When more than one audio input devices are equipped with your machine, you can choose one from them by a pop-up menu at the top in the window. Note that the devices plugged in after invocation of DT1 might not be listed in the menu. If you are using an external microphone directly connected to the sound input jack not through an amplifier, check the button labeled Raises the level of audio input by software.
The agents' behaviour is usually attracted mode where they are attracted by difference of video-input images in previous and current frames. The other modes are escaping mode where they escape from the same features, and ignoring mode where they ignore any features from video input. You can change the mode by your voice or any other sounds.
The trigger of mode change depends on the volume of sound. The sliders labelled Level#1 and Level#2 are to indicate the threshold levels of sound volume. Sound louder than level#1 is recognised as loud voice, sound smaller than level#1 and louder than level#2 is recognised as middle voice, and sound more quiet than level#2 is recognised as silent voice.
You can set-up how to change the mode by operating controls in the box entitled Rules to change mode placed under the sliders. In default, loud voice at attracted mode leads escaping, ten seconds of silent time during escaping mode leads ignoring mode, five seconds of silent time moves it back to attracted mode, and middle voice at escaping and ignoring mode also leads attracted mode.
When the drawing mode of agents is opaque, the method to mix the video-input image is different according to the behaviour mode as follows. It is drawn in usual way in attracted mode. It draws the image by inverted colour in escaping mode. It mixes nothing in ignoring mode.
| previous | top | next | © 2004-2007 Tatsuo Unemi & Daniel Bisig |