Speech Engine Services Text-to-Speech Performance Counters
Speech Engine Services Text-to-Speech Performance Counters
Requests for speech recognition sent to a Microsoft Speech Server (MSS) computer are processed individually by a single text-to-speech (TTS) engine instance. TTS engine performance counters track performance across all TTS engine instances.
The following table describes the Speech Engine Services (SES) TTS Engine Instances performance counters.
Performance Counter |
Description |
|---|---|
Audio Bytes Sent/sec |
The number of audio bytes sent by all engine instances per second. |
Audio File Cache Hits |
The cumulative number of audio files requested that were loaded from the cache. |
Audio File Cache Misses |
The cumulative number of audio files requested that required the file to be downloaded. |
Prompt DB Cache Hits |
The cumulative number of prompt databases requested that were loaded from the cache. |
Prompt DB Cache Misses |
The cumulative number of prompt databases requested that required the database to be downloaded. |
Prompt Databases Loaded |
The total number of prompt databases currently loaded into all prompt engine instances. |
Prerecorded Prompt Hits |
The cumulative number of times that part of a speech output was generated from prerecorded prompts. |
Prerecorded Prompt Misses |
The cumulative number of times that part of a speech output required synthesized speech. |