Auto transcribe
Auto transcribe is our most popular option for creating captions. Clicking on it will send an audio file to a speech recognition provider, and they'll return a raw transcript to us. Using CaptionHub's proprietary Natural Captions© technology, we'll then process that transcript into perfectly aligned, human readable captions.
We use two different speech recognition providers: Speechmatics, and Amazon Transcribe. Both have slightly different functionality, which is supported in CaptionHub. To change speech recognition provider, navigate to Team Settings, and change providers from the Transcription sub-menu.
To use Auto transcribe, select it in the "Create captions" dialogue box, or from the "Replace captions" dialogue box. You'll need to choose the language that your video is in.
![The Auto transcribe dialogue box](https://www.notion.so/image/https%3A%2F%2Fs3-us-west-2.amazonaws.com%2Fsecure.notion-static.com%2F8a77190c-cdcb-45b7-be06-33d871851063%2FScreenshot_2022-01-31_at_08.30.19.png?table=block&id=7500188c-40c9-44a2-89cb-0dc14be48f4d&cache=v2)
Supported languages
You can see which transcription engine supports which languages here.
Custom dictionaries
Depending on your subscription, you may have custom dictionaries enabled. This is a feature that biases the speech recognition towards a pre-defined list of words – like names, places, products and custom terminology. You'll need to pre-define a custom dictionary in Team Settings, and you'll need to choose it here, in order to submit it for transcription.
Reflow
Immediately after auto-transcription, you’ll see the “Reflow” option in the drop-down menu in the Caption Editor. This allows you to change the caption limits, without initiating a brand new auto-transcription. Once you’ve changed any text on that page, however, then the option will disappear from the menu.