Silencing by Motion

What to see

On the right, there is a ro­tat­ing ring con­sist­ing of many coloured discs; best fix­ate the cen­tral cross. No­tice any­thing spe­cial? Well, the ring changes ro­ta­tion di­rec­tion, we all can eas­ily agree on that.

What to do

Press the ‘stop’ but­ton. Un­sur­pris­ingly, the ring re­frains from ro­tat­ing. Now it be­comes ob­vi­ous that all the small disks change their colour; going through rain­bow so to speak.

Start­ing the ring ro­ta­tion now and look­ing closely re­veals that the colour change hap­pens all the time, but it is barely not­i­ca­ble dur­ing ro­ta­tion. This can be demon­strated by press­ing the ‘hue–/hue+’ but­ton dur­ing ro­ta­tion; this has nearly no ef­fect dur­ing ring ro­ta­tion.

In­stead of the ring you can also se­lect to ro­tate a pin­wheel in the back­ground with sim­i­lar ef­fect. So it is not rel­e­vant that the small disks them­selves move to im­pare de­tec­tion of the hue change.

Com­ments

This ef­fect was pub­lished early in 2011 by Jor­dan W. Su­chow and George A. Al­varez from Har­vard Uni­ver­sity. It sub­se­quently won, de­servedly, the “Best Il­lu­sion of the Year Con­test”.

In the paper by Su­chow and Al­varaz and their 2011-VSS poster, not only in­ter­est­ing vari­ants but also sev­eral con­crete mech­a­nisms are dis­cussed and tested, which I need not go into here. An in­tu­itive ex­pla­na­tion: pro­cess­ing global mo­tion pre­sents an over­load, which im­pairs cod­ing of the local de­tails. This is a re­ally new ef­fect, and I think it may be re­lated to “mo­tion blind­ness”.

Sources

Su­chow JW, Al­varez GA (2011) Mo­tion Si­lences Aware­ness of Vi­sual Change. Cur­rent Bi­ol­ogy 21:1–5.
That pro­vides links to demon­stra­tion movies with a num­ber of in­ter­est­ing vari­ants.
You can find them also at their demon­stra­tion page.

Burr D (2011) Vi­sual Per­cep­tion: More Than Meets the Eye. Curr Biol 21:R159–R161 (thought­ful ed­i­to­r­ial)

I am thank­ful to Jor­dan for grace­fully al­low­ing me to pre­sent this demon­stra­tion, closely mod­eled after their pub­li­ca­tion.