PerceptEQ profiles your ears and headphones together, then builds a custom EQ that corrects for both at once. The whole process takes about 3 minutes.
Stage 1Frequency Matching
Pure tones step through the frequency range in random order. Use the slider to match their loudness — this maps how your system presents each frequency.
Stage 2Band Balance
Broader frequency bands, with Stage 1 corrections already applied. A final check to make sure the overall tonal balance sounds right.
Before you start: Put on your headphones and use the button below to play some noise. Set your volume to a comfortable listening level and keep it there for the whole test.
Stage 1 of 2 Frequency Matching
Match the volume
Two tones play one after the other. Use the slider to make them feel the same loudness — that's all you're doing here. Don't overthink it.
Tip: Close your eyes if it helps. The pitches will be quite different — only loudness matters here, not the tone quality.
1000Hz · lower
⇔
1.3kHz · upper
← lower is louderupper is louder →
0 / 0 pairs done
Stage 2 of 2 Band Balance
Match the volume
Same idea, but with bands of noise rather than pure tones. Your Stage 1 corrections are already applied — nudge anything that still sounds uneven.
Tip: Focus on the overall sense of level between the two sounds. Ignore any tonal differences and just chase the centre point where they feel equally loud.
20–38Hz · lower
⇔
38–72Hz · upper
← lower is louderupper is louder →
0 / 15 pairs done
Export Done
Your correction curve
This is your personalised correction. Adjust the graph smoothing below, then download the convolution IR to load into any convolver (HeSuVi, Voicemeeter, DAW, Equalizer APO).
← precisesmoothing: 1/8 octsmooth →
Play pink noise then toggle EQ on/off to hear the correction. The graph smoothing controls both the preview and the downloaded IR.