For audio analysis, you upload the audio file to Lexeffect. In the first step, the audio file is transcribed, i.e. the speech is converted into text. This process can take a few minutes, depending on the length of the audio file.
This process runs in the background, so that you can continue working in the meantime. Via the notification icon in Lexeffect – to the left of your profile in the header – you get the information when the transcription is available.
A second reason for the separation of transcription and analysis is: Technologically, speech recognition is already very advanced, but there are currently still errors in its recognition. So that you can edit the results manually and make corrections, the text is displayed to you in the text field before you run the analysis.
The third reason also has something to do with technology. During transcription, Lexeffect recognizes the language of the audio file. This can also lead to errors, if e.g. many English terms are used in a German text. You can check the language before the analysis and change it, if necessary.
And then there is a fourth reason and that is economically justified. The accounting of the analysis with audio files takes place starting from the account professional per 1,000 characters. We only know how many characters an audio file contains after its transcription. Before the analysis, you yourself can control how many of your recorded words should be analyzed. This gives you full cost control.