TTS Receiving ERROR: Failed to write to GCS URI: gs://${bucketName}/output20.wav

Google Cloud Build with AI AI APIs
, ,
1

Hi everyone,

I came across a error on Google Text to Speech long audio synthesis that i can’t solve. hope you can help:

code is written in nest.js while using the client library

@google-cloud/text-to-speech

import textToSpeechCaller from '@google-cloud/text-to-speech';
const text = './//Really long text with a length of 907033 characters';
const texttospeechClient =
        new textToSpeechCaller.TextToSpeechLongAudioSynthesizeClient({
          credentials: cred,
        });
      const parent = `projects/${projectNumber}/locations/global`;
      const outputGcsUri = `gs://${bucketName}/output20.wav`;

      const [operation] = await texttospeechClient.synthesizeLongAudio(
        {
          parent: parent,
          // select the type of audio encoding
          audioConfig: { audioEncoding: 'LINEAR16' },
          input: { text: text },
          // Select the language and SSML voice gender (optional)
          voice: { languageCode: 'en-US', name: 'en-US-Standard-F' },
          outputGcsUri: outputGcsUri,
        },
        { timeout: 5000000, },
      );

      return operation.promise();

to give you some context the text is 907033 characters long and after running the package i get a bunch of files in my google cloud storage bucket(*image below) but then i get the error
ERROR [ExceptionsHandler] Failed to write to GCS URI: gs://${bucketName}/output20.wav

and so i thought google cloud timed out and put in the timeout in the above code and tried again then i get the error
ERROR [ExceptionsHandler] INTERNAL: [type.googleapis.com/util.ErrorSpacePayload=‘SpeechErrorSpace::TTS_BACKEND_REQUEST_RPC_ERROR’]
Error: INTERNAL: [type.googleapis.com/util.ErrorSpacePayload=‘SpeechErrorSpace::TTS_BACKEND_REQUEST_RPC_ERROR’]
at Operation._unpackResponse (C:\Users\me\Documents\projects\backend\node_modules@google-cloud\text-to-speech\node_modules\google-gax\build\src\longRunningCalls\longrunning.js:143:31)
at C:\Users\me\Documents\projects\backend\node_modules@google-cloud\text-to-speech\node_modules\google-gax\build\src\longRunningCalls\longrunning.js:123:18

and then i tried minimizing the characters and after some trial and error i found out that as long as the character length is below 600K then everything works but if i try more than that i keep getting the error so i thought maybe its a character limit as specified in the docs “Long Audio Synthesis asynchronously synthesizes up to 1 million bytes on input” but then when i checked the audios they are all generated meaning all the text has been changed to speech fully with a text length of 907033 characters but it can’t generate the file that holds it all output20.wav.

I am currently stuck here and don’t know what do next

storage2.png
storage2.png386×692 49.6 KB