Verbose
I know what u are saying verbose but let the guy experiment abit eh? with a surface mirror and a page magnifier it could work as the tollerences with the image are so small compared with a normal mirror and with the light going it through it the first time how can the image be focused on the mirror to wash it out? it will be out of focus the first time cos its so close to the mirror and therfore the light will be magnified and the main lens focuses to the front of the lcd very well lit up, the only prob ocurs is if u focus to the mirror not the front of the lcd, anyway lets see what happends, as i say i ionow what u are says verbose but i think with the surface mirror it could work.
Trev
I know what u are saying verbose but let the guy experiment abit eh? with a surface mirror and a page magnifier it could work as the tollerences with the image are so small compared with a normal mirror and with the light going it through it the first time how can the image be focused on the mirror to wash it out? it will be out of focus the first time cos its so close to the mirror and therfore the light will be magnified and the main lens focuses to the front of the lcd very well lit up, the only prob ocurs is if u focus to the mirror not the front of the lcd, anyway lets see what happends, as i say i ionow what u are says verbose but i think with the surface mirror it could work.
Trev
Your right. I dont mean to put everyones ideas down and it is a only a fresnel. I cant wait to see the results of this project if anyone tries it.
i would try it too, but i finished with my projector now, and it's awesome, as soon as i get a digi cam i will post pictures...
100" image
400W MH bulb
PC TFT Monitor
self build casing, and it works great!!!
100" image
400W MH bulb
PC TFT Monitor
self build casing, and it works great!!!
well, with help from my friends, I got my projector going tonight. What did they want to watch? Matrix, of course. This was a preliminary run to test the design. It works. Just need to seal light leaks, work on lense positioning, work on brightness, install fans, etc. Good image, but it can be so much better.
I am getting some haloing ( light bleeding in one direction) around words and such. Any suggestions?
Ken
I am getting some haloing ( light bleeding in one direction) around words and such. Any suggestions?
Ken
You know... I really dont think much. I just started a class in video editing. Guess what??? I have acess to a top of the line digi-cams... The problem? No more diy projector... 😡
On a reflective LCD system you absolutely don't want anything between the LCD screen and the mirror (except the built-in polarizing filter or a replacement one).
You want the light to go into each pixel as nearly perpendicular as possible, undergo it polarization and color filtration hit the mirror and come right back out.
The mirror should be a surface mirror, but not because of multiple reflections, but for reasons of shortening the distance the light travels between its two trips through the pixel.
A reflection perpendicular to a mirror will not have any multiple reflections to speak of. That problem only occurs when the light hits the mirror at an angle and you get two reflections--one from the glass and one from the metallic reflector on the back of the glass (the silver part).
As far as such a thing as a "one-way mirror" or "two-way mirror"--pretend that such a thing does not exist. They don't work the way you think they do if you are planning on using one like this. They block half of the light and pass the other half. They are 50% silvered on one side.
In use, they rely on a dark room / light room combination for their effect. Not much different than looking out you living room windo at night when the living room lights are on.
One problem with using a transmissive LCD panel as a reflective panel (for those of you trying this design with a panel designed to be backlit) is that the black lines between the pixels will look very different. On a transmissive panel, the black lines are opaque and get their "blackness" from blocking light from coming through. When you instead shine a light on the front of this panel, these lines will be illuminated, and tend to be a BRIGHT gray. (picture opening a window at night and shining a flashlight through the screen. You will lose a ton of contrast in the resulting image because the reflective light from the screen door will be scattered.
Slize, your design from 04/03/03 labelled "new style" will work using a reflective style LCD panel, but you need to modify it one of two directions:
[1] get the light landing on the panel more parallel and use a second lens between the panel and the final lens. It is possible to do this while sending the light throught the same lens toward the LCD (in the other direction). I'll send along a picture.
[2] go for the super diffuse lighting bombardment of the LCD panel and use a really big lens to capture all of the light coming back from the LCD to get it to the screen.
On other point about using a reflective screen-- If the arrangement is all in one box like this there will be a loss of color contrast as well. Here's an example: Imagine that you are showing an image on your screen like a Canadian flag. Lots of red and white. Inside the box, you will have lots of extra red and white bouncing around and this stray light, if it was only white could further illuminate your panel. But with a lot of red in it, it's gong to shine a bunch of extra red light on your panel, which will tend to cast a red glow on the image a bit. (As if the color temp on your source light was just lowered.)
You want the light to go into each pixel as nearly perpendicular as possible, undergo it polarization and color filtration hit the mirror and come right back out.
The mirror should be a surface mirror, but not because of multiple reflections, but for reasons of shortening the distance the light travels between its two trips through the pixel.
A reflection perpendicular to a mirror will not have any multiple reflections to speak of. That problem only occurs when the light hits the mirror at an angle and you get two reflections--one from the glass and one from the metallic reflector on the back of the glass (the silver part).
As far as such a thing as a "one-way mirror" or "two-way mirror"--pretend that such a thing does not exist. They don't work the way you think they do if you are planning on using one like this. They block half of the light and pass the other half. They are 50% silvered on one side.
In use, they rely on a dark room / light room combination for their effect. Not much different than looking out you living room windo at night when the living room lights are on.
One problem with using a transmissive LCD panel as a reflective panel (for those of you trying this design with a panel designed to be backlit) is that the black lines between the pixels will look very different. On a transmissive panel, the black lines are opaque and get their "blackness" from blocking light from coming through. When you instead shine a light on the front of this panel, these lines will be illuminated, and tend to be a BRIGHT gray. (picture opening a window at night and shining a flashlight through the screen. You will lose a ton of contrast in the resulting image because the reflective light from the screen door will be scattered.
Slize, your design from 04/03/03 labelled "new style" will work using a reflective style LCD panel, but you need to modify it one of two directions:
[1] get the light landing on the panel more parallel and use a second lens between the panel and the final lens. It is possible to do this while sending the light throught the same lens toward the LCD (in the other direction). I'll send along a picture.
[2] go for the super diffuse lighting bombardment of the LCD panel and use a really big lens to capture all of the light coming back from the LCD to get it to the screen.
On other point about using a reflective screen-- If the arrangement is all in one box like this there will be a loss of color contrast as well. Here's an example: Imagine that you are showing an image on your screen like a Canadian flag. Lots of red and white. Inside the box, you will have lots of extra red and white bouncing around and this stray light, if it was only white could further illuminate your panel. But with a lot of red in it, it's gong to shine a bunch of extra red light on your panel, which will tend to cast a red glow on the image a bit. (As if the color temp on your source light was just lowered.)
A Note on Light Efficiency
Since the intention is to get a brighter (and larger) screen image, while maintaining picture quality, I thought I would toss in some data (and design viewpoints) regarding the treatment of light efficiency.
It would seem that, electricity being so cheap and all, that the efficiency of the system can be regarded lightly, but I would assert that this is not the case.
Any light emitted by the lamp that does not show up on the screen is wasted as heat somewhere. Some of it can leak out as well and go wrong places, lowering the contrast of the viewed image.
For each loss, if we can't recover the energy as useful light, we should at least control where it goes. IR light (for instance) getting absorbed by a filter is better than IR light being absorbed by semiconductors and LCD's.
So, let's follow the light path:
You have a lamp that is rated at (for example) 10,000 lumens. How much of that is visible light and how much is outside the one-octave range of human eyeballs? For our purposes, any component of the light in the IR and UV range is at best useless and at worst harmful to the components. There are passive substances that can recycle UV and IR light into visible light. They tend to be either expensive or useless. (In the useless category, think of a glow-in-the-dark frisbee, you can shine invisible light on it in a dark room and it will glow. That's the same principle that the expensive devices use: throw photons at something to raise its energy level. As the individual atoms drop back to their non-excited states the emit different-colored photons, in the visible spectrum.)
So, let's assume that we lose about one third of our energy to UV and IR light that wouldn't entertain us if it made it to the screen anyway. The thing to do here is first, find out about the emission spectra of the available lamps (some will have a lot more visible light than others), and second, proactively waste this light by shunting or filtering it out before it gets too far from the lamp.
Our 10,000 lumens is now down to about 6667 lumens.
The efficiency of our reflectors can be made really high, but since they tend to be made from rather shallow parabolas, lots of light is going to miss the target. We'll assume that some light is bouncing around non-parallel (for you LOA guys, this will be a TON of light) and will miss the panel, or strike it so obliquely that it can't go through. All of the designs I have seen using the point-source/parabolic method are casting a circle of parallel light at a rectangular target and the area outside the rectangle just gets masked off. That 'circle-outside-of-the-rectangle' we'll call one third of the light. This depends on the aspect ratio and how much bigger the circle is than the square, but one-third will be a good working estimate. (If your mask is a mirror with a panel-sized hole in the middle, score extra points here, since the returning light has some chance of eventually returning and becoming useful).
Our 6667 lumens just went down to 4444.
Next, we start to go through the LCD. First, the polarizer.
A perfect polarizer would cut the volume of light by about half. The cheap films used on LCD's will cut off about 2/3. Can we recycle the light? Yep. It's expensive or complex. You could work out a reflection-off-of-glass scheme that would convert all of the light to the correct polarity and even remove the first polarizer from the panel. It's doable, but not easy. (Do a search on "polarized" and "Brewster's Angle" if interested in persuing this.)
Our polarizing film just took our 4444 lumens and knocked it down to 1481.
Now the light is going through the liquid crystal itself. We have already accounted for the waste of the non-perpendicular light (for the point-source/parabola guys, not the LOA'ers), but we haven't accounted for reflection off of the panel. Shine perfectly polarized light on a panel and some light will bounce back at you. We'll be optimistic and call this reflective loss only 10%.
Our 1481 lumens just became 1333.
Here comes a whammy. (stick with me here...)
Now out white light goes through the panels Red, Green and Blue filters. All of the white light going through a red filter, gets stripped of its blue and green. The same goes for the other two colors. With a perfect set of filters this loss would be 2/3 of our light. We'll say our RGB filters are only 10% less than perfect. We'll also say that the "screen door" lines between the filters are only 10% of the total area of the panel.
Our perfect filters would have taken 2/3 of our light (and converted it to heat), so our 1333 lumens became 444 lumens. Our 10% less-than-perfect filters cost us another 10%, so we are at 400 lumens. The 10% for the "screen door" brings us down to 360 lumens.
One more polarizing film on the way out of the panel... But this one doesn't hurt so much, because our light has become 'twisted' in polarity to match up to it. We take 10% off for less-than-perfect alignment and filtering.
We have 324 lumens remaining that still have a chance at being a viewable image.
To get out of the box, our light will need to go through at least one lens (or curved mirror, but I haven't seen that tried yet). The most efficient designs are using a combination of a near fresnel and a far conventional lens, to avoid having to use a BIG lens. Each of these surfaces has some unwanted reflection. Each of these materials have some transmissive losses (and get slightly warm) and each have some surface imperfections which cause a small amount of stray light. For all of these, we'll say we get 90% of the light to go through (not get reflected or absorbed) and land on the screen exactly where we want it.
Our 10,000 lumens have been reduced to 292 lumens and we haven't hit the screen yet.
For everyone that ever asked "where in the hell did all of my light go?"
There are lots of points above where these losses could be cut drastically. The current fix is to simply dump in a LOT more light. You just need to make sure that the wasted light above (97.1% of the light we started with) doesn't leak out and cause a bigger contrast problem, and doesn't damage your internal components.
Since the intention is to get a brighter (and larger) screen image, while maintaining picture quality, I thought I would toss in some data (and design viewpoints) regarding the treatment of light efficiency.
It would seem that, electricity being so cheap and all, that the efficiency of the system can be regarded lightly, but I would assert that this is not the case.
Any light emitted by the lamp that does not show up on the screen is wasted as heat somewhere. Some of it can leak out as well and go wrong places, lowering the contrast of the viewed image.
For each loss, if we can't recover the energy as useful light, we should at least control where it goes. IR light (for instance) getting absorbed by a filter is better than IR light being absorbed by semiconductors and LCD's.
So, let's follow the light path:
You have a lamp that is rated at (for example) 10,000 lumens. How much of that is visible light and how much is outside the one-octave range of human eyeballs? For our purposes, any component of the light in the IR and UV range is at best useless and at worst harmful to the components. There are passive substances that can recycle UV and IR light into visible light. They tend to be either expensive or useless. (In the useless category, think of a glow-in-the-dark frisbee, you can shine invisible light on it in a dark room and it will glow. That's the same principle that the expensive devices use: throw photons at something to raise its energy level. As the individual atoms drop back to their non-excited states the emit different-colored photons, in the visible spectrum.)
So, let's assume that we lose about one third of our energy to UV and IR light that wouldn't entertain us if it made it to the screen anyway. The thing to do here is first, find out about the emission spectra of the available lamps (some will have a lot more visible light than others), and second, proactively waste this light by shunting or filtering it out before it gets too far from the lamp.
Our 10,000 lumens is now down to about 6667 lumens.
The efficiency of our reflectors can be made really high, but since they tend to be made from rather shallow parabolas, lots of light is going to miss the target. We'll assume that some light is bouncing around non-parallel (for you LOA guys, this will be a TON of light) and will miss the panel, or strike it so obliquely that it can't go through. All of the designs I have seen using the point-source/parabolic method are casting a circle of parallel light at a rectangular target and the area outside the rectangle just gets masked off. That 'circle-outside-of-the-rectangle' we'll call one third of the light. This depends on the aspect ratio and how much bigger the circle is than the square, but one-third will be a good working estimate. (If your mask is a mirror with a panel-sized hole in the middle, score extra points here, since the returning light has some chance of eventually returning and becoming useful).
Our 6667 lumens just went down to 4444.
Next, we start to go through the LCD. First, the polarizer.
A perfect polarizer would cut the volume of light by about half. The cheap films used on LCD's will cut off about 2/3. Can we recycle the light? Yep. It's expensive or complex. You could work out a reflection-off-of-glass scheme that would convert all of the light to the correct polarity and even remove the first polarizer from the panel. It's doable, but not easy. (Do a search on "polarized" and "Brewster's Angle" if interested in persuing this.)
Our polarizing film just took our 4444 lumens and knocked it down to 1481.
Now the light is going through the liquid crystal itself. We have already accounted for the waste of the non-perpendicular light (for the point-source/parabola guys, not the LOA'ers), but we haven't accounted for reflection off of the panel. Shine perfectly polarized light on a panel and some light will bounce back at you. We'll be optimistic and call this reflective loss only 10%.
Our 1481 lumens just became 1333.
Here comes a whammy. (stick with me here...)
Now out white light goes through the panels Red, Green and Blue filters. All of the white light going through a red filter, gets stripped of its blue and green. The same goes for the other two colors. With a perfect set of filters this loss would be 2/3 of our light. We'll say our RGB filters are only 10% less than perfect. We'll also say that the "screen door" lines between the filters are only 10% of the total area of the panel.
Our perfect filters would have taken 2/3 of our light (and converted it to heat), so our 1333 lumens became 444 lumens. Our 10% less-than-perfect filters cost us another 10%, so we are at 400 lumens. The 10% for the "screen door" brings us down to 360 lumens.
One more polarizing film on the way out of the panel... But this one doesn't hurt so much, because our light has become 'twisted' in polarity to match up to it. We take 10% off for less-than-perfect alignment and filtering.
We have 324 lumens remaining that still have a chance at being a viewable image.
To get out of the box, our light will need to go through at least one lens (or curved mirror, but I haven't seen that tried yet). The most efficient designs are using a combination of a near fresnel and a far conventional lens, to avoid having to use a BIG lens. Each of these surfaces has some unwanted reflection. Each of these materials have some transmissive losses (and get slightly warm) and each have some surface imperfections which cause a small amount of stray light. For all of these, we'll say we get 90% of the light to go through (not get reflected or absorbed) and land on the screen exactly where we want it.
Our 10,000 lumens have been reduced to 292 lumens and we haven't hit the screen yet.
For everyone that ever asked "where in the hell did all of my light go?"
There are lots of points above where these losses could be cut drastically. The current fix is to simply dump in a LOT more light. You just need to make sure that the wasted light above (97.1% of the light we started with) doesn't leak out and cause a bigger contrast problem, and doesn't damage your internal components.
handling the misunderstandings
There are a few mis-conceptions about light that should be cleared up. After reading 1300 posts here and on the "part one" version of this thread, I thought a few people might be interested in some of this data.
"Lighting a small LCD is more efficient than lighting a large one."
Nope. Not true. The only real efficiency gained by using a small LCD panel is that the exit lens system can use smaller lenses, and the light source can be aimed at a smaller "target."
A small LCD panel, however, has to be enlarged MUCH more by the lens system than a large panel. You can get by with smaller diameter lenses, but they better be higher quality. This includes not only surface imperfections, but chromatic abberation correction. A single lens (made of a single type of glass) is really just a rounded prism of sorts. It bends some colors of light more than others. When you magnify more, this separation of color is proportionally more pronounced. For a given screen size, a 2 inch LCD would need to be magnified 6 times as much as a 12 inch panel.
Afraid of damaging an LCD with too much light? You should know that you will need to put 36 times as much light (per square inch) through a 2 inch panel as through a 12 inch panel.
The guys doing the OHP/big panel with a near-fresnel and a distant small lens system are getting the benefits of both large and small LCD systems (assuming a high-quality fresnel is employed).
Small LCD panels will typically also have a much more pronounced screen-door effect, since the ratio of pixel space to divider space is much worse (in available, inexpensive panels).
Also, (and this is a biggie for the "can't live with the screen door" crowd) a large panel can easily have enough extra resolution to "oversample" the pixels using line-doubling, or even line-tripling or quadrupling. When this is employed, each screen "hole" is divided up into smaller "holes," so the effect is much less pronounced.
So, even when you find that 2 inch "dream panel" with full HDTV resolution, think twice. There are drawbacks.
"Light spreads out and gets weaker with distance"
Waste that datum right now. Replace it with this one:
"Light travels in straight lines" and another "Parallel lines don't spread out" and "Light doesn't get weaker unless something converts it into something else (like heat)"
So, if you are worried about having you light source too far from your LCD, but it (and its reflector) are putting out parallel lines, stop worrying. If you have a really good parabolic reflector and a point source lamp, you could put it a on the moon (properly aimed at your LCD on earth, and it would work just the same, as long as nothing was between them to absorb or scatter you light.
You can make a projector that keeps all of the light in parallel rays all of the way from the lamp reflector to the final lens, and the spacing between all of the elements in between would be a non-issue. (Your final lens(es) would need to be as big as your panel, but all of the placements get much simpler.)
By the way, clean air doesn't interfere with light too much at all, (except when it is unevenly hot and cold.) The size of your projector might get out of hand, but don't worry about going through "extra air" between elements.
"Lenses bend light"
This one is a bit of a subtle distinction. Light will bend when going from one medium to another if the two materials have a different index of refraction. The difference between the index of refraction of the two materials will also delineate how much the light will get bent. So, light going from air to glass gets bent a lot. If you took that lens under water if would bend light less. Out in space, it would bend light more.
And, all of the bending happens at the surface where the two materials meet each other. The light bends when it leaves the air and enters the glass, not inside of the glass, and bends again when it leaves the glass and enters the air.
How much the light gets bent is also determined by the angle that the ray of light crosses from one material into another. If light hits the glass perpendicular to the surface where it enters, it will not get bent at all.
That also explains how a Fresnel lens works-- it's nearly all surface and no middle, but on a lens, the surfaces do all of the work. (By the way, if you start designing a projection system that uses lasers, stay away from fresnels; they won't work well with monochromatic light.)
And a flat pane of glass is really only a special type of lens. It does bend the light entering it, but the light gets bent the other direction when it goes out the other side. The ray of light does get shifted over a little bit, but will always come out of the other side in parallel with the ray that went in. With really thick glass, it gets shifted over a bunch. Try looking though the thick bullet-proof stuff at the bank and you'll see what I mean.
"You could use a one-way mirror here..."
A one-way mirror (as popularly imagined) does not exist. Light doesn't care which direction it goes through a piece of glass. Imagine taking this magic one-way mirror glass and making a hollow ball out of it, with the shiny side on the inside. You could just shine light at it all day long and all of the light would go inside and bounce around forever and never come out.
That would be fun. You wouldn't be able to see it (since the light would bounce back at you) and it would just be a really black ball. Then one day you poke a hole in it, and BLAMMO! all of the stored-up light finally escapes.
Find me some of this sort of glass, and I'll get you on TV and make you rich. I promise.
One-way mirrors in police interrogations rooms, stores and even in dressing rooms are much less magic. They are half-silvered so that half of the light goes through (in both directions). They rely on the viewer being in a darker room and the person being viewed being in a brighter room.
There are, however types of glass with a thin film applied to them that will pass certain colors (frequencies) of light and not others. This is where we get the semi-magical "cold mirrors" and "hot mirrors." Be careful in ordering and using these because their effects only work when the light hits them at a specific angle (nearly always 45 degrees or 90 degrees). If you don't put parallel light rays through a cold mirror, it ain't gonna work right.
"Some types of glass does not have chromatic aberration."
Nope. Still no magic glass out there. Anything that refracts light will bend some colors more than others. All prisms, all lenses, all drops of water, a bubble of air underwater, etc.
A color-correcting lens is actually made up of two lenses made of two types of glass operating together. The two materials have two different indexes of refraction, so they mess up the color by different amounts. Typically, one lens is concave and one is convex, so the color problems from the two lenses cancel each other out.
By the way, you can do almost anything with a curved mirror that you can do with a lens and mirrors don't have color problems; they "bend" all frequencies of light by the exact same amount.
"An LCD panel will still work if I take off the polarizing file from the font or the back."
Not unless you replace it somewhere. You could wear polarized sunglasses while watching it. If fact, no one else would even see the image on the screen, until they put on their polaroids.
You could put a piece of polarized material elsewhere in the light path (as long as it is on the same side of the LCD as the one you removed). It could go at the light source, or at the lens.
You could also use something else to polarize the light. At the correct angle you could bounce the light off of a piece of glass or a bucket of water, or make it go through certain crystals. Lots of things can polarize light. Light why polarized sunglasses were invented, they cut down on the "glare" of reflected light bouncing off of glass and water, because they block light polarized in a certain direction.
When light comes out of a LCD screen, it is polarized. If you put on your polarized sunglasses and tilt your head at a 45 degree angle, the picture on the screen goes nearly black.
That also why nearly all LCD panels have their polarization diagonally, rather than vertically or horizontally; so people can still use them with their sunglasses on.
All that an LCD screen really does is twist the polarization of the light. A white pixel is a twisted pixel. Light that came in that little hole in the back way polarized one way, then twisted 90 degrees so that it could get out through the front.
This also give of the contrast ratio problem with LCD's. The darkest a pixel can ever be is the amount that polarized light from the back gets blocked by the film on the front. If you were to hold up two sheets of polaroid flim and look through them while turning one of them, you would see it get darker and lighter. At it darkest, you can still se throught it somewhat, and at the lightest, it still blocks a bunch of light. The difference between the darkest and the lightest is what defines the LCD's rather wimpy contrast ratio. And that why the blackest of the black part of your movies look pretty darn gray.
If you take off the front and back polarizing films from your LCD panel and look through it, all of the light will come through. All you will see is the "screen door."
There are a few mis-conceptions about light that should be cleared up. After reading 1300 posts here and on the "part one" version of this thread, I thought a few people might be interested in some of this data.
"Lighting a small LCD is more efficient than lighting a large one."
Nope. Not true. The only real efficiency gained by using a small LCD panel is that the exit lens system can use smaller lenses, and the light source can be aimed at a smaller "target."
A small LCD panel, however, has to be enlarged MUCH more by the lens system than a large panel. You can get by with smaller diameter lenses, but they better be higher quality. This includes not only surface imperfections, but chromatic abberation correction. A single lens (made of a single type of glass) is really just a rounded prism of sorts. It bends some colors of light more than others. When you magnify more, this separation of color is proportionally more pronounced. For a given screen size, a 2 inch LCD would need to be magnified 6 times as much as a 12 inch panel.
Afraid of damaging an LCD with too much light? You should know that you will need to put 36 times as much light (per square inch) through a 2 inch panel as through a 12 inch panel.
The guys doing the OHP/big panel with a near-fresnel and a distant small lens system are getting the benefits of both large and small LCD systems (assuming a high-quality fresnel is employed).
Small LCD panels will typically also have a much more pronounced screen-door effect, since the ratio of pixel space to divider space is much worse (in available, inexpensive panels).
Also, (and this is a biggie for the "can't live with the screen door" crowd) a large panel can easily have enough extra resolution to "oversample" the pixels using line-doubling, or even line-tripling or quadrupling. When this is employed, each screen "hole" is divided up into smaller "holes," so the effect is much less pronounced.
So, even when you find that 2 inch "dream panel" with full HDTV resolution, think twice. There are drawbacks.
"Light spreads out and gets weaker with distance"
Waste that datum right now. Replace it with this one:
"Light travels in straight lines" and another "Parallel lines don't spread out" and "Light doesn't get weaker unless something converts it into something else (like heat)"
So, if you are worried about having you light source too far from your LCD, but it (and its reflector) are putting out parallel lines, stop worrying. If you have a really good parabolic reflector and a point source lamp, you could put it a on the moon (properly aimed at your LCD on earth, and it would work just the same, as long as nothing was between them to absorb or scatter you light.
You can make a projector that keeps all of the light in parallel rays all of the way from the lamp reflector to the final lens, and the spacing between all of the elements in between would be a non-issue. (Your final lens(es) would need to be as big as your panel, but all of the placements get much simpler.)
By the way, clean air doesn't interfere with light too much at all, (except when it is unevenly hot and cold.) The size of your projector might get out of hand, but don't worry about going through "extra air" between elements.
"Lenses bend light"
This one is a bit of a subtle distinction. Light will bend when going from one medium to another if the two materials have a different index of refraction. The difference between the index of refraction of the two materials will also delineate how much the light will get bent. So, light going from air to glass gets bent a lot. If you took that lens under water if would bend light less. Out in space, it would bend light more.
And, all of the bending happens at the surface where the two materials meet each other. The light bends when it leaves the air and enters the glass, not inside of the glass, and bends again when it leaves the glass and enters the air.
How much the light gets bent is also determined by the angle that the ray of light crosses from one material into another. If light hits the glass perpendicular to the surface where it enters, it will not get bent at all.
That also explains how a Fresnel lens works-- it's nearly all surface and no middle, but on a lens, the surfaces do all of the work. (By the way, if you start designing a projection system that uses lasers, stay away from fresnels; they won't work well with monochromatic light.)
And a flat pane of glass is really only a special type of lens. It does bend the light entering it, but the light gets bent the other direction when it goes out the other side. The ray of light does get shifted over a little bit, but will always come out of the other side in parallel with the ray that went in. With really thick glass, it gets shifted over a bunch. Try looking though the thick bullet-proof stuff at the bank and you'll see what I mean.
"You could use a one-way mirror here..."
A one-way mirror (as popularly imagined) does not exist. Light doesn't care which direction it goes through a piece of glass. Imagine taking this magic one-way mirror glass and making a hollow ball out of it, with the shiny side on the inside. You could just shine light at it all day long and all of the light would go inside and bounce around forever and never come out.
That would be fun. You wouldn't be able to see it (since the light would bounce back at you) and it would just be a really black ball. Then one day you poke a hole in it, and BLAMMO! all of the stored-up light finally escapes.
Find me some of this sort of glass, and I'll get you on TV and make you rich. I promise.
One-way mirrors in police interrogations rooms, stores and even in dressing rooms are much less magic. They are half-silvered so that half of the light goes through (in both directions). They rely on the viewer being in a darker room and the person being viewed being in a brighter room.
There are, however types of glass with a thin film applied to them that will pass certain colors (frequencies) of light and not others. This is where we get the semi-magical "cold mirrors" and "hot mirrors." Be careful in ordering and using these because their effects only work when the light hits them at a specific angle (nearly always 45 degrees or 90 degrees). If you don't put parallel light rays through a cold mirror, it ain't gonna work right.
"Some types of glass does not have chromatic aberration."
Nope. Still no magic glass out there. Anything that refracts light will bend some colors more than others. All prisms, all lenses, all drops of water, a bubble of air underwater, etc.
A color-correcting lens is actually made up of two lenses made of two types of glass operating together. The two materials have two different indexes of refraction, so they mess up the color by different amounts. Typically, one lens is concave and one is convex, so the color problems from the two lenses cancel each other out.
By the way, you can do almost anything with a curved mirror that you can do with a lens and mirrors don't have color problems; they "bend" all frequencies of light by the exact same amount.
"An LCD panel will still work if I take off the polarizing file from the font or the back."
Not unless you replace it somewhere. You could wear polarized sunglasses while watching it. If fact, no one else would even see the image on the screen, until they put on their polaroids.
You could put a piece of polarized material elsewhere in the light path (as long as it is on the same side of the LCD as the one you removed). It could go at the light source, or at the lens.
You could also use something else to polarize the light. At the correct angle you could bounce the light off of a piece of glass or a bucket of water, or make it go through certain crystals. Lots of things can polarize light. Light why polarized sunglasses were invented, they cut down on the "glare" of reflected light bouncing off of glass and water, because they block light polarized in a certain direction.
When light comes out of a LCD screen, it is polarized. If you put on your polarized sunglasses and tilt your head at a 45 degree angle, the picture on the screen goes nearly black.
That also why nearly all LCD panels have their polarization diagonally, rather than vertically or horizontally; so people can still use them with their sunglasses on.
All that an LCD screen really does is twist the polarization of the light. A white pixel is a twisted pixel. Light that came in that little hole in the back way polarized one way, then twisted 90 degrees so that it could get out through the front.
This also give of the contrast ratio problem with LCD's. The darkest a pixel can ever be is the amount that polarized light from the back gets blocked by the film on the front. If you were to hold up two sheets of polaroid flim and look through them while turning one of them, you would see it get darker and lighter. At it darkest, you can still se throught it somewhat, and at the lightest, it still blocks a bunch of light. The difference between the darkest and the lightest is what defines the LCD's rather wimpy contrast ratio. And that why the blackest of the black part of your movies look pretty darn gray.
If you take off the front and back polarizing films from your LCD panel and look through it, all of the light will come through. All you will see is the "screen door."
neededandwanted
May I ask your proffession? You've served the members some prety "hard won" information
Salute!
Good information!
zardoz..."goes to bed a little less stupid tonight" I try to every day...
May I ask your proffession? You've served the members some prety "hard won" information
Salute! Good information!
zardoz..."goes to bed a little less stupid tonight" I try to every day...
Lens for sale
Here's an opportunity to buy a lens that's perfect for LCD projection. 18" focal length, very efficient / fast.
This lens was removed from a device meant to project illuminated opaque pages, so it really makes a bright image. I believe it's in great optical condition.
I'm offering it here, so someone who really wants it wont have to compete on ebay - they can just buy it from me.
I want $75 for it. Probably the best single investment you can make for your DIY projector project using a backlit 14-15" LCD panel!
First to respond to joejas@ihpc.net get it.
Here's an opportunity to buy a lens that's perfect for LCD projection. 18" focal length, very efficient / fast.
This lens was removed from a device meant to project illuminated opaque pages, so it really makes a bright image. I believe it's in great optical condition.
I'm offering it here, so someone who really wants it wont have to compete on ebay - they can just buy it from me.
I want $75 for it. Probably the best single investment you can make for your DIY projector project using a backlit 14-15" LCD panel!
First to respond to joejas@ihpc.net get it.
Attachments
fresnel
Verbose:
It's the original Elmo- I separated the 2 halves. It's a pain in the a**., but well worth the effort.
Verbose:
It's the original Elmo- I separated the 2 halves. It's a pain in the a**., but well worth the effort.
well slap me around and call me susie,........... when you seperated the fresnels how did you orient(SP) them, smooth side toward the lcd panel top and bottom etc?
Re: fresnel after LCD
Proto: You have just GOTTA post this pic over on the "Double Fresnel vs. Two single ones" thread on this board. They need to see this.
Hopefully, someone can do some before and after pics...?
VFreak: take a look at that thread as well. MUCH discussion on this. Basically, the orientation of the fresnels should remain unchanged after separation. In about 100.5% of the cases, it will be the opposite of how you asked--that is the smooth sides get the non-parallel light, and the side with the ridges will face that LCD panel (both of them).
proto5 said:
Proto: You have just GOTTA post this pic over on the "Double Fresnel vs. Two single ones" thread on this board. They need to see this.
Hopefully, someone can do some before and after pics...?
Video Freak said:
VFreak: take a look at that thread as well. MUCH discussion on this. Basically, the orientation of the fresnels should remain unchanged after separation. In about 100.5% of the cases, it will be the opposite of how you asked--that is the smooth sides get the non-parallel light, and the side with the ridges will face that LCD panel (both of them).