How to transcribe Voice to Text

• Posted by Common causes of Refrigerator Failure August 6, 2021 Filed in Business #Voice to Text  172 views
  click to rate

  In the last few years, the accuracy of voice to text transcription has improved markedly. Voice to text transcription (Aka VTT, V2T, Speech to Text, S2T, STT) is almost a commodity in 2021. When you pick up your iPhone and speak to Siri, undertake dictation on MS Word, speak to Alexa, turn on auto-captions on Google Meet, you are engaging in readily available voice to text services.

   

  Now, that voice to text is pervading almost every part of our digital lives, it becomes important to know the benefits of using this powerful, AI driven, tool:

   

  • Better comprehension and better user experience
  • Accessibility for deaf and hard of hearing users
  • Improve content SEO and discoverability
  • Increase useability and engagement of content – turn long form, spoken word content into blogs, articles social posts, extracts, snippets and more.

   

  It is now easier than ever to transcribe voice to text from your audio and video files. Firstly, choose a online transcription engine. There are many available, with many different features and languages to choose from. Sonnant is one that offers a free trial. It has great speed, low cost, timestamps, ability to search and use custom vocabulary and accuracy.

   

  Secondly, make sure you have your audio and video files ready on your computer, have the URL for the file ready to paste, or have the ability to upload the file from a linked cloud service like DropBox, GDrive or OneDrive.

   

  Upload your file and wait for your chosen transcription service to turn the spoken word into usable text. It is almost as simple as this, these transcription engines are able to covert voice to text quickly. The accuracy can depend on may factors.

   

  Depending on your service, your next set of options will be what to do with the transcript, or how your transcript gets represented. Here are some ideas:

   

  • Edited – No matter how good your transcription engine, there are some things that will be unknown to a transcription engine. Special nouns such as unique names maybe transcribed incorrectly.
  • Spelling – Some words may be phonetically transcribed such a names. E.g. Geoff, could be spelled Jeff.
  • Verbatim transcription – exactly every single word that is spoken is repeated. If you repeat a word, say “um” “ahh”, “yes, yes, yes, yes”, it doesn’t matter – you get an exact replica of the spoken word.
  • Consumable transcript – an editing process where what is spoken is interpreted to be more akin to a written piece of content. So this would involved removing “ums” and repeated word to get a better, more cohesive, meaning.

   

  Once you have the format of your finished product, the next step is using the files in a format the you need. Export your transcript into .doc / pdf / txt or use caption files in VTT / SRT format for creating engaging content for other platforms.

   

  The ability to transcribe voice to text is a vital element of turning the spoken word into discoverable content that will target and build your audience. Further, the ability to re-package and re-purpose spoken word audio and video into short-form posts and snippets can very quickly drive organic reach which will save you time and money.