(Page 8 of 8 in this chapter)

7.7 System Restrictions

The following system restrictions should be noted regarding voice record and playback: Only one function can drive the output to the telephone line (an MVIP inband output stream); therefore, the following functions are mutually exclusive: Voice Recording (since the record beep prompt drives the output) Voice Playback Tone Generation FSK Sending The following functions are mutually exclusive: Voice Recording Voice Playback If the DTMF detector is disabled, the voice functions cannot terminate when digits are entered. The following functions are typically configured to share the same group of task processors: Call Progress Voice Record Voice Playback Tone Generation MF Detection For the typical configuration, DSP capacity is allotted under the assumption that every CTA context is running no more than one of these functions at any given time. There is nothing preventing the application from concurrently executing some combinations of these functions on some CTA contexts. If, however, multiple CTA contexts concurrently execute a combination of these functions, the DSP capacity will most likely be exhausted.
For the typical configuration, DSP capacity is allotted under the assumption that every CTA context is running no more than one of these functions at any given time. There is nothing preventing the application from concurrently executing some combinations of these functions on some CTA contexts. If, however, multiple CTA contexts concurrently execute a combination of these functions, the DSP capacity will most likely be exhausted.

7.7.1 Delays in Data Processing

Delays between initiating a command and its execution are influenced by the following factors: 10 ms batching - There are two (three on an AG-8) 10 ms delays between the line and the application that result from DSP data being "batched" in 10 ms blocks. For example, if a command is sent to the board to change the play volume, there will be a 20 ms delay (30 ms on AG-8) before the change appears on the line. Function start and stop - There is a 10-20 ms delay starting a DSP input function (record, for example) and a 20-30 ms delay to start an output function (play, for example). This is in addition to the batching delay explained above. For example, when play is started on an AG-8, 50-60 ms elapse before the data appears on the line (20-30 ms start-up delay, plus 30 ms batching delay). Load-related queueing delay - A heavily loaded AG board or host will not be able to process data as soon as the data is available. The AG board is designed to normally complete processing commands, events and data for all channels in each 10 ms time period. Occasionally, some processing may be delayed by 10 or 20 ms when a board is heavily loaded. Voice buffering - Although the DSPs process data in 10 ms blocks, it is not practical for the host to deliver the data in such small batches. Typically, play and record buffers hold 2-5 seconds of data. It is possible to use buffers as small as 100ms with a fast and/or lightly loaded host. Codec and switch delays - These add up to less than .5 millisecond.

(Page 8 of 8 in this chapter)

Tech_Support@nmss.com