Written by Marcel Goya
Updated May 6, 2024
LinguSocial Meetings transcribes calls in real-time and several factors determine the accuracy of your transcription. If you're experiencing trouble with the accuracy of transcriptions, consider making some changes that will assist LinguSocial's models to clearly understand what you're saying. Let's take a look at what you can do to improve LinguSocial Ai's ability to transcribe your calls.
One of the best things you can do to improve your transcript accuracy is to use a high-quality microphone. Laptops rely on tiny built-in microphones, which significantly limits the audio quality captured and attracts background noise easily as they're located away from your face. Either through a USB port or a standalone headset, use an external microphone and speak directly into it to largely reduce any background noise.
Speak loud and clear; if you're too quiet or if you mumble, AI might struggle to detect the words you're saying. Talk directly into the microphone from a distance of 1 to 3 inches for a headset microphone, or of 6 to 12 inches for a non-headset external microphone. Remember to breathe! Since our technology works best with conversational speech, take breaks when speaking for a long time or avoid monologuing completely if possible.
Create a custom dictionary to clearly define tricky or unusual words and phrases often used in your company or industry. Custom dictionaries allow LinguSocial Meetings to better recognize what you're saying in order to accurately transcribe a call that includes jargon not understood by the average person.
Conduct calls in a quiet environment, away from sources of noise as well as in an area that will not produce an echo. Find a smaller room that offers silence so that LinguSocial Meetings understands precisely what you're saying.
Audio can become choppy when your connection is poor. If possible, utilize a wired internet connection through an ethernet cable. Otherwise, try connecting to a 5GHz wireless network to ensure your connection is strong and stable.