Lompat ke konten Lompat ke sidebar Lompat ke footer

How to Easily Transcribe Voice to Text on a Computer (Voice To Text)

download image

Voice to Text- The existence of artificial intelligence or abbreviated as AI is very useful for fulfilling our daily needs. As is the case when we want to transcribe sound or voice into text immediately and easily.

With the existence of AI that is focused on recognizing sounds and giving commands to them to write those sounds into text, transcribing sounds is really easy thanks to the help of this tool.

Google has created a tool that makes full use of AI to be able to easily transcribe voice into text.

We can find these tools in the Chrome browser, by accessing a certain address we can use these tools easily.

The name of the tool isWeb Speech API Demonstration, this tool is useful for transcribing voices, both voices that have been recorded or sounds exclusively through the computer that we have.

How to Transcribe Voice to Text

To be able to use this tool, the first thing we have to provide is to use the Google Chrome Browser, if you don't have it, please install it on your computer.

You can search for the Google Chrome application via Google Search by typing the keyword "Google Chrome" then heading to the link that has been shown by Google.

Next, please access this linkSpeech, we will be directed to a simple looking site similar to the image below:

As we have seen, the appearance of the tools capable of transcribing voice to text is very simple. But do not be a priori to the abilities that he can do.

By doing the right settings, we can do many things with this tool.

Get to know the Web Speech API Demonstration Features


I explain step by step the features indicated by the arrows above:

  1. The arrow shown in number one serves to determine the sound that arises. By clicking on the mic sign, we have activated this tool to recognize the voice and then write the text automatically in arrow number 2.
  2. The arrow indicated by number 2 serves to store the text written from the voice recognition. The text is gradually written in the column.
  3. Number 3 indicated by the arrow above serves to determine what language is used in the voice or voice. If the language spoken is Indonesian, then choose Indonesian

How to Use Voice to Text Correctly

The first thing you prepare after successfully opening this tool is to plan the sound that you will transcribe into text.

This voice can come from your own voice, the sound of a speech, the sound of a recorded discussion. For the sound itself, it doesn't need to be recorded because we can say it right away.

Next, please confirm the language that arises from the sound. There are 60 language options with various dialects, including Javanese, which is also available in this tool.

After determining the language, the next step is to press the number one button, then listen to the tool's voice. That way the tool automatically transcribes voice to text. The text will appear in column arrow number 3.

Warning: The text in column three will disappear when you press the on button on the number 1 button.

So when you have finished transcribing the sound and deactivating the button, immediately copy and secure it in Microsoft Word or another note-taking application.

Voice to Text Voice Source Recognition Settings

These tools can work optimally on two types of input sound sources in Windows. The sound source can come from an external microphone and an internal microphone.

Sound input from an external microphone uses other tools not found on the computer, for example headphones that have sound input.

While the sound input from the Internal Microphone comes from within the computer itself. For example, the sound of music from Youtube that is heard on laptop speakers can be used as sound input.

For Windows 10 users, we can make settings to adjust where the sound input source comes from.

Please follow the settings below:

To go to settingsSoundplease right click on the speaker icon in the lower right corner, then selectOpen Sound Settings.

On the settings chat buttonSoundplease select Sound and get the Input menu. There are two types of microphones.

If you want to take sound input from your own voice exclusively choose Microphone (Realtek High Definition Audio or depending on the type of hardware on the computer).

If you want to take sound from media sources on your laptop, for example music player applications, youtube and others, please select Stereo Mix (adjusting to the name of the hardware you are using).

The description of these settings is as shown in the image below:


Some Weaknesses Of Voice To Text

No one is perfect except Gusti Allah SWT, whatever is created by humans always has weaknesses.

Likewise withWeb Speech API Demonstration, to be able to use it to its full potential, there are special prerequisites that must be met.

Like the voice to be transcribed, it must have a clear and bright sound, at a speed that the machine can understand.

The sound that is not too clear, too clean and not even too loud cannot be recognized by the AI.

Another weakness is, AI only does the job of transcribing sound into text, that's all. It is incapable of performing tasks such as adjusting punctuation or controlling paragraphs.

Because this is part of the aesthetics of writing, while the machine has no aesthetic value. So for the work of setting punctuation and paragraphs we have to do it ourselves.

But hand in hand, assisted by AI in transcribing voice to text, makes us more time efficient. We no longer need to type text.

Good luck friends.

download image