curious about that. Does everyone have a copyright claim on anyone who sounds like them?? It's all ridiculous. OP JUST MAKE THE MOD, DON'T WORRY ABOUT THE BILLIONAIRES AND THEIR GAMES!! If they remove your mod, just shrug your shoulders and move on. The worst that can happen is your mod gets removed and that's fine. You aren't hurting the livelihood of these voice actors, you are just making a free mod for people to enjoy. Link to comment Share on other sites More sharing options...
DirebearCoat Posted July 1 Author Share Posted July 1 A question. Would there be legal issues if a modder used AI to make the existing voice say existing lines, but in a whisper? Let's say you're in a Dwemer dungeon, and there are Falmer around. You want Lydia to carry a few things for you. Then she'd say her existing lines, but in a Whisper. Lydia whispers, "How may I serve you, my thane?" Would there be legal issues with doing that? New dialogue isn't being said, but instead of shouting - while in a cave full of bandits - the follower would whisper their replies... Link to comment Share on other sites More sharing options...
Recommended Posts