Closed jcuccia closed 4 years ago
Having exactly the same problem here.
`tts:
Hey there @robbiet480, mind taking a look at this issue as its been labeled with a integration (amazon_polly
) you are listed as a codeowner for? Thanks!
Error message appears in 0.100.1 regardless of whether you have 'engine' defined in your configuration or not...I get the exact same error with it defined as 'engine: neural' and with that line commented out.
(Appreciating this has been said - I guess this is my way of saying 'me too')
@bbrendon did the last update to this component, adding neural engine support. It worked before that.
Brendon, if you're listening, can you please take a look?
This is the same issue as reported in https://github.com/home-assistant/home-assistant/issues/27418
Can the issues be combined?
I can't test your issue because of secrets but these are the supported regions. Are you using one of them?
US East (N. Virginia): us-east-1 US West (Oregon): us-west-2 EU (Ireland): eu-west-1
Also @pums12 Nicole is not a supported voice for Neural.
Have tried by hard coding to us-east-1 with the same result.
Though it’s no longer working even without neural...
And yes, this appears to be the same as 27418...though arguably if Polly isn’t working at all, it’s hard to tell if neural is working too ;)
@bbrendon thanks, I added the "engine: neural" line whilst trying to get it working again.
Prior to the upgrade to 0.100.1, my working configuration was:
I have also just tried:
but am still getting the same error "Unknown parameter in input: "Engine", must be one of: LexiconNames, OutputFormat, SampleRate, SpeechMarkTypes, Text, TextType, VoiceId, LanguageCode"
@Coch I have changed back to the standard engine so should be working ok with Nicole but am still getting the same error.
See the other issue on this for details. The problem is figured out. Solution unknown.
As a temporary solution, I have added amazon_polly as a custom component and edited tts.py replacing:
resp = self.client.synthesize_speech( Engine=self.config[CONF_ENGINE], OutputFormat=self.config[CONF_OUTPUT_FORMAT], SampleRate=self.config[CONF_SAMPLE_RATE], Text=message, TextType=self.config[CONF_TEXT_TYPE], VoiceId=voice_id, )
with: resp = self.client.synthesize_speech( OutputFormat=self.config[CONF_OUTPUT_FORMAT], SampleRate=self.config[CONF_SAMPLE_RATE], Text=message, TextType=self.config[CONF_TEXT_TYPE], VoiceId=voice_id, )
This gets it working again but only when using the 'standard' engine.
Same issue here. Using pt-br voice.
- platform: amazon_polly
aws_access_key_id: !secret
aws_secret_access_key: !secret
voice: "Vitoria"
cache: true
engine: 'standard'
My config didn't have "engine" before. I added now to try to solve the issue but, no change.
Closing this duplicate issue. See https://github.com/home-assistant/home-assistant/issues/27418 for updates.
Home Assistant release with the issue: 0.100.1
Last working Home Assistant release (if known): 0.99.3
Operating environment (Hass.io/Docker/Windows/etc.): NUC/Ubuntu18.04/DockerCE
Integration: Amazon Polly https://www.home-assistant.io/integrations/amazon_polly/
Description of problem: tts.amazon_polly_say does not work. It logs errors.
Problem-relevant
configuration.yaml
entries and (fill out even if it seems unimportant):Traceback (if applicable):
Additional information: I tried with engine: neural in the tts: config section, but that did not make any difference. I have reverted to version 0.99.3 and the Polly integration is working again.