AI Tutor
If a singer whispers into a microphone, you hear a whisper β€” but when they suddenly belt out a high note, how does an audio compressor tame the volume without completely crushing the performance?
Audio engineers visualize this dynamic control using an input-output transfer characteristic graph, which maps the incoming signal level directly to the outgoing level.
What does a completely untouched audio signal look like on this grid?

The 1:1 Unity Line

In digital audio, volume is measured in dBFS (decibels relative to full scale), where 0Β dBFS0\text{ dBFS} is the absolute maximum level and all quieter signals are negative numbers like βˆ’20Β dBFS-20\text{ dBFS}.
When a processor applies no volume change, the input level equals the output level (Output=Input\text{Output} = \text{Input}) β€” a state called unity gain.
This linear unity gain line (1:1 line) passes through the origin/reference coordinates with a slope of m=1m = 1 (Ξ”Output/Ξ”Input=1\Delta Output / \Delta Input = 1), representing an uncompressed state where Output=InputOutput = Input.
This linear 1:1 line passes through the reference coordinates with a slope of m=1m = 1 (Ξ”Output/Ξ”Input=1\Delta\text{Output} / \Delta\text{Input} = 1), representing an uncompressed state.
So how do we show the exact moment a compressor steps in to reduce the level?

Threshold and Compression Slope

The threshold is the set input level where compression begins; any signal below it passes through unchanged along the 1:1 line.
Above the threshold, the curve flattens according to the compression ratio (R:1R:1), which determines how many decibels of input increase are required to produce a 1Β dB1\text{ dB} increase in output.