Correlogram Museum

About Correlograms Bibliography Credits

Correlogram Introduction

This web site is a repository for information about the correlogram. The correlogram is a model of auditory perception that explains how the auditory system might process and understand the temporal information in the audio signals we hear. The auditory correlogram was "invented" by J. C. Licklider in 1951, but at the time he could only draw pictures of the representation. More recently the correlogram was popularized by Richard F. Lyon, Malcolm Slaney and Roy Patterson to model and explain pitch perception. Many of the examples presented here were first presented in the Apple Hearing Demo Reel.

To read more about what a correlogram is, and how it is computed go to this page about the correlogram, or to the bibliography.

The web link below (Leonardo) leads to a simple example of the correlogram, showing how it represents sounds along two axes: frequency and time delay.

• Leonardo

Correlogram Examples These examples demonstrate the basic properties of a correlogram. Duda Tones Leonardo's Vowels Leonardo's Pitch Duda Vowels McAdams-Reynolds Oboe Pierce Tones

ASA Examples The Acoustical Society of America (ASA) created a wonderful collection of auditory demonstrations. Some of these demonstrations are useful when thinking about the temporal properties of perception. If you are interested in sound you should have a copy of this audio CD. Ordering information is available from the ASA. ASA 1 - Cancelled Harmonics ASA 9 - Asymmetry Of Masking By Pulsed Tones ASA 10 - Backward And Forward Masking ASA 11 - Pulsation Threshold ASA 12 - Dependence Of Pitch On Intensity ASA 13 - Pitch Salience And Tone Duration ASA 14 - Influence Of Masking Noise On Pitch ASA 15 - Octave Matching ASA 16 - Stretched And Compressed Scales ASA 17 - Frequency Difference Limen Or Jnd ASA 18 - Logarithmic And Linear Frequency Scales ASA 19 - Pitch Streaming ASA 20 - Virtual Pitch ASA 21 - Shift Of Virtual Pitch ASA 22 - Masking Spectral And Virtual Pitch ASA 23 - Virtual Pitch With Random Harmonics ASA 24 - Strike Note Of A Chime ASA 25 - Analytic Vs Synthetic Pitch ASA 26 - Scales With Repetition Pitch ASA 27 - Circularity In Pitch Judgment ASA 28 - Effect Of Spectrum On Timbre ASA 29 - Effect Of Tone Envelope On Timbre ASA 30 - Change In Timbre With Transposition ASA 31 - Tones And Tuning With Stretched Partials ASA 32 - Primary And Secondary Beats ASA 33 - Distortion ASA 34 - Aural Combination Tones ASA 35 - Effect Of Echoes

Auditory Image Examples Most of the correlograms presented here were computed using autocorrelation. While this is a straight forward mathematical algorithm, it is certainly not how the brain would do the calculation. Roy Patterson's Auditory Image Model (AIM) proposes a biologically more realistic computational implementation, based on strobed temporal integration (STI). The videos in this section were computed using an early model of STI. Duda Tones with AIM Leonardo's Vowels with AIM

Miscellaneous Examples These videos show simple examples from Lyon's cochlear modeling. This cochlear model is the basis of the (autocorrelation) correlogram videos shown here. We also have a three-panel correlogram showing the difficulty of sound separation, and a video from an early VLSI chip that computed the correlogram. Gray Scale Propagation Single Tone Propagation Two Tone Propagation Running to You Correlogram Chip