The diffusers are probably structured light diffusers, like those made by Luminit. Optotune works but is expensive and also uses a similar diffuser structure on the membrane.
The diffusers are probably structured light diffusers, like those made by Luminit. Optotune works but is expensive and also uses a similar diffuser structure on the membrane.
I'm still not sure how a diffuser will not destroy the near perfect collimation required to achieve a focus free projection.
Think about what causes speckle: it is individual waves of light interfering with one another, producing constructive and destructive interference.
Constructive = multiple peaks = bright spot.
Destructive = peaks and troughs = dark spot.
Realize that the slightest movement, on the order of just a few hundred nanometers, is enough to begin to change this interference pattern.
Now imagine you have the ability to vibrate the entire beam ever so slightly in a random pattern at very high speeds so that you can average out the bright and dim spots over time.
That would be a way to get rid of speckle without changing the divergence at all.
Alas, it's not very practical to mount the diode on some sort of "micron-scale shaker table"... (Reliability might prove to be an issue as well.)
The next best thing is to diffract the beam *slightly* so that you cause the interference patterns to change/move as the diffraction element is moved. Now keep that diffraction element moving quickly and over time once again the bright and dim spots will average out. But remember that we're dealing with very small scales here.
Now, if you just use a piece of glass that has been sand-blasted, then yeah - you're going to greatly increase the divergence, which is bad. But that is overkill.
Instead you want to use a lightly-diffractive element that yields just enough variation to allow you to average out the speckle without significantly altering your divergence.
Adam
Okay, makes sense. Thanks for the explanation.