What’s new in SG Com 5.1?

 

We've released the latest update to SG Com, our real-time animation system that converts audio into facial animation with a latency of only 50 milliseconds.

Version 5.1 includes:

  • Breath animation

  • Improved nonverbal movement

  • Laughter auto mode & laughter animation

  • InsertExpression function

  • Status update types

  • API changes

Breath animation

SG Com now animates breath. The only requirement to use this feature is to have the Chest Breath muscle defined in your character (pictured below).

Pictured: Adding a Chest Breath muscle enables breath animation to be driven by audio input.

Continuous breathing motion is generated by a biological model constrained by the audio analysis. The breaths are synched to audible inhales and exhales, and to the pauses in the speech. The intensity and rapidity of breath are determined by the vocal performance.

Below are two examples of automatic breath animation: panting breath, and breath mixed with effort and speech.

Improved nonverbal movement

In SG Com 5.1 the dynamic muscle models governing real-time nonverbal behavior have been improved. The muscles have lower response times resulting in more interactive behavior. Movements are also more natural and less glitchy.

Laughter auto mode & laughter animation

As part of our mission to automate animation for all of the vocal sounds a human can make, SG Com 5.1 adds a new auto mode: Laugh. This is triggered whenever laughter is detected in the audio. The behavior of the new auto mode is defined by mapping it to a behavior mode with laugh-like expressions. Check out the video below of real-time laughter in action (0:24).

InsertExpression function

The InsertExpression function gives users more direct control over nonverbal expressions. Normally, SG Com fully automates the choice, timing and scale of expressions based on the audio. The new function allows users to manually trigger expressions, as well as specifying:

  • exactly which expression to trigger

  • the exact duration of the expression

  • the exact scale of the expression

Status update types

To get more metadata out of SG Com, we have added two new status types - ‘breath changed’ and ‘auto mode changed’.

  • ‘Breath changed’ is triggered whenever breath changes from inhale to exhale or vice-versa.

  • ‘Auto mode changed’ is triggered whenever a new auto mode is detected.

API changes

Finally, in the SG_COM_Modifier enum, the magnitude and speed modifiers have been divided into more specific modifiers giving the user more control over the magnitude and speed of various motions. It allows users to individually:

  • Adjust the magnitude of speech motion

  • Adjust the magnitude of nonverbal motion

  • Adjust the magnitude of breath motion

  • Adjust the speed of nonverbal motion

  • Adjust the speed of breath motion

Try SG Com now

Overall, the new features in SG Com 5.1 enhance realism and user control to improve real-time animation. Breath animation synchronized with vocal performance adds lifelike subtlety, while improved nonverbal motion ensures smoother, more natural interactions.

These advancements mark a significant step forward in making real-time animated performances more realistic and engaging.

Get in contact with our team to try out SG Com 5.1.

Previous
Previous

Speech Graphics in 2024

Next
Next

The Journey of Speech Graphics and Rapport