The embedded microphone should work without any problems for the casual transcription user for capturing shopping lists and notes, for example. But if you are using the software more intensively for regularly transcribing meetings or sending long emails, then you should look for something that will pick up the sound waves of your voice more effectively.
For less than $100, there are many headset or external microphone options available, such as those used by YouTube interviewers and podcasters.
Best tips for speech Recognition software
Although speech-to-text software recognizes different voices, it isn’t 100% perfect, but you can make it work better by following the tips below.
Use of natural voice
Dictation apps learn the user's voice and voice patterns over time. I you are using them for some time, you would want to be comfortable; therefore, speaking naturally is essential. Having said that, clearer enunciation is vital if the software isn't giving the user at least 90% accuracy from the start.
Punctuate
When transcribing, it's essential that you verbally add all punctuation marks, such as question marks, full stops, commas, and so forth. The speech-to-text software cannot usually figure that on its own
Practice a few commands
The users should ensure they learn basic commands like "new line" to move to the following line. Commands for editing, operating a device, and composing are very different. Commands can differ from one application to another application. So it's essential to find out the ones applicable to the tool of your choice.
Know the limits
On mobile devices in particular, some software and apps have a time limit on how long they can listen – sometimes just 10 seconds. It's essential to keep checking the screen continuously to ensure the session hasn't timed out.
Practice
Adjusting to the speech recognition software takes some time, but practising more makes it easier. You should go through help menus, on-screen tutorials, and cheat sheets.