🎚️ Psychoacoustic Masking Simulator

About this simulation

Psychoacoustic masking happens when one sound, the masker, raises the ear's detection threshold for nearby frequencies so much that a second, quieter sound — the target — becomes completely inaudible even though its physical energy is well above the threshold of hearing in silence. This is not a flaw in hearing; it is how the cochlea's overlapping frequency filters (the "critical bands") work, and it is the single biggest reason lossy audio codecs like MP3 and AAC can throw away most of a signal without anyone noticing.

How it works

• The purple ribbon is the masking threshold curve, computed from the masker's frequency and level using a Bark-scale spreading function — the same kind of model used inside real perceptual audio codecs.
• The tall orange spike is the masker tone; the thin spike is the target tone.
• When the target spike pokes above the purple threshold surface it stays bright and is marked AUDIBLE. When it falls under the shaded volume beneath the curve it dims and is marked MASKED — inaudible in the presence of the masker.
• The spreading function is asymmetric: masking spreads further toward higher frequencies than lower ones, so a loud low tone masks more of the spectrum above it than below.

Controls

• Masker frequency / level — sets the pitch and loudness of the dominant tone.
• Target frequency / level — sets the pitch and loudness of the quiet tone you're trying to hear.
• 🔊 Play tones — synthesizes both tones with the Web Audio API so you can actually hear masking happen.
• ↺ Reset — restores the default masker/target settings. Drag to rotate, scroll to zoom.

Did you know?

Lossy audio compression exploits masking directly: encoders estimate a masking threshold for every short time-frequency block and allocate almost no bits to anything predicted to be inaudible, which is how a 5-minute song can shrink from 50MB to 4MB with no audible loss.