- Voice Engine
- Advanced Configuration
Select a Character Model to assign to your Avatar.
- Click the “Gallery” button to see images of the currently available avatars. If you need additional customizations to an avatar, contact email@example.com.
- Select the desired model.
- Select Close.
Linking an avatar allows multiple avatars to be displayed in the same window. To link an avatar, first create at least two avatars, then pick an avatar from the list below. Click the X to remove links. This option is only available for Super Admins.
Select an Amazon Polly or Azure voice for your Avatar. Codebaby provides access to large library of male and female voices in English as well as many other languages.
- Select the Voice drop-down menu.
- Select the desired voice.
Play Voice #
Play Voice plays a sample of the currently select voice so that you can hear the sound qualities.
Voice Engine #
Some voices have a “neural” version, which is closer to human speech but cannot be as customized by SSML tags as standard voices. We recommend starting with neural voices when available. The display in this field changes between Standard or Neural, depending on the currently selected voice.
Azure voices with a neural version from ChatGPT #
Microsoft currently provides neural versions of Nuance Nina and Azure Speech Services’ neural voices, both US and UK English voices. This includes formant-synth hybrid neural voices, as well as deep neural text to speech (TTS) voices coming soon. Additionally, Microsoft plans to add more neural voices to its selection in the future.
Advanced Configuration #
This advanced configuration can be used to make very specific changes to your avatar and should only be used if you are advised to do so by a CodeBaby team member.
Cameras and Loading Images #
The simplest method to edit the values for the camera angle is to go to Video Generator and alter the view in the avatar display area. Use the mouse scroll wheel to zoom in and out, the left mouse button to rotate the view and the right mouse button to pan. When the display is set to the desired view, go back to Character/Cameras and Loading Images and copy the alpha, beta, radius and target values from the videoGenerator section and paste into the section for fullbody or headShoulders and refresh the view.
The Lights settings provide Position and Intensity values for three light sources.
Idle Animations #
Idle Animations let you define a sequence of animations that are used when the avatar is waiting and not speaking. You can enter the Codebaby markup language code for the gestures to be used in the idle animations. The weight values are all relative to each other, with the lower numbers indicating that the gesture should be used less, higher numbers playing the gestures more.
Talking Idle Animations #
Idle Animations let you define a sequence of animations that are used when there are no specific gestures coded into text responses. You can enter the Codebaby markup language code for the gestures to be used in the idle animations.
Lip Sync Mapping #
Lip Sync Mapping is used to control the mouth movement when the avatar is speaking. It works by defining the amount of movement for each blendshape (between 0-1) on an Avatar to create the correct mouth shape for each syllable of a word.
Here is an example of blendshapes and their values for the letter E.
In this example “E” creates a sound like the e in ‘dress’ and to create that sound the avatar will create a mouth shape using a value of .22 on the “Open” blendshape and 1 on the “Wide” blendshape. By adjusting the value on each blendshape you can change what shape the mouth makes. For example, if the “Open” value is changed from .22 to a value of .5 the mouth will be open wider.
A table that shows what sound each Viseme character makes can be found here.
Grab Image #
Grab Image allows you to save an image of the current page.
Grab Again #
Grab Again recaptures the screen with an update gesture.
Save saves the file as a PNG directly to the download folder.
Use As Loading Image for Active Camera #
Assigns the recently grabbed image as the loading image.
Cancel cancels the Grab Image function and closes the dialog.
Upload Loading Image #
Upload Loading Image let’s you define a background image for the avatar. Once the image is defined, the file name is included in the Cameras and Loading Images dialog.
Import Scene #
Scenes can be used to shared data between projects.
Import Scene imports the scene from a JSON file. The scene includes the model, voice selection, camera, lights, viseme and gesture maps.
Export Scene #
Export Scene exports the scene to a JSON file.