Mach Banding Illusions and our Visual Perception of Gradients

Have you ever noticed the appearance of an extra bright area at the edge of a gradient, or a darker looking area at the border between shapes of differing brightness?

Take a look at this video:

Mach Banding Illusions demonstrate some interesting aspects of how our visual systems process information.  We often describe the eye like a camera, and in certain ways, it is an accurate model.  Unfortunately our visual system is not a camera, and we don't always interpret light in direct accordance with photon reception.  This demonstration will show the effects Mach Banding, and the process of Lateral Inhibition which induces the phenomenon.  Gradients like the one in the example don't occur naturally in nature.  Gradient detection in the real-world likely evolves from our visual system attempting to use lighting gradients as indicators of an object's depth/shape.   We use software to induce and enhance the Mach Banding effect for easier identification.

This process has a bigger impact than one may first imagine.  In the Healthcare Industry, for example, Mach Banding can create false positives in medical imaging/diagnostics due to incorrectly perceived levels brightness in gradients.  Lives have been lost due to incorrect analysis/interpretations of perceived highlighting, increasing the contrast and sharpness of gradient edges, introduced by gradient Mach Banding.  With appropriate knowledge of Mach Banding, healthcare professionals can avoid misdiagnosis by referencing the original image through waveform, vectorscopes, or histogram monitoring to ensure that the perceived brightness accurately reflects actual image brightness.

  • How do you think that graphic design could be used to emphasize this visual effect?  
  • Can Mach Banding be used to emphasize/accent 3d modeling/shading?
  • Could you imaging using projection mapping gradients to accentuate or de-accentuate form details/edges?

You can Find a link to the OSX Mach Banding Simulation Application we made here:

1.  Download, extract, and run the app(it looks best in Fullscreen due to color sampling/scaling).

Bonus Easter Egg:  You can press 'h' in order to view a histogram of the rendered image which shows the luminance of the image live, in real-time.

Further Reading:

Salience of Mach bands as a function of the relative linearity of the luminance gradient stimulus.