Level ABX Test — Loudness Drill
Can you truly hear tiny 1.0, 0.5 or 0.2 dB volume changes? Double-blind ABX loudness trials with strict binomial scoring. Free in your browser, no download.
Protect your hearing. Start at a low volume. Never raise the volume to make a tone audible. Levels are shown in dBFS, relative to full scale — not SPL.
How it works
Double-blind ABX loudness test synthesizing exact decibel attenuation levels. Runs 100% locally with binomial p-value significance calculation.
What is Loudness Discrimination?
The Just Noticeable Difference (JND) for sound pressure level represents the minimum change in decibels ($\Delta\text{dB}$) required for a listener to reliably perceive that one sound is louder or quieter than another.
According to Weber’s Law, the perceived intensity of a stimulus is proportional to the relative change rather than the absolute physical magnitude. In acoustics, decibels express this ratio logarithmically:
$$\Delta L = 20 \log_{10}\left(\frac{V_1}{V_2}\right)$$
Perceptual Threshold Guidelines
- 3.0 dB: Clearly audible to virtually all listeners; represents a doubling of electrical power.
- 1.0 dB: The standard benchmark threshold for untrained listeners in quiet conditions.
- 0.5 dB: Accessible to trained recording and mix engineers under headphone monitoring.
- 0.2 dB: Mastering engineer perceptual territory; requires top-tier monitoring and focused concentration.
- 0.1 dB: Generally considered the physical perceptual limit of the human basilar membrane.
Frequently asked questions
What is the smallest volume difference a human ear can detect?
For broadband signals and clean mid-frequency tones (around 1 kHz), the human Just Noticeable Difference (JND) is approximately 0.5 dB to 1.0 dB. Highly trained audio engineers in quiet listening environments can sometimes detect level changes down to 0.2 dB.
Why is level matching critical in audio comparisons?
Due to the Fletcher-Munson equal-loudness contours, listeners perceive slightly louder sound as having punchier bass and brighter treble, even when the frequency response is identical. The ABX blind test removes this bias.
How many trials do I need to prove I can hear the difference?
To achieve statistical significance at α = 0.05, you need at least 5 correct answers out of 5, or 9 correct out of 10. A score of 3 out of 4 has a p-value of 0.31, which is statistically consistent with coin tossing.