This document is intended to give guidance on how many channels can simultaneously perform these popular tasks on a single Prosody Processor:
In a typical conferencing application, each participant can exercise control using DTMF digits. This may allow them to adjust their volume, drop out of the conference, or do more unusual operations such as talk privately to another conference participant. The channel capacity below assumes that DTMF detection is required. The allocation of channels to conferences has no significant effect on resource usage. Ten channels can be arranged into a single conference of ten parties, two conferences of five, or five conferences of two, with all these using approximately the same quantity of resources. Therefore the capacity of a Prosody processor is given in channels and you can arrange these in any combination of conferences.
| Operation | Channels |
|---|---|
| Conference participant | 64 |
There are several applications which use the same resources. Here are some examples:
Each of these applications needs the same basic features in Prosody: the ability to play messages to the caller and the ability to recognise the incoming DTMF digits. Therefore they appear in our table called Play with DTMF control.
| Operation | Channels |
|---|---|
| Play with DTMF control | 64 |
Each of the basic interactive applications described above can be controlled using speech recognition instead of (or as well as) DTMF digits. Speech recognition is performed on a host system, with the data being fed from Prosody.
| Operation | Channels |
|---|---|
| Play with speech recognition feed | 64 |
| Play with speech recognition feed and DTMF | 64 |
When an application is controlled by speech, 'barge-in' is the ability to speak and be recognised while the system is still playing a message. Speech is more challenging to recognise at this time as the message is typically speech, so its echo might be thought to be the user's response. To avoid this problem, the system must use echo cancellation. This can be done with or without DTMF detection, and the typical capacity of a Prosody Processor is:
| Operation | Channels |
|---|---|
| Play with speech recognition feed and barge-in | 27 |
| Play with speech recognition feed, barge-in, and DTMF | 24 |
There are several variations possible for a service like voicemail. In addition to the basic facility to record a message, you can allow the user to speed-up and slow-down the replay of messages, and you can use speech recognition to control it. The capacity of a Prosody processor is typically:
| Operation | Channels | |
|---|---|---|
| Voicemail | Basic | 64 |
| With fast and slow replay | 24 | |
| With speech recognition feed | 64 | |
| With speech recognition feed and barge-in | 27 | |
| With speech recognition feed, barge-in, and DTMF | 24 | |
In addition to the basic fax abilities of sending and receiving, you can have a system which automatically detects whether a call is a fax or voice call. The entries in the table below assume that such calls do not send faxes.
| Operation | Channels | |
|---|---|---|
| Send fax | V.27ter and V.29 | 48 |
| V.27ter, V.29, and V.17 | 42 | |
| Receive fax | V.27ter | 21 |
| V.27ter and V.29 | 12 | |
| Auto select: receive fax or speech recognition feed | V.27ter | 21 |
| V.27ter and V.29 | 12 | |
| Auto select: receive fax or speech recognition feed with barge-in | V.27ter | 21 |
| V.27ter and V.29 | 12 | |
If you want to do something a bit different, you might like to try using the TiNG channel capacity calculator. It has the popular tasks at the top, with the individual operations below, so you can see exactly what the popular tasks do. You can then modify the set of operations to reflect what you want to do, or even add your own sets. After you have entered a description of the operations for each of the types of call you intend to handle, when you go to its summary page you will see the result of its calculations.