Using virtual audio cable for synthetic speech demonstrations


One of the challenges in accessibility is showing someone how technologies such as screen readers work. This challenge is exacerbated when the recipient is blind. How do you transmit the audio from your computer over an audio chat client such as Google Hangouts, Skype or Team Talk? Many people use a mixer which is dedicated hardware for this kind of work. You can add as many channels as the mixer supports so you can combine multiple sound sources. Virtual audio cable allows you to simulate patch cords among different sources. You can speak into a microphone, have your screen reader talking and broadcast the entire thing via Skype without any echo.
This article applies to version 4.15 of Virtual Audio Cable.

Setting up virtual audio cable

The first thing you need to do is to determine how many cables you need. Think of how many patch cords you need to keep in your toolkit. You then need to decide how many repeaters you need. Each repeater represents a connection from one sound source to another. You can always add more. The way to think about this is to calculate the number of sound out sources. You then add one more repeater to feed to your final output such as Skype. Let us take an example.
You want to demonstrate the working of the track changes feature with a screen reader in Microsoft Word to a client. Let us consider the sound sources. You have your voice which will be sent through a microphone into Skype. You have the screen reader that you need to send through Skype. Microsoft Office does not interact with the sound card. I am assuming that you have not put any audio animation.
In terms of repeaters, you will need one repeater for the microphone, one repeater for the screen reader and the final one to connect the 2 sound cables.
Virtual audio Cable consists of a few small programs that you launch individually. In most cases, you use the item called “Audio Repeater (MME)”
When you launch the program, you will be placed in a dialogue that has several controls. Be warned, at least in this version of Virtual Audio Cable, the labels of the controls will not be spoken.
Set the input sound. In this case, you will choose the microphone device. This is a combo box and usually, your screen reader should speak the current selection. This on my computer is “Microsoft Sound Mapper” The label for this combo box is “wave in.”
Tab once to the next combo box labeled “wave out” where nothing is selected. Here, you need to set the output device, which in our case is line2.
Tab across the dialogue box until you hear a value of “500.” This is the value of audio buffers which you must reduce to 200 or lower to avoid lag.
Hit the “start button” to start the repeater.

Repeat the above process for all the repeaters you want to start. In our case, we have the microphone going to line 2, the screen reader going to line 1 and line 1 going to line 2.
You will need to adjust the screen reader and your other audio software to use the repeater that you choose like you choose a soundcard. You will launch a separate instance of the audio repeater for each source you want to transmit.


To exit a repeater, hit alt+f4.
If you have set your screen reader’s output to a virtual cable, you should change it back to a real sound card or else you will temporarily lose speech.
In Windows 10, keep repeaters on a separate desktop to reduce clutter.


Virtual Audio Cable home page
The CAVI cast on Virtual Audio Cable

One thought on “Using virtual audio cable for synthetic speech demonstrations”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.