Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

DarkBASIC Professional Discussion / Dark Voices

Author
Message
SteverEno
19
Years of Service
User Offline
Joined: 3rd Mar 2006
Location:
Posted: 3rd Mar 2006 15:30
Link
I made a comparatively crude, non-automatic program for generating lip sync data a while back called Lip Sync Lab. That's what I used to do the lip sync for Uncle Weevy. If you want a free but labor-intensive way to generate mouth shape data for a handful of short samples, you can get it here: http://www.stevetiffany.com/darkbasic.html

Uncle Weevy made up sentences by stringing together short samples of individual words, so Lip Sync Lab had a functional sample length limit of two or three seconds. For anything beyond that, Dark Voices sounds like a great tool. The amount of time it will save me in the future is staggering.

CJB-- I once tried your idea of creating speech by playing a string of sampled phonemes. No one could understand it but me, and the only reason I could understand it is that I knew what it was supposed to be saying. Cool idea, though.
Back to top
Profile PM Website
CJB
Valued Member
21
Years of Service
User Offline
Joined: 10th Feb 2004
Location: Essex, UK
Posted: 3rd Mar 2006 22:06
Link
I guess the same process could be used as a step toward speech recognition too?

Back to top
Profile PM Email Website
SteverEno
19
Years of Service
User Offline
Joined: 3rd Mar 2006
Location:
Posted: 4th Mar 2006 14:27
Link
If Dark Voices is able to "read lips," it's halfway to speech recognition already. Going the rest of the way has been an ongoing challenge for the large companies dedicated to solving it, though.
Back to top
Profile PM Website
BatVink
Moderator
22
Years of Service
User Offline
Joined: 4th Apr 2003
Location: Gods own County, UK
Posted: 4th Mar 2006 22:35
Link
Quote: "I made a comparatively crude, non-automatic program for generating lip sync data a while back called Lip Sync Lab"


It's in my folder of useful tools I've never needed to use it, but that's down to my projects, not the tool!

Back to top
Profile PM Email
MR elevetor
19
Years of Service
User Offline
Joined: 4th Sep 2005
Location: you wouldn\'t belive me if I told you
Posted: 6th Mar 2006 14:47
Link
this is tha best program! (I still want you C++)

So hard to choose!
Fpsc or DBPRO or Source or Havok or C++-....maybe its C++....
Back to top
Profile PM
Lurtz Tha Uruk Hai
19
Years of Service
User Offline
Joined: 22nd Feb 2006
Location: Under the Mud beneath Orthanc-Isengard
Posted: 6th Mar 2006 23:16
Link
I really hope this tool will be made for fpsc to
Back to top
Profile PM Email
QuothTheRaven
22
Years of Service
User Offline
Joined: 2nd Oct 2002
Location: United States
Posted: 6th Mar 2006 23:19
Link
Quote: "---------------------------
Error
---------------------------
Failed to load DLL (4: DBProCameraDebug.dll)
---------------------------
OK
---------------------------"


Back to top
Profile PM
BatVink
Moderator
22
Years of Service
User Offline
Joined: 4th Apr 2003
Location: Gods own County, UK
Posted: 6th Mar 2006 23:23
Link
Quote: "Failed to load DLL (4: DBProCameraDebug.dll)"


You need DirectX 9.0c December 2005

Back to top
Profile PM Email
D Ogre
21
Years of Service
User Offline
Joined: 19th Nov 2003
Location:
Posted: 7th Mar 2006 01:01 Edited at: 7th Mar 2006 02:07
Link
How much technical documentation will be accompanying this product? I am more concerned with setting up a custom 3D character to use it. I have a project that I shelved years ago that could benefit from this.

<EDIT>

Quote: "It doesn't rely on anything. DarkVOICES is raw data which you can apply to anything. You could use the data in other apps too, it's not tied to DB Pro or any particular model or method."


Hmmm... You still need to know how the data is applied to the model. There needs to be some sort of method in the process. The raw data by itself isn't going to automatically make any 3D object talk. You need to know the what, why, where, and when to apply the information correctly.

I am wondering how helpful the source code and documentation is for helping the programmer use this in their own projects that's all.

-----------------------------------------------------------------------

As far as voice recognition goes... I remember back in the eighties there was a hardware plugin for the Commodore 64 called the COVOX VOICE MASTER that would let you digitize and play back your own voice as well as do voice recognition. It came with software and a BASIC wedge extension that allowed you to program your own software with it. This little gadget worked reasonably well. If they did it back then, why is it so hard to come up with something simple that works well today?
Back to top
Profile PM
BatVink
Moderator
22
Years of Service
User Offline
Joined: 4th Apr 2003
Location: Gods own County, UK
Posted: 7th Mar 2006 10:43
Link
Quote: "The raw data by itself isn't going to automatically make any 3D object talk"


Take a look at the newsletter tutorial, it shows one method of applying the lipsync data to a sprite-based model. You can use this - the method itself is very similar to the code in the DarkVOICES tool. IIRC, the only thing I changed was reading the file in, applying it in realtime is identical.

Back to top
Profile PM Email
LeeBamber
TGC Lead Developer
25
Years of Service
User Offline
Joined: 21st Jan 2000
Location: England
Posted: 8th Mar 2006 01:40
Link
The demo version contains the full version manual, which details all the technical information you need regarding integration with your project. If you want to create your own widget from scratch, check out the SAPI SDK technology microsoft release for free (also the heart of DarkVOICES). This SDK also deals with full voice recognition too, if you feel daring!

"Small, smart, and running around the legs of dinosaurs to find enough food to survive, bedroom programmers aren't extinct after all "
Back to top
Profile PM Email Website
CJB
Valued Member
21
Years of Service
User Offline
Joined: 10th Feb 2004
Location: Essex, UK
Posted: 9th Mar 2006 14:17
Link
@D Ogre> Hehe... COVOX VOICE MASTERterterterter... I used to have one of those! The voice recognition was terrible!

Good 'blast from the past' moment!

Back to top
Profile PM Email Website
D Ogre
21
Years of Service
User Offline
Joined: 19th Nov 2003
Location:
Posted: 10th Mar 2006 18:22
Link
Thanks for the info guys...

CJB, I never really had any problems with the voice recognition with CVM. If I remember correctly, you could fine tune the ACCEPT level of error. I know it wasn't 100%, but for the time it was okay.
Back to top
Profile PM
JulesD
20
Years of Service
User Offline
Joined: 7th Feb 2005
Location:
Posted: 29th Apr 2006 08:29
Link
Just wondering.

If I had a game where a Character said something using a sound file but did not use Dark Voices on the Character. Then I took the same game and added the Dark Voices, on average how many frames per second am I going to loose?
Back to top
Profile PM Website
Back to top

Login to post a reply

Server time is: 2025-08-08 12:52:34
Your offset time is: 2025-08-08 12:52:34