Why Are Automatic Captions Not Working For Me?

Brivvio's automatic captions service uses speech recognition services to transcribe your spoken words and turn them into visual captions across the bottom of your video.

Automatic captioning works well as long as there is minimal background noise and the audio quality is good. Sentences need to be spoken clearly and the subject should not be too far from the microphone. Captions currently work with English language only. The caption service also requires a solid and reliable internet connection.

If you get a message saying that generating captions has failed, it may be due to one of the following issues:

  • Too much background noise 
  • Standing too far away from the phone's microphone
  • Non-English language
  • A strong dialect or accent 
  • Poor or interrupted internet service while generating captions
  • Too many other apps open on your phone, or low battery power 
  • Your video was captured on a different device or app. 

Here are some suggestions of how to deal with these issues...

Too much background noise 

A loud air-conditioner, fridge, machinery, car engine running, traffic on the street or other people talking in the background, can all be enough to prevent speech recognition software from being able to process the voice from your entire video.

Make sure you are recording in a quiet location. Avoid having multiple people talking over the top of each other.

Standing too far away from the phone's microphone

Your phone has several microphones, (one for the front facing camera, one for the away facing camera, and one at the bottom end for when you are on a call with the phone held up to your face). 

Brivvio uses the microphone that corresponds to the camera view you have selected. For example, if you are recording the front facing camera, it will use the front facing microphone. So Brivvio will record audio on that side of the camera more clearly. The same applies for using the away facing camera.

In a quiet space, the built in microphones of your phone are usually sufficient to capture voice clear enough for voice recognition up to about 2 metres. Speech recognition may not work for softly spoken people, or people further than 2 metres from your phone's microphone. The solution in this case is to speak louder or move closer to the camera.

Using a lapel microphones can also help improve audio quality and volume. There are several types of lapel microphones available specifically for phones, including wired and wireless types. We have found good results with the RØDE Smartlav+ and RØDE Wireless Go.


Mini directional microphones can also help. We do not recommend the RØDE VideoMic Me/Me-L unless the phone is mounted on a tripod, because it picks up vibrations from your hands if you're holding the phone while recording. Other options include the RØDE VideoMic Go and RØDE VideoMicro which have vibration reducing mounts. Another brand that have a similar range of microphones that we've found works well is Comica.

Non-English language

Currently, Brivvio speech recognition only recognises English language.

A strong dialect or accent 

Unfortunately, the cloud based speech recognition service we use is not yet able to recognise all accents and dialects.

Poor or interrupted internet service while generating captions

The speech recognition service requires solid and uninterrupted internet access in order to process speech.  

Check that you have internet access and do not have airplane mode turned on.

If you think you are in an area with poor internet access, then try re-generating captions when you are in an area with reliable and fast internet access.

Too many other apps open on your phone, or low battery power 

Speech recognition processing requires a chunk of your phone's CPU power to complete the process. If you have a large number of apps open in the background, they may be inhibiting CPU access and cause the process to fail.

If your phone's battery power is low, this may impact it's processing speed and cause automatic captions to fail. 

Try closing all non essential apps and then re-generate captions, and make sure your battery is charged to green level.

Your video was captured on a different device or app

If you recorded your video on a different camera, device, or another app, there may be issues with the audio quality or video file format, that cause the speech recognition to fail. Try recording in Brivvio to see if automatic captions works.

If you still have issues generating captions in Brivvio, reach out to us at or send us a message in the Brivvio app. We're always looking for ways to improve the app, so any additional information you can provide about your particular case is very helpful.