Amazon Polly properties
3rd party integrations / Purple Manager
This article lists and explains the Properties for setting up TTS with Amazon Polly in Purple. You can use them for both ways of setting up TTS:
There are a few settings which need to be done either for the 'Team' or for each 'Publication'. If the settings are set for both, the publication settings are preferred.
Example: awsPollyVoices=Hans|male,Marlene|female
This setting is used to select the voices you have selected for text-to-speech and can be adjusted per selected voice and language, and it is also possible to select neutral and non-neutral voices. In this example, the voices for the German language are: 'Voice Hans' for the male voice and 'Voice Marlene for the female voice. The overview of all available voices can be found on the external website of the provider: https://docs.aws.amazon.com/polly/latest/dg/voicelist.html
Example: awsPollyLanguageCode=de-DE
This setting is used to select the language output you have selected for text-to-speech and can be adjusted per language selected. In this example, it is 'de-DE' for the German language. The list of language codes for all available languages can be found on the provider's external website: https://docs.aws.amazon.com/polly/latest/dg/voicelist.html
Example: awsPollySampleRate=22050
This setting indicates the quality of the MP3 file generated for the speech output and can be configured to your liking. The audio frequency is specified in Hz. In this example, we are working with 22050 Hz, which we recommend as the default configuration for both neural and non-neural voices. The higher the sample rate, the better the quality of the speech output. However, it also increases the size of the file, which can degrade the performance of your app. For more information on sample rate, see this external article: https://docs.aws.amazon.com/polly/latest/dg/API_SynthesizeSpeech.html#API_SynthesizeSpeech_RequestSyntax
Example: awsPollyAccessKey=AKIA2UY364J26EWPFLMN
The Access Key must be generated per team and by the customer himself/herself and entered into the AWS bucket set up in Purple Manager. To generate the keys, please go to your Amazon Web Services account. You can find more information on the external website of the provider: https://docs.aws.amazon.com/en_en/polly/latest/dg/install-voice-plugin2.html
Example: awsPollySecretKey=ml6ofPZ/+g6gxh1LD2IIb2yNJl7LkhcMN9pMeE6l
The Secret Key must be generated per team and by the customer himself/herself and entered into the AWS bucket set up in Purple Manager. To generate the keys, please go to your Amazon Web Services account. You can find more information on the external website of the provider: https://docs.aws.amazon.com/en_en/polly/latest/dg/install-voice-plugin2.html
Team setting | Name of custom property for publication | default | Example values |
Bucket Key Prefix | - | tts | - |
Sample Rate | awsPollySampleRate | no default, needs to be filled either for team or publication | 22050 |
Voices | awsPollyVoices | no default, needs to be filled either for team or publication no default, needs to be filled either for team or publication no default, needs to be filled either for team or publication no default, needs to be filled either for team or publication | English voices: Emma#Neural|female Amy#Neural|female German Voices: Marlene|female Hans|male Comma separated values are possible |
LanguageCode | awsPollyLanguageCode | no default, needs to be filled either for team or publication | en-GB de-DE |