1
00:00:00,000 --> 00:00:05,280
Voice conversation from the chat. Voice
conversation requires a microphone, audio

2
00:00:05,360 --> 00:00:09,720
playback, and a stable connection. The
experience may vary depending on the

3
00:00:09,760 --> 00:00:14,640
browser and sound output. Step 1. Open the
voice conversation.

4
00:00:14,880 --> 00:00:19,760
In Ask, there are different controls:
Dictate question fills the text field.

5
00:00:20,240 --> 00:00:23,200
Quick voice responds without consulting
your material.

6
00:00:23,760 --> 00:00:27,360
Intelligent voice uses the library,
internet, and memory.

7
00:00:27,920 --> 00:00:32,800
Choose the mode according to your needs;
the indicated shortcuts are Alt plus M

8
00:00:32,880 --> 00:00:38,080
to dictate, Alt plus V for quick voice, and
Alt plus S for intelligent

9
00:00:38,160 --> 00:00:42,360
voice. Step 2. Register your voice if
requested.

10
00:00:42,560 --> 00:00:46,800
Intelligent voice may open a guided
registration if you do not yet have a voice

11
00:00:46,880 --> 00:00:52,240
associated. Follow the five prompts:
record, stop, and move to the next

12
00:00:52,360 --> 00:00:57,280
one. Use your own voice in a quiet
environment and wait for confirmation.

13
00:00:57,360 --> 00:01:01,560
This registration alone does not guarantee
that any other voice or echo will be

14
00:01:01,600 --> 00:01:05,840
ignored. Step 3. Speak and follow the
status.

15
00:01:05,960 --> 00:01:10,120
Wait for the listening indication, speak
naturally, and leave a pause to end the

16
00:01:10,160 --> 00:01:14,640
turn. Follow the listening, processing, and
response statuses.

17
00:01:14,680 --> 00:01:18,480
The dictation that fills a field and a
continuous conversation are different

18
00:01:18,560 --> 00:01:22,800
controls. Step 4. Interrupt with an
explicit control.

19
00:01:23,200 --> 00:01:27,600
If voice detection does not cut off the
response when you want, use the available

20
00:01:27,680 --> 00:01:32,640
button to stop or interrupt. Avoid opening
two voice conversations or two

21
00:01:32,720 --> 00:01:36,760
captures at the same time. Close the dialog
when finished to release the

22
00:01:36,800 --> 00:01:40,400
microphone. Step 5. Adjust the environment.

23
00:01:40,880 --> 00:01:44,800
Using headphones usually makes it easier to
prevent the microphone from picking

24
00:01:44,880 --> 00:01:49,520
up the assistant's voice again. The browser
requests echo cancellation and noise

25
00:01:49,600 --> 00:01:53,120
suppression, but their effectiveness
depends on the device.

26
00:01:53,160 --> 00:01:57,200
A synthetic test does not demonstrate how
it works in a noisy room.

27
00:01:57,240 --> 00:02:01,040
The enablement and transport of the chat
voice and the tutor voice may be

28
00:02:01,320 --> 00:02:06,080
different. The tutor chapter explains its
pilot continuous conversation; one

29
00:02:06,160 --> 00:02:09,360
should not assume that enabling one
activates the other.
