- LSL 97.7%
- JavaScript 2.3%
| Allophone | ||
| Assets | ||
| Objects | ||
| scripts | ||
| .gitignore | ||
| CHANGELOG.md | ||
| LICENSE.md | ||
| README.md | ||
Synthesized Voice Box
Emulate speech within Second Life
Background
This was a small research project to emulate speech within Second Life based. Just under sixty 8-bit sounds from the General Instrument SPO-256-AL2 were strung together to form words. With the existing limitations, speech was not fluid. The project was donated as open-source to Second Life's third anniversary and was in the SL3B vault at the time. As I became familiar with various techniques to push Second Life platform to its limits, I found ways to work around some of the limitations and improved the voice box to produce speech that could be understood.
Visual Changes
See the original (1.0) and latest (1.6) versions in a video: Voice Emulation with SPO-256-AL2 sounds
1.0 - 1.3
1.4 and later
Allophones
The allophones are parts of speach that are stitched together to form words and pauses. There are a total of three sets in this repo - sounds, sam, and lewie. I'm uncertain of the difference between "sounds" and "sam". It may have just been 8-bit vs 16 bit, or a format of the audio before uploading into Second Life. However, "Lewie" is my own voice trying out all of the sounds in a monotone voice, and then normalized before converting to 8-bit PCM mono. The original sounds are unable to be retrieved, as the Second Life viewer only receives OGG formatted sounds in a higher bitrate.
Allophone - SPO 1.0
The object, Allophone - SPO 1.0 appears to only be an object to offer the allphone sounds from the sounds folder. It's included for archival purposes.
Forked
After creating this repo with the various versions, I searched my GitHub account and found version 1.2 already had it's own repo at Eliza-Chatbot-LSL.

