This guide describes how to configure and use the NVIDIA Riva Speech Synthesis (SS) plugin to the UniMRCP server. The document is intended for users having a certain knowledge of NVIDIA Riva Speech Skills and UniMRCP.
For installation of the packages, use one of the manuals below.
Instructions provided in this guide are applicable to the following versions.
UniMRCP 1.7.0 and above
UniMRCP Riva SS Plugin 1.0.0 and above
This is a brief check list of the features currently supported by the UniMRCP server running with the Riva SS plugin.
SPEAK
STOP
PAUSE
RESUME
BARGE-IN-OCCURRED
SET-PARAMS
GET-PARAMS
SPEECH-MARKER
SPEAK-COMPLETE
Kill-On-Barge-In
Completion-Cause
Voice-Gender
Voice-Name
Prosody-Rate
Prosody-Volume
Speech-Language
Logging-Tag
Cache-Control
Plain text (text/plain)
SSML (application/ssml+xml or application/synthesis+ssml)
Check out the voices supported by the NVIDIA Riva Speeck Skills:
The configuration file of the Riva SS plugin is located in /opt/unimrcp/conf/umsrivass.xml. The configuration file is written in XML.
The root element of the XML document must be <umsrivass>.
Attributes
Name | Unit | Description |
---|---|---|
license-file | File path | Specifies the license file. File name may include patterns containing '*' sign. If multiple files match the pattern, the most recent one gets used. |
Parent
Children
Name | Unit | Description |
---|---|---|
synth-settings | String | Specifies synthesis parameters employed via gRPC. |
waveform-manager | String | Specifies parameters of the waveform manager. |
sdr-manager | String | Specifies parameters of the Synthesis Details Record (SDR) manager. |
monitoring-agent | String | Specifies parameters of the monitoring manager. |
license-server | String | Specifies parameters used to connect to the license server. The use of the license server is optional. |
Example
This is an example of a bare document.
<umsrivass license-file="umsrivass_*.lic">
</umsrivass>
This element specifies synthesis parameters.
Attributes
Name | Unit | Description |
---|---|---|
language | String | Specifies the default language to use, if not set by the client. |
bypass-ssml | Boolean | Specifies whether to transparently bypass or parse received content in order to determine voice parameters set in SSML. |
normalize-ssml | Boolean | Specifies whether to normalize SSML. The parameter is observed only when bypass-ssml is set to false. |
voice-name | String | Specifies the default voice name. Can be overridden by client. |
voice-gender | String | Specifies the default voice gender. Can be overridden by client. |
http-proxy | String | Specifies the URI of HTTP proxy, if used. |
grpc-log-redirection | Boolean | Specifies whether to enable gRPC log redirection. |
grpc-log-verbosity | String | Specifies gRPC logging verbosity. One of DEBUG, INFO, ERROR. See GRPC_VERBOSITY for more info. |
grpc-log-trace | String | Specifies a comma separated list of tracers producing gRPC logs. Use 'all' to turn all tracers on. See GRPC_TRACE for more info. |
caching | Boolean | Specifies whether to enable caching of synthesized waveforms. |
prosody-rate | Double | Specifies the default prosody rate (speaking_rate) in the range [0.25, 4.0]. Can be overridden by client. |
prosody-volume | Double | Specifies the default prosody volume (volume_gain_db) in the range [-96.0, 16.0]. Can be overridden by client. |
prosody-pitch | Double | Specifies the default prosody pitch in the range [-20.0, 20.0]. |
deadline | Time interval [msec] | Specifies the gRPC call deadline. Defaults to 0 (disabled). |
reattempt | Boolean | Specifies whether to reattempt the gRPC call if the original attempt fails. Disabled by default. |
service-uri | String | Specifies the service endpoint and defaults to 127.0.0.1:50051. |
secure-transport | Boolean | specifies whether to use secure (TLS) or insecure transport. Defaults to insecure. |
Parent
<umsrivass>
Children
Example
This is an example of synthesis parameters.
<synth-settings
language="en-US"
bypass-ssml="true"
normalize-ssml="true"
voice-name=""
voice-gender=""
/>
This element specifies parameters of the waveform manager.
Attributes
Name | Unit | Description |
---|---|---|
save-waveforms | Boolean | Specifies whether to save waveforms or not. |
purge-existing | Boolean | Specifies whether to delete existing records on start-up. |
max-file-age | Time interval [min] | Specifies a time interval in minutes after expiration of which a waveform is deleted. Set 0 for infinite. |
max-file-count | Integer | Specifies the max number of waveforms to store. If reached, the oldest waveform is deleted. Set 0 for infinite. |
waveform-folder | Dir path | Specifies a folder the waveforms should be stored in. |
file-prefix | String | Specifies a prefix used to compose the name of the file to be stored. Defaults to 'umsrivass-', if not specified. |
use-logging-tag | Boolean | Specifies whether to use the MRCP header field Logging-Tag, if present, to compose the name of the file to be stored. |
Parent
<umsrivass>
Children
Example
The example below defines a typical utterance manager having the default parameters set.
<waveform-manager
save-waveforms="false"
purge-existing="false"
max-file-age="60"
max-file-count="100"
waveform-folder=""
/>
This element specifies parameters of the Synthesis Details Record (SDR) manager.
Attributes
Name | Unit | Description |
---|---|---|
save-records | Boolean | Specifies whether to save recognition details records or not. |
purge-existing | Boolean | Specifies whether to delete existing records on start-up. |
max-file-age | Time interval [min] | Specifies a time interval in minutes after expiration of which a record is deleted. Set 0 for infinite. |
max-file-count | Integer | Specifies the max number of records to store. If reached, the oldest record is deleted. Set 0 for infinite. |
record-folder | Dir path | Specifies a folder to store recognition details records in. Defaults to ${UniMRCPInstallDir}/var. |
file-prefix | String | Specifies a prefix used to compose the name of the file to be stored. Defaults to 'umsrivass-', if not specified. |
use-logging-tag | Boolean | Specifies whether to use the MRCP header field Logging-Tag, if present, to compose the name of the file to be stored. |
Parent
<umsrivass>
Children
Example
The example below defines a typical utterance manager having the default parameters set.
<sdr-manager
save-records="false"
purge-existing="false"
max-file-age="60"
max-file-count="100"
waveform-folder=""
/>
This element specifies parameters of the monitoring agent.
Attributes
Name | Unit | Description |
---|---|---|
refresh-period | Time interval [sec] | Specifies a time interval in seconds used to periodically refresh usage details. See <usage-refresh-handler> . |
Parent
<umsrivass>
Children
<usage-change-handler>
<usage-refresh-handler>
Example
The example below defines a monitoring agent with usage change and refresh handlers.
<monitoring-agent refresh-period="60">
<usage-change-handler>
<log-usage enable="true" priority="NOTICE"/>
</usage-change-handler>
<usage-refresh-handler>
<dump-channels enable="true" status-file="umsrivass-channels.status"/>
</usage-refresh-handler >
</monitoring-agent>
This element specifies an event handler called on every usage change.
Attributes
Parent
<monitoring-agent>
Children
<log-usage>
<update-usage>
<dump-channels>
Example
This is an example of the usage change event handler.
<usage-change-handler>
<log-usage enable="true" priority="NOTICE"/>
<update-usage enable="false" status-file="umsrivass-usage.status"/>
<dump-channels enable="false" status-file="umsrivass-channels.status"/>
</usage-change-handler>
This element specifies an event handler called periodically to update usage details.
Attributes
Parent
<monitoring-agent>
Children
<log-usage>
<update-usage>
<dump-channels>
Example
This is an example of the usage change event handler.
<usage-refresh-handler>
<log-usage enable="true" priority="NOTICE"/>
<update-usage enable="false" status-file="umsrivass-usage.status"/>
<dump-channels enable="false" status-file="umsrivass-channels.status"/>
</usage-refresh-handler>
This element specifies parameters used to connect to the license server.
Attributes
Name | Unit | Description |
---|---|---|
enable | Boolean | Specifies whether the use of license server is enabled or not. If enabled, the license-file attribute is not honored. |
server-address | String | Specifies the IP address or host name of the license server. |
certificate-file | File path | Specifies the client certificate used to connect to the license server. File name may include patterns containing a * sign. If multiple files match the pattern, the most recent one gets used. |
ca-file | File path | Specifies the certificate authority used to validate the license server. |
channel-count | Integer | Specifies the number of channels to check out from the license server. If not specified or set to 0, either all available channels or a pool of channels will be checked based on the configuration of the license server. |
http-proxy-address | String | Specifies the IP address or host name of the HTTP proxy server, if used. |
http-proxy-port | Integer | Specifies the port number of the HTTP proxy server, if used. |
security-level | Integer | Specifies the SSL security level, which defaults to 1. Applicable since OpenSSL 1.1.0. |
Parent
<umsrivass>
Children
Example
The example below defines a typical configuration which can be used to connect to a license server located, for example, at 10.0.0.1.
<license-server
enable="true"
server-address="10.0.0.1"
certificate-file="unilic_client_*.crt"
ca-file="unilic_ca.crt"
/>
This section outlines common configuration steps.
The default configuration should be sufficient for the general use.
Synthesis language can be specified by the client per MRCP session by means of the header field Speech-Language set in a SET-PARAMS or SPEAK request, or inline in the SSML data. Otherwise, the parameter language set in the configuration file umsrivass.xml is used. The parameter defaults to en-US.
Sampling rate is determined based on the SDP negotiation. Refer to the configuration guide of the UniMRCP server on how to specify supported encodings and sampling rates to be used in communication between the client and server. Either 8 or 16 kHz can be used by the NVIDIA Riva Text-to-Speech API for synthesis.
Global Settings
The default voice name and gender can be specified from the configuration file umsrivass.xml using the voice-name and voice-gender attributes of the synth-settings element.
MRCP Header Fields
The voice name and gender can be specified by the MRCP client in SET-PARAMS and SPEAK requests.
SSML Content
The voice name and gender can also be specified using the corresponding attributes of the voice element in SSML content. In order to parse and determine the parameters and pass them forward to the NVIDIA Riva Text-to-Speech API accordingly, the bypass-ssml attribute of the synth-settings element must be set to false in the configuration file umsrivass.xml.
If the bypass-ssml attribute is set to false and the normalize-ssml attribute is set to true, then the voice element, if present, is stripped off from the SSML content passed to the service in order to conform to the subset of SSML supported by the NVIDIA Riva Text-to-Speech API.
The voice name and gender can also be specified via the header field Vendor-Specific-Parameters.
The following prosody parameters can be specified by the MRCP client in SET-PARAMS and SPEAK requests.
The following parameters can optionally be specified by the MRCP client in SET-PARAMS, DEFINE-GRAMMAR and RECOGNIZE requests via the MRCP header field Vendor-Specific-Parameters.
Name | Unit | Description |
---|---|---|
service-uri | String | Specifies the URI of the service endpoint. |
voice-name | String | Specifies the voice name. |
voice-gender | String | Specifies the voice gender. |
Speech data can be specified by the MRCP client in SPEAK requests using one of the following content types:
plain/text
application/ssml+xml (or application/synthesis+ssml)
Collection of waveforms is not required for regular operation and is disabled by default. However, enabling this functionality allows to save synthesized speech received from the Riva service endpoint and later listen to them offline.
The relevant settings can be specified via the element waveform-manager.
Collection of synthesis details records (SDR) is not required for regular operation and is disabled by default. However, enabling this functionality allows to store details of each synthesis attempt in a separate file and analyze them later offline. The SDRs ate stored in the JSON format.
The relevant settings can be specified via the element sdr-manager.
Synthesized waveforms can be stored and re-used for consecutive speech synthesis requests, when applicable. In order to use this functionality, the attribute caching of the element synth-settings must be set to true. The attribute defaults to false.
The lifetime and size of cached records are controlled by the attributes max-file-age and max-file-count of the element waveform-manager.
The cached records are persistent and populated on initial loading, unless the attribute purge-existing of the element waveform-manager is set to true.
The following speech synthesis parameters are observed while searching for a cached record.
language
voice-name
voice-gender
sampling-rate
prosody-rate
prosody-volume
content
The following cache control directives are observed while searching for a cached record.
max-age
min-fresh
The cache control directives can be specified by the client per individual speech synthesis request via the MRCP header field Cache-Control. By default, no cache control directives are applied.
The number of in-use and total licensed channels can be monitored in several alternate ways. There is a set of actions which can take place on certain events. The behavior is configurable via the element monitoring-agent, which contains two event handlers: usage-change-handler and usage-refresh-handler.
While the usage-change-handler is invoked on every acquisition and release of a licensed channel, the usage-refresh-handler is invoked periodically on expiration of a timeout specified by the attribute refresh-period.
The following actions can be specified for either of the two handlers.
The action log-usage logs the following data in the order specified.
The number of currently in-use channels.
The maximum number of channels used concurrently.
The total number of licensed channels.
The following is a sample log statement, indicating 0 in-use, 0 max-used and 2 total channels.
[NOTICE] RIVASS Usage: 0/0/2
The action update-usage writes the following data to a status file umsrivass-usage.status, located by default in the directory ${UniMRCPInstallDir}/var/status.
The number of currently in-use channels.
The maximum number of channels used concurrently.
The total number of licensed channels.
The current status of the license permit.
The license server alarm. Set to on, if the license server is not available for more than one hour; otherwise, set to off. This parameter is maintained only if the license server is used.
The following is a sample content of the status file.
in-use channels: 0
max used channels: 0
total channels: 2
license permit: true
licserver alarm: off
The action dump-channels writes the identifiers of in-use channels to a status file umsrivass-channels.status, located by default in the directory ${UniMRCPInstallDir}/var/status.
This examples demonstrates how to perform speech synthesis by using a SPEAK request with an SSML content.
C->S:
MRCP/2.0 309 SPEAK 1
Channel-Identifier: 4dde51f37d1a9546@speechsynth
Content-Type: application/ssml+xml
Voice-Age: 28
Content-Length: 163
<?xml version="1.0"?>
<speak version="1.0" xml:lang="en-US" xmlns="http://www.w3.org/2001/10/synthesis">
<p>
<s>Welcome to Uni MRCP.</s>
</p>
</speak>
S->C:
MRCP/2.0 83 1 200 IN-PROGRESS
Channel-Identifier: 4dde51f37d1a9546@speechsynth
S->C:
MRCP/2.0 122 SPEAK-COMPLETE 1 COMPLETE
Channel-Identifier: 4dde51f37d1a9546@speechsynth
Completion-Cause: 000 normal
This example demonstrates how to perform speech synthesis by using a SPEAK request with a plain text content.
C->S:
MRCP/2.0 155 SPEAK 1
Channel-Identifier: 85667d0efbf95345@speechsynth
Content-Type: text/plain
Voice-Age: 28
Content-Length: 20
Welcome to Uni MRCP.
S->C:
MRCP/2.0 83 1 200 IN-PROGRESS
Channel-Identifier: 85667d0efbf95345@speechsynth
S->C:
MRCP/2.0 122 SPEAK-COMPLETE 1 COMPLETE
Channel-Identifier: 85667d0efbf95345@speechsynth
Completion-Cause: 000 normal
The following sequence diagrams outline common interactions between all the main components involved in a typical synthesis session performed over MRCPv1 and MRCPv2 respectively.
All the data is carried over an HTTP/2 connection via the gRPC streaming.