First consider a bubble, CD or layer of oil in water... they create interesting colours because of the 'thin film effect' wherein light waves reflecting off the top surface interferes with light reflecting off the bottom surface, either cancelling-out or constructively adding.

If the depth of the thin film could be varied then images could be formed.

The change in depth of the film could be achieved by an array of piezoelectric crystals; each crystal could squeeze the thin film at a particular location. The screen would be 'bathed' in light of narrow bandwidth (eg red). If a piezoelectric crystal is activated the light constructively interferes, hence turning a pixel on. If a piezoelectric crystal is not activated, the light interferes destructively, turning the pixel off.

A small gap between the bottom of a transparent sheet and the top of the array of piezoelectric crystals could be used instead of a 'thin film'.
-- xaviergisz, Jul 15 2004

Why do CDs reflect rainbow colors? http://science.hows....com/question52.htm
[xaviergisz, Oct 05 2004]

Diffraction Gratings with LCD Displays http://lfw.pennnet....ay&ARTICLE_ID=28629
[philmckraken, Oct 05 2004]

Interferometric modulator display http://en.wikipedia...c_modulator_display
looks like this technology is now available [xaviergisz, Mar 12 2008]

Grating light valve https://en.wikipedi...Grating_light_valve
another related technology, which uses electrostatic rather than piezoelectric actuation [notexactly, Mar 18 2018]

This one sounds much more promising than the first one.
Each pixel could be a module, mechanically robust to stop finger pressure (or air pressure) from disrupting the effect, with the air gap corresponding to the colour that you want.
Bath with white light, and relect only the colour required at that pixel.
The advantage would be that it could display any colour, and not just a combination of RGB. +
<Edit> Sorry, I forgot about brightness: how would you control brightness?
-- Ling, Jul 15 2004


Is that link right about colours reflecting off a CD? I thought it was a diffractive effect.

It seems to be an extremely complicated way to produce a colour display (and as you mentioned it's probably better to make it single colour), which are pretty much baked. + for a novel method, - for no clear incentive to produce it
-- scubadooper, Jul 15 2004


[scubadooper], I think you are right. The construction of a CD is approx 1.2mm thick polycarbonate on the bottom, then there is a layer of reflective aluminium on top of that (according to the buried link). 1.2mm is too thick to get the 'Rainbow Oil colour or Newton ring (?)' effect.
-- Ling, Jul 15 2004


I thought so, I'm still amazed at the fact that the pits in a CD are only 110 nanometers deep, people don't realise that everytime they put a CD or DVD into a player they're dealing with nano-engineering.
-- scubadooper, Jul 15 2004


I think natural nano-engineering is far more impressive. Butterfly wings and so on.
-- Ling, Jul 15 2004


I agree, I think the link I provided is a poor (or incorrect) explantion of why a CD reflects rainbow colors but the explanation of thin film interference is OK.

It might be possible to backlight rather than using front-on reflection. This would make a colour display easy... each crystal would be backlit by either a red, green or blue light (and rather than a backlight for each pixel, use one red, one green and one blue light each split and redirected using fibre optics etc.)

[subadooper], incentive for this display is to *redirect* light at each pixel rather than *produce* it at each pixel. This should keep power consumption (and heat) down, make it scalable (ie big or small) and keep it flat and thin. It may have some advantages over LCD...
-- xaviergisz, Jul 15 2004


//each crystal would be backlit by either a red, green or blue light //
? I don't follow that bit. Then you may as well have a standard display behind your display....

I thought your idea was to produce coloured light from incident white light, like an oil film does.
-- Ling, Jul 15 2004


Ling, using incident white light might work pretty well but would have to used in a dark environment (which I was trying to get around).

I agree with you that the backlit version probably wouldn't work... I think I'll stick with the incident white light version.
-- xaviergisz, Jul 15 2004


I think a problem with colours created by interference is that the colour you see will vary depending on the angle at which the display is viewed (see also link).
-- philmckraken, Jul 15 2004


phil, the 'white light' I mentioned is actually a combination of narrow-bandwidth red, green and blue light. The interference effect causes the light to be either reflected or not, ie. no chance of angle-dependent 'rainbow effect'.
-- xaviergisz, Jul 15 2004


[philmckraken] like the link.

[Ling] you're right the structures that nature can make are almost beyond beauty.

[xaviergisz] I'm still not convinced.
-- scubadooper, Jul 15 2004



random, halfbakery