Apologies -to those who noticed- for posting twice about the same darn thing, but I return to the subject armed with marginally more information.
The ugly and disheartening phenomenon I'm facing is rather aptly named, though I cannot say for certain that your average home-theater-projection-technology buff takes into account coherent light's strong need to diffract in response to every edge it encounters. A digital micro-mirror device with tens of thousands of micron-scale mirrors has lots of edges. My projected image really does look as though it's being shone through a window screen.
So far I've tried shaping my input beam several different ways, I've tried a small diffuser just outside the inlet hole and a diffuser just inside the inlet. Shaping the beam in an attempt to mimic the LED output did nothing to reduce the screen door deal. The diffuser just outside the inlet has so far yielded the best results, though at the cost of considerable loss in resolution and a lot of light loss.
Everything is very very compact inside the optical module of this portable projector, and most of the inner workings are inaccessible. A micro-lens array is the closest I've found so far on the internet to a solution, and even if I could manage to pull one of those from my hindquarters, I'd have no idea where or how to insert it into this highly miniaturized optical system.
I truly hope there is a better solution to this dilemma than just blurring the image until the screening isn't (as) noticeable.
I'm very interested in suggestions.
Thanks, everyone.
Laser on DLP "screen door effect"
Laser on DLP "screen door effect"
How big are you blowing it up? Even when we used the TI DMD XGA device we never went bigger than 6 by 8 inches, 15 by 20 cm. If I put my near-sighted eyes close to the screen, or final hologram, I could see the pixels. Mainly it was used for dinky postage stamp size security stuff.
It sounds almost too elementary, but are you sure you're in focus? That's how we could tell when the DMD was in focus, the diffracted artifacts disappeared. Also with a film projector device, there would be interference fringes at borders of objects, but they disappeared when the image was in focus. Something to do with a Fourier transform of a Fourier transform gets you back to the original, although it isn't exactly the Fourier case, but I don't have my reference library in front of me so I can't say exactly what's happening. (Hecht & Zajac or Born & Wolf has it I'm sure!)
In either of the cases above, there were no diffusers. In the Craig Newswanger-engineered CFC stereo set up with the DMD, a spatially filtered beam was incident at the appropriate angle so that the image carrying beam came out normal to the chip. The projecting lens was a pair of achromats used in the typical projector lens formula, don't have my reference books with me to tell what the name of the system is.
For the film version, which was the stereogram machine at the SAIC, a spatially filtered beam, no diffuser, went through the movie film. The projection lenses we used were fixed focal length Canon 35 mm camera lenses, 50 mm for vertical format for portraits, 35 mm fl for horizontal frame filling. Zoom lenses from all the camera manufacturers had so many elements that the internal bouncing made some pretty ugly artifacts.
Speaking of Fourier transforms, something we often discussed but never attempted at CFC was to expose a plate in the image beam path where there were all sorts of patterns that looked like if the right position were selected, and the developed plate were replaced, with the exposed areas developed black to block the pattern of the grid, another spatial filtering technique a la van der Lugt, they would disappear. But we were either too busy or too apathetic to try, maybe you might be.
It sounds almost too elementary, but are you sure you're in focus? That's how we could tell when the DMD was in focus, the diffracted artifacts disappeared. Also with a film projector device, there would be interference fringes at borders of objects, but they disappeared when the image was in focus. Something to do with a Fourier transform of a Fourier transform gets you back to the original, although it isn't exactly the Fourier case, but I don't have my reference library in front of me so I can't say exactly what's happening. (Hecht & Zajac or Born & Wolf has it I'm sure!)
In either of the cases above, there were no diffusers. In the Craig Newswanger-engineered CFC stereo set up with the DMD, a spatially filtered beam was incident at the appropriate angle so that the image carrying beam came out normal to the chip. The projecting lens was a pair of achromats used in the typical projector lens formula, don't have my reference books with me to tell what the name of the system is.
For the film version, which was the stereogram machine at the SAIC, a spatially filtered beam, no diffuser, went through the movie film. The projection lenses we used were fixed focal length Canon 35 mm camera lenses, 50 mm for vertical format for portraits, 35 mm fl for horizontal frame filling. Zoom lenses from all the camera manufacturers had so many elements that the internal bouncing made some pretty ugly artifacts.
Speaking of Fourier transforms, something we often discussed but never attempted at CFC was to expose a plate in the image beam path where there were all sorts of patterns that looked like if the right position were selected, and the developed plate were replaced, with the exposed areas developed black to block the pattern of the grid, another spatial filtering technique a la van der Lugt, they would disappear. But we were either too busy or too apathetic to try, maybe you might be.
"We're the flowers in the dustbin" Sex Pistols
Laser on DLP "screen door effect"
I mean not might be!
"We're the flowers in the dustbin" Sex Pistols
Laser on DLP "screen door effect"
I just read your earlier post, and it appears that you are just diffracting off the mirrors without a lens to focus the image onto a screen. Try using the lens from the projector, or does this one use one at all? Or camera lenses, or even single element positive lenses, etc.
"We're the flowers in the dustbin" Sex Pistols
-
Justin W
Laser on DLP "screen door effect"
Hi Ed!
I'm not using any focusing lenses that would be additional to what the little unit has inside (what I would guess to be a set of matched singlets in a plastic mount that slides on three metal pins to focus the unit). It's all so darn tiny.
The green LED is intended to shine in right next to where the projector outputs. Green light travels straight back through the micro-lens deal and finds its way to the DMD in the back, to be collected, reflected, then gathered by the focusing slider and outputted. The red/blue LED shines in perpendicular to the green, is reflected 90 deg by a couple of thin glass thingers (that the green is in fact transmitting on through), then through the micro-lens and to the DLP chip.
Each of the LED inlet holes has a wee lens. I've tried coating that lens with raw beam. I've tried a spatial-filtered 5x, 20x. I've tried a converging beam. All yielded the same result: a bright laser image of whatever the DLP was displaying with a very noticeable pattern of mesh-like shadows. This is the case regardless of projected image size or focus. I can defocus enough to reduce the screen pattern, but by then the image is a mere blur.
I'm not using any focusing lenses that would be additional to what the little unit has inside (what I would guess to be a set of matched singlets in a plastic mount that slides on three metal pins to focus the unit). It's all so darn tiny.
The green LED is intended to shine in right next to where the projector outputs. Green light travels straight back through the micro-lens deal and finds its way to the DMD in the back, to be collected, reflected, then gathered by the focusing slider and outputted. The red/blue LED shines in perpendicular to the green, is reflected 90 deg by a couple of thin glass thingers (that the green is in fact transmitting on through), then through the micro-lens and to the DLP chip.
Each of the LED inlet holes has a wee lens. I've tried coating that lens with raw beam. I've tried a spatial-filtered 5x, 20x. I've tried a converging beam. All yielded the same result: a bright laser image of whatever the DLP was displaying with a very noticeable pattern of mesh-like shadows. This is the case regardless of projected image size or focus. I can defocus enough to reduce the screen pattern, but by then the image is a mere blur.
Laser on DLP "screen door effect"
The light you’re putting thru is spatially coherent, try another light source? LED, laser pointer (non coherent) and see what you get? Fiber optic? Try different source and see if the interference goes away. This way you’ll know if it is truly a diffraction issue or optical train issue ....................
-
Justin W
Laser on DLP "screen door effect"
A green LED aimed into the hole gives me a mesh-less image.
-
PinkysBrain
Laser on DLP "screen door effect"
Don't you want an always in focus projection rather than an imaging of the DLP? If you simply have a purely diverging beam coming from the DLP you can also focus it through an aperture to get rid of all the light diffracted at the DLP.
-
Justin W
Laser on DLP "screen door effect"
I haven't removed the DLP from its optical module or anything like that. It still rests in its little housing with all of the optics unmodified save for the removal of the LED light sources that plugged into input holes on the exterior.
-
Johnfp
Laser on DLP "screen door effect"
To be honest with you. I assume you are planning on making stereograms. WIth that being said, once you make the sets of slits and then the final hologram, a healthy person (with two eyeballs that work) in viewing the finaly hologram, in essence by viewing through two of the slits at one time to see the 3D image, will be looking at two different images at once from slightly different angles. So I would imaging most of that screen door effect will not be seen. In essence there will be rare occasions that the two dark boxes (the wires on a screen door) will overlap exactly.
I am not sure how much of the negative effect will go away but I can bet it will dimminish.
I am not sure how much of the negative effect will go away but I can bet it will dimminish.