...ok fair enough. If that was the cheapest, I imagine there won't be cheaper out there. I just wonder about the geometry that we are experiencing with these lens. Do we need more precisely manufactured lens before we start coating them? I will check out the link again and see if it looks do-able. I have some spare time now so I could look for something in Toronto if possible. You never know.
stevodude said:
I haven't seen anything like that with my prisms. I have a computer connected and can put up all sorts of different test images. And while I still have slight pincushioning issues in the corners, I don't see any issue with changes to the aspect ratio as you move across the screen.
I put up a grid pattern on the computer and then measured the grid "squares" and it seems pretty uniform across the screen. All I can think of is that you don't have the prisms set up correctly. I know that it took me a couple of attempts to get them into the correct position.
Many of the recent posts illustrate how different this is on different systems. I'm guessing that the people having the most trouble might have the shortest throw distance, or maybe just got bad prisms. But it would be useful to post more details on your system setup to see if there is something in common. Posting your projector model, screen size, throw distance, etc, might help us determine what might be causing problems.
When I bought 4 prisms, 2 of them provided slightly blurry spots when used together, so I don't think the surfaces are as optically perfect as we might like. So it is certainly possible to get "bad" prisms.
I wish there was some way to easily post more exact descriptions or pictures so that we could see the setups that are not working and provide suggestions. There are so many variables involved. For example, I have found that the color "convergence" issue at the edges of the screen become worse as I move the prisms further away from each other.
I definitely encourage people to play with only 2 prisms first to get the hang of it. The 4-prism system has a lot more variables, and, of course, you need to have 4 matched prisms instead of just 2. And the reflections are always going to be worse with the 4 prism setup.
So try it first with 2 and see if you can play with it to reduce the color issues.
I've got to say that when I first read this thread, I was really concerned about the CA problems that people were reporting with the 2-prism lens, but honestly I just don't see any color problem with my setup. Either I'm blind or I just got lucky with 2 good prisms. The only issue I wasn't able to solve is some pincushioning in the corners, even with tilting the enclosure. But it's pretty minor and with the masking on my screen, it's not noticeable at all during movies.
Finally, we shouldn't overstate the reflection issue. With a 2-prism system the single reflection is very minor and only occurs in very specific scenes that are very dark with a bright spot on the right or left of the image (depending upon your prism orientation). Keep in mind that the reflection is something like 0.1% of the intensity of the normal image. I've watched many movies over the past week and have only seen a dim reflection a couple of times, and I know exactly what to look for (my wife hasn't noticed them at all).
Sorry for the long post, but I just wanted to make sure we were all on track here and didn't scare away anyone new who might want to try this. Yes, there are minor issues...but are these issues worth >$1000 for a "real" anamorphic lens? Not for me. I'm *extremely* happy with the image that I'm getting with a 2-prism system.
Projector: Infocus 7205. Screen: 53" tall. Throw distance: 16 ft.
Projector Panasonic ae700
throw distance 3200mm
Screen size 2500mm width, height 1060mm
2 crystal glass lenses 170mmx140mm, 13deg lense
in order from projector to screen
straight through
lense 1 90deg corner towards projector parallel to projector lense
lense 2 90deg corner towards projector, with 13deg away from parallel to lense thin end out - squares are uniform squares over whole screen
Horizontal stretch
lense 1 90deg corner still towards projector with center rotated 19deg and fat end away from projector
lense 2 90deg corner still towards projector with center rotated 17deg and fat end away from projector
I tried it with 90deg lense 1 towards projector and 90deg lense 2 towards screen, but couldn't get a horizontal stretch on left hand side.
maybe I have the projector too close to the screen for this to work, I can move the whole system from 3meters to 9 meters off the screen...
throw distance 3200mm
Screen size 2500mm width, height 1060mm
2 crystal glass lenses 170mmx140mm, 13deg lense
in order from projector to screen
straight through
lense 1 90deg corner towards projector parallel to projector lense
lense 2 90deg corner towards projector, with 13deg away from parallel to lense thin end out - squares are uniform squares over whole screen
Horizontal stretch
lense 1 90deg corner still towards projector with center rotated 19deg and fat end away from projector
lense 2 90deg corner still towards projector with center rotated 17deg and fat end away from projector
I tried it with 90deg lense 1 towards projector and 90deg lense 2 towards screen, but couldn't get a horizontal stretch on left hand side.
maybe I have the projector too close to the screen for this to work, I can move the whole system from 3meters to 9 meters off the screen...
Here is a picture of my lens setup:
Sounds like it's similar to what you described, although the exact angles might be different.
To decrease color problems, I found that having the second prism as close as possible to the first (so that the prisms are actually touching) helps. Also, the horizontal position of the second prism is important. Use a white piece of paper in front of each prism surface to ensure that the incoming light beam is roughly centered in the prism.
With the projector pointing at the prisms as shown in the image, I adjust prism A to move the right edge of the screen image, then adjust prism B to move the left edge of the screen image.
I wouldn't think that the throw distance would effect the geometry uniformity. The throw distance probably has more of an effect on the color aberations. I honestly can't think of how the geometry would be effected as you described.
But you should place your projector as far back from the screen as you can, then move it forwards just a bit so that your projector is not at the maximum zoom. That's the best compromise between the longest throw (will give the best geometry) without getting into the non-linear portion of your projector's zoom.
Can you measure one of your "squares" at the middle vs the edge of the screen to see how much of a difference you are talking about?
Sounds like it's similar to what you described, although the exact angles might be different.
To decrease color problems, I found that having the second prism as close as possible to the first (so that the prisms are actually touching) helps. Also, the horizontal position of the second prism is important. Use a white piece of paper in front of each prism surface to ensure that the incoming light beam is roughly centered in the prism.
With the projector pointing at the prisms as shown in the image, I adjust prism A to move the right edge of the screen image, then adjust prism B to move the left edge of the screen image.
I wouldn't think that the throw distance would effect the geometry uniformity. The throw distance probably has more of an effect on the color aberations. I honestly can't think of how the geometry would be effected as you described.
But you should place your projector as far back from the screen as you can, then move it forwards just a bit so that your projector is not at the maximum zoom. That's the best compromise between the longest throw (will give the best geometry) without getting into the non-linear portion of your projector's zoom.
Can you measure one of your "squares" at the middle vs the edge of the screen to see how much of a difference you are talking about?
thinking off hand - as I have put projector back in roof mount ( I had it on a desk testing it). I am currently making a aluminium frame around each lense so I can rotate them with a bolt underneath.
the correct geometry square (measured at 1/3 & 2/3) horizontal postion is 110mm square, center squares are 110mm high & 103mm wide, at the outsides L & R the squares are 110mm high & 120mm wide, the overall dimensions are correct at 2.35:1 ratio, but internal geometry is as above.
geometry as below roughly in mspaint:
the correct geometry square (measured at 1/3 & 2/3) horizontal postion is 110mm square, center squares are 110mm high & 103mm wide, at the outsides L & R the squares are 110mm high & 120mm wide, the overall dimensions are correct at 2.35:1 ratio, but internal geometry is as above.
geometry as below roughly in mspaint:
An externally hosted image should be here but it was not working when we last tested it.
Even though I have the thicker prisms that Steveodude has, right now I have them aligned the same as MikeP where the apex of the P2 overlaps the base of P1 and physically touches and the right angles are in.
Last night I watched Batman Begins, and there was NO stray light reflections and very uniform stretch...
Mark
Last night I watched Batman Begins, and there was NO stray light reflections and very uniform stretch...
Mark
and where is your rotation point on each prism.
I basically set the center of the prism ( centre point measured from each side with 90deg line then where those 90deg cross ) to make the center rotation point?
or would it be better to have the rotation point closer to the skinny end to get more angle for the large end to make the Horizontal expansion?
I basically set the center of the prism ( centre point measured from each side with 90deg line then where those 90deg cross ) to make the center rotation point?
or would it be better to have the rotation point closer to the skinny end to get more angle for the large end to make the Horizontal expansion?
stevodude said:
I have mine in case so I then move the case left or right depending on what gives the best results. Right now it is off centre to the left so with the primary lens (projector's lens) at the thin end on P2, but the thick end of P1...
I also realigned the prisms with the case hard against the projector. The image is slightly offshooting the right side of the screen to get the full stretch, but that could be brought back in by a slight tweak of P1 (rear prism).
I seems that I can remove the and re-install the lens without any issues for alignment the way I have them configured now...
HERE is the development page of my BLOG...
Mark
Well Stevedude, I take back what I said. You are right.
I had thought you were talking about something more dramatic. But when I saw your actual "square" dimensions, I decided to measure mine more precisely.
The computer grid image that I was using was using much smaller squares (about 15mm by 15mm on the screen). So it was hard to measure much difference here. But when I saw your measurements, I decided to use 7 of my grids to get a size closer to your 100mm.
When I actually got my nose an inch from the screen to measure the grid, I was surprised to find the same result that you had.
In 16:9 mode, my "squares" are 104x104mm in size. When I put the prisms in front, the "squares" in the center are now 104x132mm whereas the "sqaures" on the edge of the screen are 104x148mm. The correct aspect ratio would give 104x137mm which happens in the middle-left and middle-right just as you mentioned.
So yes, the stretching is not perfectly uniform. But you know, if I hadn't actually measured it, I wouldn't have believed it and I still never notice it when watching movies or even when using the computer with the prisms in front of it.
I'd be interested in seeing measurements of commercial lenses to see if they do any better. The above numbers only represent a few percent change in aspect ratio across the screen, so I don't see this as a real problem.
A bigger problem is that when I got this close to the screen, I *did* notice more color artifacts then I thought I had seen in my initial tests. In my initial tests I had a grid of white lines on a black background, and that looked fine.
But this new test has a grid of black lines on a white background, and the black lines towards the right and left sides of the screen have definitely color bands. I'll try to take some closeup pictures with my new camera this weekend (if it arrives through the snow storm we are having in Colorado).
Again, I have not noticed these color effects when watching a movie, and I can only see the color bands when I'm closer than about a foot from the screen. The screen-door effect from my DLP is much more noticeable than the color bands.
So yes, technically they are there. But again, I don't notice any color problems at my normal 12 ft seating distance.
I had thought you were talking about something more dramatic. But when I saw your actual "square" dimensions, I decided to measure mine more precisely.
The computer grid image that I was using was using much smaller squares (about 15mm by 15mm on the screen). So it was hard to measure much difference here. But when I saw your measurements, I decided to use 7 of my grids to get a size closer to your 100mm.
When I actually got my nose an inch from the screen to measure the grid, I was surprised to find the same result that you had.
In 16:9 mode, my "squares" are 104x104mm in size. When I put the prisms in front, the "squares" in the center are now 104x132mm whereas the "sqaures" on the edge of the screen are 104x148mm. The correct aspect ratio would give 104x137mm which happens in the middle-left and middle-right just as you mentioned.
So yes, the stretching is not perfectly uniform. But you know, if I hadn't actually measured it, I wouldn't have believed it and I still never notice it when watching movies or even when using the computer with the prisms in front of it.
I'd be interested in seeing measurements of commercial lenses to see if they do any better. The above numbers only represent a few percent change in aspect ratio across the screen, so I don't see this as a real problem.
A bigger problem is that when I got this close to the screen, I *did* notice more color artifacts then I thought I had seen in my initial tests. In my initial tests I had a grid of white lines on a black background, and that looked fine.
But this new test has a grid of black lines on a white background, and the black lines towards the right and left sides of the screen have definitely color bands. I'll try to take some closeup pictures with my new camera this weekend (if it arrives through the snow storm we are having in Colorado).
Again, I have not noticed these color effects when watching a movie, and I can only see the color bands when I'm closer than about a foot from the screen. The screen-door effect from my DLP is much more noticeable than the color bands.
So yes, technically they are there. But again, I don't notice any color problems at my normal 12 ft seating distance.
OK...totally off topic!
First...Happy New Year to everyone.
Now, on with the show...has anyone using a painted screen, used reflective paint? And if so, any problems(hotspotting, etc)? For Xmas, I got a BAG of fine glass beads o mix with my paint to make a reflective screen. Checked the Internet & highway paint is 6 lbs of beads per gallon of paint, so....thought i'd run this by people here & see if anyone had experience. As for my prisms...I currently have 2 setups, a 2 prism & a 4 prism. Dispite the reflections, I think I still prefer the 4 prism set, it just seems, to me, to have a better more vivid picture. But that's just my opinion.
Looking forward to progress on the anti-reflective persuit.
Bud
First...Happy New Year to everyone.
Now, on with the show...has anyone using a painted screen, used reflective paint? And if so, any problems(hotspotting, etc)? For Xmas, I got a BAG of fine glass beads o mix with my paint to make a reflective screen. Checked the Internet & highway paint is 6 lbs of beads per gallon of paint, so....thought i'd run this by people here & see if anyone had experience. As for my prisms...I currently have 2 setups, a 2 prism & a 4 prism. Dispite the reflections, I think I still prefer the 4 prism set, it just seems, to me, to have a better more vivid picture. But that's just my opinion.
Looking forward to progress on the anti-reflective persuit.
Bud
OK thats cool, so it is basically there no mater which way around the lenses are you will probably get some variation in horzontal compression across the screen, unless you use perfectly curved lenses...
just building a pot-rivet alluminium L channel frame around my lenses so I can mount a bolt on the bottom center so I can adjust through or stretch mode.
My only issue here is the verticle high of the lense of each ends (fat end vs thin end) is about 10mm difference, so making a 5mm lift at the thin end hopefully will center them better and parallel the glass to the projector lense.
just building a pot-rivet alluminium L channel frame around my lenses so I can mount a bolt on the bottom center so I can adjust through or stretch mode.
My only issue here is the verticle high of the lense of each ends (fat end vs thin end) is about 10mm difference, so making a 5mm lift at the thin end hopefully will center them better and parallel the glass to the projector lense.
Happy new year everyone 🙂
Non linear stretch occurs with pro lenses as well as our DIY lenses...
Mark
Non linear stretch occurs with pro lenses as well as our DIY lenses...
Mark
yea happy new year to all.
coupla pics of my down and dirty lense frame (one of them anyway)
on a piece of 50x18mm pine length just for getting the distance correct between the two lenses, then I'll mount in a box & probably cover the edges in black electricians tape.
I'll also be replacing the zippy ties with bolts when the shops open tomorrow, and also get longer bolts for the rotation action so I can put a permanent dial/adjuster on the end, I bought some that were too small, so only just fit a nut on the end.
so far it is rock solid and stays in place when I move it.
can't wait to finish the other fram and throw them up for a test run.
I also made them square, only because that was easier, cause the three overlapping each other would create a twist in the angle, so this way it's perfectly square and will leave the prisms perfectly horizontal.
more to come 🙂
coupla pics of my down and dirty lense frame (one of them anyway)
on a piece of 50x18mm pine length just for getting the distance correct between the two lenses, then I'll mount in a box & probably cover the edges in black electricians tape.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I'll also be replacing the zippy ties with bolts when the shops open tomorrow, and also get longer bolts for the rotation action so I can put a permanent dial/adjuster on the end, I bought some that were too small, so only just fit a nut on the end.
so far it is rock solid and stays in place when I move it.
can't wait to finish the other fram and throw them up for a test run.
I also made them square, only because that was easier, cause the three overlapping each other would create a twist in the angle, so this way it's perfectly square and will leave the prisms perfectly horizontal.
more to come 🙂
stevodude said:
How would you get a twist? Perhaps I'm simply not seeing fully how you assembled this. Are you notching at the corners? As I think about your statement, you must not be, and notching would solve this issue. Single piece of L notched for the 4 corners. To make it a solid loop, leave an extra half inch or so of the vertical face (let the surface the prism sits on sit flush) and rivet there. Should be able to do the primary piece with just the one rivet, leaving only the 2 you used to secure the piece the adjustment bolt goes through. Not critical for function unless you have troubles getting the narrow edges close enough, which may be an issue.
Regardless, nice work and *great* idea. I work with this stuff all the time and know exactly how I'd work this. It fixes the last remaining discomfort I had (I tend to get started and not follow through if something is too miserable) as far as being able to properly adjust - just didn't like the concept of trying to glue bolts directly to the glass.
So now I need to order some prisms and fortify my psyche for more rolled eyes. 😎 NE IL is gonna have a lot of these suckers.
Oh... yeah, I'll need a new screen too.
C
I am one who has adhered the bolts directly to the glass, and I am almost to the point where I can get the whole contraption up and mounted with my ceiling mounted projector. But, I am holding back because adhering directly to glass has made me a bit nervous that heavy glass prisms are going to start raining down on people watching a movie.
I followed the idea of whoever mentioned putting the bolt between the prism and a piece of cardboard, except that I used plexiglass instead of cardboard.
I cut triangular pieces of 0.22" acrylic (from Home Depot) to match the top and bottom surfaces of the prism. I drilled and countersunk holes in the plexiglass. You feed the bolt through the plexiglass, then epoxy the entire piece of plexiglass (and screw head) to the prism.
Because of the countersink, the screw head is trapped between the plexiglass and the prism. And with a good epoxy, it's not going anywhere.
I did have trouble with snapping the plexiglass off from the prism (it came off very cleanly). But this was from the same surface that I had tried using a "universal adhesive" and then been forced to remove with GoofOff, and I think some of that GoofOff hadn't been fully cleaned off, leading to a bad epoxy joint. I just used a fresh epoxy to glue it back and it's been rock solid since then.
But the increased surface area of the plexiglass/prism bond really helps. You mainly just need to use enough epoxy around the screw head so that it doesn't break and start turning by itself.
Also, I found that it was *very* important to match the shape of the plexiglass with the triangular shape of the prism top/bottom. When I first tried just using a rectangular piece of plexiglass, then I couldn't get the two prisms close enough together. Since I have small prisms, I don't have much margin...the farther apart the prisms are, the worse the color problems on the edges are.
My camera still hasn't arrived due to the combination of Colorado blizzard and then the New Years holiday. But it's not just the screws that hold the prisms. The screws help provide the adjustment of the prism angle. But when the enclosure top is screwed down, the prisms are also held between the top and bottom of the enclosure with felt inside against the prism/plexiglass top/bottom. There is absolutely no way that these prisms could fall out of the enclosure that I have made.
Still hope to have a full set of pictures in a few days, along with an entire web page describing the construction process.
I cut triangular pieces of 0.22" acrylic (from Home Depot) to match the top and bottom surfaces of the prism. I drilled and countersunk holes in the plexiglass. You feed the bolt through the plexiglass, then epoxy the entire piece of plexiglass (and screw head) to the prism.
Because of the countersink, the screw head is trapped between the plexiglass and the prism. And with a good epoxy, it's not going anywhere.
I did have trouble with snapping the plexiglass off from the prism (it came off very cleanly). But this was from the same surface that I had tried using a "universal adhesive" and then been forced to remove with GoofOff, and I think some of that GoofOff hadn't been fully cleaned off, leading to a bad epoxy joint. I just used a fresh epoxy to glue it back and it's been rock solid since then.
But the increased surface area of the plexiglass/prism bond really helps. You mainly just need to use enough epoxy around the screw head so that it doesn't break and start turning by itself.
Also, I found that it was *very* important to match the shape of the plexiglass with the triangular shape of the prism top/bottom. When I first tried just using a rectangular piece of plexiglass, then I couldn't get the two prisms close enough together. Since I have small prisms, I don't have much margin...the farther apart the prisms are, the worse the color problems on the edges are.
My camera still hasn't arrived due to the combination of Colorado blizzard and then the New Years holiday. But it's not just the screws that hold the prisms. The screws help provide the adjustment of the prism angle. But when the enclosure top is screwed down, the prisms are also held between the top and bottom of the enclosure with felt inside against the prism/plexiglass top/bottom. There is absolutely no way that these prisms could fall out of the enclosure that I have made.
Still hope to have a full set of pictures in a few days, along with an entire web page describing the construction process.
Yes, if your gonna do a rotating prism design, then you need rotating pins on both the top and the bottom of each prisms I have not done this as doing so would make the case larger and it is already 200mm x 200mm x 175mm...
Mark
Mark
MikeP: Great idea, and thanks--I will be doing that. I already have the bolts adhered to the prism tops (and I have a flat-headed rivet on the bottom to provide the point of rotation), so all I need to do is add the plexiglass. And I have a few scraps of plexiglass left over in my scrap pile downstairs, so I could do this tonight (if my wife didn't keep forcing me to watch Battlestar Galactica <--not a joke! Well, the "forcing" part is a joke, since I am as willing as she is--I just can't believe my non-scifi wife is so obsessed over this show--we are one show away from finally being caught up; we've been tivo'ing the episodes all year...)