Clipping is a waveform distortion that occurs when an audio signal is amplified beyond the maximum allowed limit. This creates a deviation in the shape of audio which can introduce undesirable, staticky, fizzy, or crunchy harsh sounds. Peaking is another term that is sometimes used to describe this phenomenon.
Minor clipping doesn't typically affect a person's ability to understand audio content, but it does cause a sense of annoyance among some listeners. Major clipping events can badly distort audio and create speech intelligibility issues.
The Media APIs can help you identify clipping problems.
Clipping can occur at any stage of audio capture and processing. Typically, audio equipment has a maximum amount of electrical voltage that can be handled. If volume is the level of audio coming out of a speaker, the gain is the level of audio that is going into the equipment. If the gain is at a high enough level to exceed the recording threshold, this can introduce clipping.
For professional quality audio, Input Monitoring is a good practice to measure input levels coming into the audio equipment before hitting record. Not all audio can be captured in a professional setting to avoid clipping though, which is where the Media APIs can help in scenarios such as:
- conference calls
Any scenario where the microphone is close to a person's mouth could suffer from clipping distortion. Alternatively, when a microphone is far away from the talker such as in a large venue like a lecture hall, there is less likely to be clipping events.
You can use /media/analyze to determine the number or count of clipping events. A couple of clipping events over the course of a minute may be negligible, but if the number starts to exceed five for example, it may indicate a problem with the content.
Updated 8 months ago