...
- Automatic speech recognition (ASR) captions are machine-transcribed using ASR technology. They can be created relatively quickly, typically taking just 15 minutes to generate captions for an hour-long video. ASR captions are - at best - 70% accurate. They should be considered a starting point from which 100% accurate captions can be created - with some editing.
- Human generated captions take longer to produce but are at least 99% accurate – taking into consideration any typing errors and audio quality. The turnaround for most caption transcription services is usually 24 hours for a one hour-long video. If you script a video prior to recording the script can also be used to create human generated captions.
...