> For the complete documentation index, see [llms.txt](https://autoedit.gitbook.io/documentation/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://autoedit.gitbook.io/documentation/transcriptions/transcriptions-intro/transcriber.md).

# Transcriber

## Component/part description

This is the module responsible for generating the transcription, generally using a third party service or API such as IBM Watson one.

{% embed url="<https://trello.com/c/7jSJ7Bc1>" %}

It is composed of 2 main components

* [**Audio converter**](/documentation/transcriptions/transcriptions-intro/transcriber/audio-to-video.md) Convert audio or video to audio specs for stt API

{% embed url="<https://trello.com/c/1fLY5mLQ>" %}

* [**STT sdk**](/documentation/transcriptions/transcriptions-intro/transcriber/stt-api.md) audio to STT API/Service, to receive time-coded transcription.

With Extra:

* **Speaker diarization** can either happen at the STT API level or as a separate module to be interpolated with the transcription.

{% embed url="<https://trello.com/c/jFRiJJRV>" %}

And optional:

* **Srt parsing**. Allow srt as input. In case transcription comes from elsewhere. Can use module [srtParserComposer to refactor](https://github.com/pietrop/srtParserComposer)

{% embed url="<https://trello.com/c/lZQYawmy>" %}

* **Plain text as input**, if you already have the transcription, use something like Gentle to re-align and generate transcription json.

## Related projects

It was Initially prototyped as a standalone app to test quality of speech to text. see [Transcriber](https://github.com/pietrop/Transcriber).

## Implementations Options considered

NA

## Current implementation

[See component](https://github.com/OpenNewsLabs/autoEdit_2/tree/master/lib/interactive_transcription_generator/transcriber)

## What needs refactoring

Perhaps look into compositor pattern to bring together the components of this module.