VC WIP Shine o' Vice

[Dominik_]

27 minutes ago, ArkianHue said:

I wonder how you guys did it?

Is it like normal animating in (3dsmax)? Or something new? Like a plugin or a tool?

It's a mocap, you need to have a special set of tools to do the animation in-real-life.

Edited June 24, 2020 by Dominik_

[Speccy]

3 hours ago, ArkianHue said:

I wonder how you guys did it?

Is it like normal animating in (3dsmax)? Or something new? Like a plugin or a tool?

Even though ilufir successfully ported an animation entirely done through mocap using an xbox kinect, the ones used at the moment are edited rockstar cutscenes from VC & SA. The camera can be changed, just as the actual animations and actors appearing in the scene. On top of that, there's CharAndAnimHelper which I haven't tried yet.

 

Edited June 24, 2020 by Speccy

[mister.t.mark12345]

6 hours ago, Speccy said:

Even though ilufir successfully ported an animation entirely done through mocap using an xbox kinect, the ones used at the moment are edited rockstar cutscenes from VC & SA. The camera can be changed, just as the actual animations and actors appearing in the scene. On top of that, there's CharAndAnimHelper which I haven't tried yet.

 

Can you edit tommy house?

[ahmedali]

what is the news about shine o vice

[deltaCJ]

8 hours ago, ahmedali said:

what is the news about shine o vice

Wait.

[Lucas 444]

I only wanna ask Speccy, in percentage value, how much is the mod done so far?

[jaydeedilla]

Not sure if it has been suggested before, but have you guys ever considered messing with speech synthesis? You could get believable voice clips of the Vice City actors speaking their respective character's lines - and get voices for the new characters as well. I think this would really enhance the entire narrative to such a higher level.

Edited July 12, 2020 by afxtwin

[Maverick512]

1 hour ago, afxtwin said:

Not sure if it has been suggested before, but have you guys ever considered messing with speech synthesis? You could get believable voice clips of the Vice City actors speaking their respective character's lines - and get voices for the new characters as well. I think this would really enhance the entire narrative to such a higher level.

They have talked about that before I think the best way is to get voice actors within this community.

[Guest]

1 hour ago, afxtwin said:

Not sure if it has been suggested before, but have you guys ever considered messing with speech synthesis? You could get believable voice clips of the Vice City actors speaking their respective character's lines - and get voices for the new characters as well. I think this would really enhance the entire narrative to such a higher level.

Bleh... it's one of two things:

1- You either get real voice actors.

                             OR

2- Make the characters silent.

 

~♡~ Speech synthesis is kinda... unrealistic, and you can't make voices believable and alive. Also, what do you mean by

:

"You could get believable voice clips of the Vice City actors speaking their respective character's lines -."

 

And as @Maverick512 said, voice actors could be from within these forums.

 

*** They (the community) provided voice actors for an old machinima/series called (PEDS) WAYYY back in 2008. So, it's possible to get real voice actors for projects like these and mods.

Edited July 12, 2020 by ArkianHue
Thought I explain some sh*t... 😛

[jaydeedilla]

Just now, ArkianHue said:

Bleh... it's one of two things:

1- You either get real voice actors.

                             OR

2- Make the characters silent.

 

~♡~ Speech synthesis is kinda... unrealistic, and you can't make voices believable and alive. Also, what do you mean by

:

"You could get believable voice clips of the Vice City actors speaking their respective character's lines -."

I disagree with you on that. With what's available now, you could totally make speech synthesis sound very convincing, especially with some post audio production. And I meant that the Vice City characters featured in this mod, such as Umberto, can deliver their lines with delivery similar to (in this case) Danny Trejo himself or any other VC actor/actress.

Edited July 12, 2020 by afxtwin

[Guest]

Just now, afxtwin said:

I disagree with you on that. With what's available now, you could totally make speech synthesis sound very convincing, especially with some audio post production. And I meant that the Vice City characters featured in this mod (such as Umberto) can deliver their lines with delivery similar to (in this case) Danny Trejo himself.

It could... but, sometimes you want the characters to deliver the lines in very specific way or a tone you have in mind. But doesn't speech synthesis need voice clips to provide accurate simulation of the voice? Sometimes you need to use a certain word or a sentence and there's not a single voice clip for the voice actor speaking it, it could sound off or weird. Maybe you're right, maybe not...

[Maverick512]

1 hour ago, ArkianHue said:

It could... but, sometimes you want the characters to deliver the lines in very specific way or a tone you have in mind. But doesn't speech synthesis need voice clips to provide accurate simulation of the voice? Sometimes you need to use a certain word or a sentence and there's not a single voice clip for the voice actor speaking it, it could sound off or weird. Maybe you're right, maybe not...

Deniska made a dub in 2008 using real actors and voice synthesizers for Anderius, but as you say, it's not realistic, If I could, I would do a voice collaboration but I don't have a microphone and my English is not good.

[deltaCJ]

DENISKa's mod is a prime example of why voice actors can make/break a mod/game, sometimes speech synthesis isn't really good, and sometimes throws you off more than anything. But he used the sci-fi setting to his advantage, usually aliens have the robotic voice, etc.

 

 

EDIT: Also I think what you speak of is the same thing those weird apps try to achieve in the "Sound like a Celebrity" aspect. I found those apps hilarious because of the robotic sound of Barack Obama, and Spongebob Squarepants. I'm pretty sure you'd need hours of audio, and a whole lot of time to perfect that audio because if you think you just plop the audio in a program, and then it comes out as a exported file, you're dead wrong.

 

The same way a Deep Fake needs HIGH QUALITY, and usually 2x the amount of footage you usually need for the amount of time the deep fake is happening. And that's usually!

 

Corridor Digital did a video on resurrecting 2Pac for a music video. First they needed a Stunt Double (For Audio Synthesis they actually might still need a voice actor...), and then they ravaged through interviews and anything else they can find, but since it was for a music video it had to look good. They had to find all HD versions of most interviews and footage where they can really see his face clearly for a good quality Deep Fake. The result was pretty good.

Here's a snippet of one of the comments:

 

Now audio synthesis may and probably is completely different. Dessa made a pretty realistic and cool Joe Rogan voice synthesizer which turned out pretty good, BUT this is a million dollar company with no "Download" section, and is definitely not letting some modders just using their software (not that it would be that easy of course.)   

 

EDIT2: After more digging Dessa actually released a SDK for development in more Deep Learning.

So if you'd like to mess around and try to mess around with the Atlas program, go ahead! You'd be taking one for the team

https://github.com/dessa-oss/atlas

https://github.com/JustusThies/NeuralVoicePuppetry

 

This stuff is actually pretty interesting, but check this out. This guy tried to FAKE HIMSELF and still had to use over 50 SAMPLES ! 

 

 

Edited July 14, 2020 by deltaCJ

[TommyVercetty]

When Shine O Vice Full Version

[Maverick512]

14 hours ago, deltaCJ said:

DENISKa's mod is a prime example of why voice actors can make/break a mod/game, sometimes speech synthesis isn't really good, and sometimes throws you off more than anything. But he used the sci-fi setting to his advantage, usually aliens have the robotic voice, etc.

 

 

EDIT: Also I think what you speak of is the same thing those weird apps try to achieve in the "Sound like a Celebrity" aspect. I found those apps hilarious because of the robotic sound of Barack Obama, and Spongebob Squarepants. I'm pretty sure you'd need hours of audio, and a whole lot of time to perfect that audio because if you think you just plop the audio in a program, and then it comes out as a exported file, you're dead wrong.

 

The same way a Deep Fake needs HIGH QUALITY, and usually 2x the amount of footage you usually need for the amount of time the deep fake is happening. And that's usually!

 

Corridor Digital did a video on resurrecting 2Pac for a music video. First they needed a Stunt Double (For Audio Synthesis they actually might still need a voice actor...), and then they ravaged through interviews and anything else they can find, but since it was for a music video it had to look good. They had to find all HD versions of most interviews and footage where they can really see his face clearly for a good quality Deep Fake. The result was pretty good.

Here's a snippet of one of the comments:

 

Now audio synthesis may and probably is completely different. Dessa made a pretty realistic and cool Joe Rogan voice synthesizer which turned out pretty good, BUT this is a million dollar company with no "Download" section, and is definitely not letting some modders just using their software (not that it would be that easy of course.)   

 

EDIT2: After more digging Dessa actually released a SDK for development in more Deep Learning.

So if you'd like to mess around and try to mess around with the Atlas program, go ahead! You'd be taking one for the team

https://github.com/dessa-oss/atlas

https://github.com/JustusThies/NeuralVoicePuppetry

 

This stuff is actually pretty interesting, but check this out. This guy tried to FAKE HIMSELF and still had to use over 50 SAMPLES ! 

 

 

Using Deepfake would be a better option than hiring the original voice actors, GTA Mushroomia is another good example of how voice actors can break a mod unfortunately I couldn't find a suitable gameplay to show

[Guest]

14 hours ago, deltaCJ said:

DENISKa's mod is a prime example of why voice actors can make/break a mod/game, sometimes speech synthesis isn't really good, and sometimes throws you off more than anything. But he used the sci-fi setting to his advantage, usually aliens have the robotic voice, etc.

 

 

EDIT: Also I think what you speak of is the same thing those weird apps try to achieve in the "Sound like a Celebrity" aspect. I found those apps hilarious because of the robotic sound of Barack Obama, and Spongebob Squarepants. I'm pretty sure you'd need hours of audio, and a whole lot of time to perfect that audio because if you think you just plop the audio in a program, and then it comes out as a exported file, you're dead wrong.

 

The same way a Deep Fake needs HIGH QUALITY, and usually 2x the amount of footage you usually need for the amount of time the deep fake is happening. And that's usually!

 

Corridor Digital did a video on resurrecting 2Pac for a music video. First they needed a Stunt Double (For Audio Synthesis they actually might still need a voice actor...), and then they ravaged through interviews and anything else they can find, but since it was for a music video it had to look good. They had to find all HD versions of most interviews and footage where they can really see his face clearly for a good quality Deep Fake. The result was pretty good.

Here's a snippet of one of the comments:

 

Now audio synthesis may and probably is completely different. Dessa made a pretty realistic and cool Joe Rogan voice synthesizer which turned out pretty good, BUT this is a million dollar company with no "Download" section, and is definitely not letting some modders just using their software (not that it would be that easy of course.)   

 

EDIT2: After more digging Dessa actually released a SDK for development in more Deep Learning.

So if you'd like to mess around and try to mess around with the Atlas program, go ahead! You'd be taking one for the team

https://github.com/dessa-oss/atlas

https://github.com/JustusThies/NeuralVoicePuppetry

 

This stuff is actually pretty interesting, but check this out. This guy tried to FAKE HIMSELF and still had to use over 50 SAMPLES ! 

 

 

 

I don't think this will be used for making character voice for the mod, right?

[NqkoJ]

17 hours ago, deltaCJ said:

DENISKa's mod is a prime example of why voice actors can make/break a mod/game, sometimes speech synthesis isn't really good, and sometimes throws you off more than anything. But he used the sci-fi setting to his advantage, usually aliens have the robotic voice, etc.

 

 

EDIT: Also I think what you speak of is the same thing those weird apps try to achieve in the "Sound like a Celebrity" aspect. I found those apps hilarious because of the robotic sound of Barack Obama, and Spongebob Squarepants. I'm pretty sure you'd need hours of audio, and a whole lot of time to perfect that audio because if you think you just plop the audio in a program, and then it comes out as a exported file, you're dead wrong.

 

The same way a Deep Fake needs HIGH QUALITY, and usually 2x the amount of footage you usually need for the amount of time the deep fake is happening. And that's usually!

 

Corridor Digital did a video on resurrecting 2Pac for a music video. First they needed a Stunt Double (For Audio Synthesis they actually might still need a voice actor...), and then they ravaged through interviews and anything else they can find, but since it was for a music video it had to look good. They had to find all HD versions of most interviews and footage where they can really see his face clearly for a good quality Deep Fake. The result was pretty good.

Here's a snippet of one of the comments:

 

Now audio synthesis may and probably is completely different. Dessa made a pretty realistic and cool Joe Rogan voice synthesizer which turned out pretty good, BUT this is a million dollar company with no "Download" section, and is definitely not letting some modders just using their software (not that it would be that easy of course.)   

 

EDIT2: After more digging Dessa actually released a SDK for development in more Deep Learning.

So if you'd like to mess around and try to mess around with the Atlas program, go ahead! You'd be taking one for the team

https://github.com/dessa-oss/atlas

https://github.com/JustusThies/NeuralVoicePuppetry

 

This stuff is actually pretty interesting, but check this out. This guy tried to FAKE HIMSELF and still had to use over 50 SAMPLES ! 

 

 

Ressurrecting 2pac?I was just listening to him right now while reading lol.

[deltaCJ]

Guys, no. This -in no way -is not what the devs are doing for the mod. Right now it isn't known, but I know for a fact that "Deep Faking" the dialogue is not happening.

Edited July 14, 2020 by deltaCJ

[BlackStallionDX]

22 hours ago, Maverick512 said:

Using Deepfake would be a better option than hiring the original voice actors, GTA Mushroomia is another good example of how voice actors can break a mod unfortunately I couldn't find a suitable gameplay to show

Only your personal IMO, mate. Mooshromania have good voice acting. It's fit for mod atmoshpere. Druggy voices is good for druggy mod.

So, without voiced cutscenes every mod looks like cheap work, like Hell on Wheels.

[Ash_735]

Back when I shared an earlier release of this mod a voice actor I know actually took notice of it and shared some interest, they're actually a proper voice actor too, mainly does anime but they also like doing side sh*t to keep busy, if you have a finished script for SOV I can see it ends up in their DM's and see if they can get a small team for it.

[deltaCJ]

Woah, Mushroomia was excellent in their voice acting, even going as far as getting a woman for the stripper. I mean it wasn't top quality, or top notch acting, but you still got the idea. Any ways, voice actors are the best way to go in this case. 

[Speccy]

 

ToMmY dEePfAkEs CONFIMRED?!????

Edited July 16, 2020 by Speccy

[TommyVercetty]

When shine o vice full version

[Guest]

On 7/14/2020 at 4:44 PM, TommyVercetty said:

When Shine O Vice Full Version

 

3 hours ago, TommyVercetty said:

When shine o vice full version

Dude, that's the second time!

 

It'll be released when the developers have some information for us about the release date here in the topic and when it's finally finished, it'll be released. Either Speccy or BEGINit idk, will tell. So don't ask again, please!

[ahmedali]

i see the shine o vice is dead

[Ash_735]

Just now, ahmedali said:

i see the shine o vice is dead

Because they've not released it fast enough for you?

[ahmedali]

exacly

[thalilmythos]

I'd reckon that there are more than enough samples of voice from tommy just in Vice City to make a decent sounding speech synthesis, if you look into more voicework by Ray Liotta, hell, you can even look for HQ sound on movies of him (shouldn't be too difficult) then you could even have just enough for a good "impersonation" i believe.

 

Is just a matter of opening your mind and trying it, because the content IS THERE.

Edited July 30, 2020 by thalilmythos

[deltaCJ]

Just now, ahmedali said:

exacly

Can you please f*ck off? 

[ahmedali]

what is problem with you i am ask why shine o vice doesn't have any news