This source of audio always introduces noise with the content you want to recognize.
Typically the audio captured via microphone and the recording files with noise should be using this type of source.
This source of audio should be the original files or the audio without noise.
Recorded Audio is more sensitive than Line-in Audio in recognition, but that doesn’t mean you can always use Recorded Audio in all of scenarios, it would be more false positions if you use Recording Audio in the scenarios should be using Line-in Audio.
- You want to make a Shazam alike app to recognize music or TV from the audio captured via microphone.
- You want to recognize background music over the TV drama, movies or TV shows.
- Although in this scenario you want to recognize something in the recording files, but background music in this kind of content always with people talks and other kind of noise alone with the music.
- You want to recognize original files of music to get metadata.
- You want to recognize music or commercials from recording files of radio live streams.