Brivvio captions are of the highest accuracy and speed that you'll find anywhere.
And while most of the time, they're pretty darn good, they're not 100% perfect. The AI that turns the speech in your videos into text captions relies on three things to do its job well.
If the AI can't recognise speech in a section of your video, you might get incorrect words, or even a gap in your AI generated captions.
Here are the things to look out for...
1. Speaker volume and clarity
The speaker must not be too far away form the microphone such that the volume is too low, and must speak clearly enough so that the speaker's words can be easily recognised.
2. Minimal background noise
Things like traffic, machinery, people talking or laughing in the background can sometimes throw a spanner in the works for our AI when it comes to recognising the speaker's words.
3. One speaker at a time.
Just like us humans, our AI can sometimes have difficulty distinguishing two speakers talking at the same time.
File formats on imported videos
This only applies to imported videos. Sometimes, we come across a peculiar video file format that encodes the audio in an an uncommon format (codec). Brivvio will recognise the audio in most standard video files, but if you find it's just not picking up words at all in the captions, and the first three points above are taken care of, then it might be to do with the audio codec of the video file you're importing.
In this case, try converting your video file to a standard MP4 video file with AAC audio codec, and then import the converted file into Brivvio and see if that works better. And if that doesn't work, please submit a support ticket here - we're happy to troubleshoot things like this.