Hey there! This channel allows you to ask Kapa.ai for assistance. But it looks like you haven't connected your Discord profile to your Deepgram account yet. You can do so at https://community.deepgram.com - by doing so, you'll be able to ask questions in #1115960287183511643 if you need human assistance!
#empty transcript
vivid gulchBOT
marsh oakBOT
Hi @fringe loom. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
There are several common reasons why you might receive an empty (blank) transcript when using Deepgram's Speech-to-Text (STT) API with JavaScript in the Chrome browser:
-
Incorrect Audio Format or Encoding
If the audio you are sending is not in the expected format or encoding, Deepgram may not be able to process it, resulting in an empty transcript. For example, if you are streaming audio, ensure the sample rate and encoding match what Deepgram expects (e.g., 16kHz or 8kHz, linear16, mulaw, etc.)problem with transcribing mp3 files with Deepgram, Empty transcript when streaming RTP audio. -
Base64 Encoding Issue
If you are sending audio data as base64-encoded text instead of raw binary, Deepgram will not be able to transcribe it. Make sure to send raw binary audio data over the WebSocket, not base64-encoded datanova-2 model started returning empty strings as transcription results. -
Incorrect Language Setting
If the language parameter is set incorrectly (e.g., defaulting to English for non-English audio), Deepgram may return an empty transcript. Set the correct language code using thelanguageparameter or try enabling language detectionSpanish audio - no transcription in the response. -
Audio Quality Issues
If the audio is very noisy, too quiet, or the speech is distant from the microphone, the model may not detect speech and return an empty transcript. Try using a different model (e.g.,enhancedinstead ofnova-2) if your audio has challenging acoustic conditionsNova 2 Spanish long transcriptions return an empty transcription.
-
Sample Rate Mismatch
For streaming, if thesample_rateparameter does not match the actual audio sample rate, you may get empty results. Try increasing the sample rate if possibleEmpty transcript when streaming RTP audio. -
Recent API or SDK Changes
If your code was working previously and suddenly started returning empty transcripts, check if there have been any changes to the Deepgram API or SDK. For example, removing unnecessary parameters likemimetypein the SDK call has resolved similar issuesDeepgram Nova Returning Blank Transcripts.
What to do next:
- Double-check your audio encoding, sample rate, and ensure you are sending raw binary data.
- Verify the language parameter matches your audio.
- Test with a known good audio file (e.g., clear English speech) to isolate the issue.
- Review your code and recent changes to the Deepgram SDK or API.
If you continue to have issues, providing a request ID, sample audio, and your code will help Deepgram support diagnose the problem more specifically.