I cannot transcribe a local file of PCM data.

[couldbejake]

Which Deepgram product are you using?

Deepgram API

Details

I am attempting to transcribe the local file 48k_peter_jackson.pcm.
The deepgram node.js client returns no transcription.

The audio file contains 40 seconds of spoken text in a New Zealand accent. The audio file transcribes perfectly on other services such as Google SST, Whisper and Rev SST. The audio is sampled at 48khz, yet I see no option to set the sample rate of prerecorded audio on deepgram.

The response from the API shows:

{
  metadata: {
    transaction_key: 'deprecated',
    request_id: '583ff5e3-defb-4a1d-b6ef-f99c7a4bd5ba',
    sha256: 'de49bb76eb8b0aac2797f89f6d370e3b53b81c6d70a0e2b4392ca8431517a1a8',
    created: '2023-11-08T20:23:07.552Z',
    duration: 2.8628125,
    channels: 1,
    models: [ '96a295ec-6336-43d5-b1cb-1e48b5e6d9a4' ],
    model_info: {
      '96a295ec-6336-43d5-b1cb-1e48b5e6d9a4': { name: 'general', version: '2023-02-22.3', arch: 'base' }
    }
  },
  results: {
    channels: [
      {
        alternatives: [ { transcript: '', confidence: 0, words: [] } ]
      }
    ]
  }
}

No transcription is given.

If you are making a request to the Deepgram API, what is the full Deepgram URL you are making a request to?

No response

If you are making a request to the Deepgram API and have a request ID, please paste it below:

583ff5e3-defb-4a1d-b6ef-f99c7a4bd5ba

If possible, please attach your code or paste it into the text box.

const fs = require('fs')
const { Deepgram } = require('@deepgram/sdk');

// The API key we created in step 3
const deepgramApiKey = '<API KEY>';

// Initializes the Deepgram SDK
const deepgram = new Deepgram(deepgramApiKey);

console.log('Requesting transcript...')
console.log('Your file may take up to a couple minutes to process.')
console.log('While you wait, did you know that Deepgram accepts over 40 audio file formats? Even MP4s.')
console.log('To learn more about customizing your transcripts check out developers.deepgram.com.')

deepgram.transcription.preRecorded(
    {
        buffer: fs.readFileSync("./48k_peter_jackson.pcm"),
        mimetype: "audio/linear16",
    },
  {smart_format: true, model: 'nova', language: 'en-US', sample_rate: 48000 },
)
.then((transcription) => {
  console.dir(transcription, {depth: null});
})
.catch((err) => {
  console.log(err);
});

If possible, please attach an example audio file to reproduce the issue.

No response

[DamienDeepgram]

For Batch Transcription you will need to convert the file to a non-raw format

Here is an example how you can convert the file from raw to wav

ffmpeg -ar 16000 -f s16le -ac 1 -i test.pcm test.wav

[couldbejake]

For Batch Transcription you will need to convert the file to a non-raw format

Here is an example how you can convert the file from raw to wav

ffmpeg -ar 16000 -f s16le -ac 1 -i test.pcm test.wav

Thank you, works perfectly