Manual

But for several reasons, sidetone is always generated locally within the speaker’s equipment, rather than on the far end of the call. This
is because humans have a very hard me handling even the smallest delay in this sidetone signal. In tesng, we’ve found that any delay
over 10 ms starts to create an eect called “slapback” where the speaker is unable to maintain conversaon and begins to halt and
stuer.
Even in “old-fashioned” analog telephone circuits, it’s possible to create a 10 ms round-trip delay on a long distance call. Now add in the
requirement that modern VoIP-based or web audio systems have inherent windowing and buer delays, and it’s easy to pile up over 100
ms round-trip on a call. A delay of this length will typically not impede interacve conversaon, but will certainly create an intolerable
“slapback” environment if the caller hears his own voice delayed.
Many users installing a studio-based remote audio system for the rst me make the mistake of applying audio to the outgoing “send’
port that contains the main program feed (the same audio used to feed the transmier or webstream). Since this mix contains the
caller’s own audio, and there’s an inherent delay in modern digital systems, the “slapback” eect is immediate.
The soluon here is mix-minus (a term used for a special mix of audio that explicitly excludes one source) the audio coming from the
place the mix-minus is being sent. To put it another way, mix-minus is the enre studio mix minus one audio source.
So how do we create this special audio mix? On modern studio systems, this is usually well dened and easy to do. Many consoles feature
channels dedicated to telephone, ISDN or codec interface, and part of the channel is an automacally created mix-minus output.
In less full-featured consoles, a mix-minus can oen be created with an auxiliary or “audion” bus funcon. By selecng all relevant
incoming sources on the bus except for the fader ed to the caller’s output, you can do this easily. The following gure shows the block
diagram of a single mix-minus feed being generated on a mixing console.
The use of Opal complicates things a bit because it requires a pair of mix-minus feeds if both channels are to be used in the same studio.
This is because Opal does not conference the two callers it receives, and each caller needs a custom mix-minus in order to hear the other,
along with the studio-generated audio. This is shown in the following diagram:
16