NOT WORKING: This program is currently not working. I've opened an issue on @google-cloud/speech the node repo detailing issue and

Serverless audio to text

This is a test to turn this tutorial on Optical Character Recognition with server-less into a speech to text converter.

General application flow

1. Prepare audio file according to notes
2. Define cloud function on staging bucket (name: gs://influx-staging-bucket)
3. Upload .flac file into gcloud bucket (name: gs://influx-audio-upload)
4. Server-less function kicks off, pulls audio file, calls speech API
5. Function stores converted text into another bucket (name gs://influx-text-out)

Notes

1. To find sample rate, number of channels mediainfo myfile.flac
2. To convert between formats use the fre:ac program.
3. To remove an extraneous channel use the audacity program.

Next steps

1. Currently program outputs transcription to stdout, need to modify to store text in a file on cloud storage
2. Currently you need to manually set file type and sampleRateHertz. Make function identify this from tags or other from uploaded file