New theories for keystone correction, merged topic Options

[Rox]

Well charlie, brain said that the standard triplet is capable of getting about a seventeen inch lcd into focus when the edges will begin to dim.

well, i can focus a 29" flat crt tv with this standar lens as well, it only depends on the trhow. if you project to 2 feet, you will have everithing focused.

I mean that 17" focused is it posible but you need to say what throw was used when tested, this information is very significant.

Rox

[Mikau]

Well when I asked brain just said 17 inches, I assume that means at the standard distance for the fresnel.

See that would work if only the throw distance was larger and the projection bigger. But putting it that far away, like you said, will result in a tiny projection at a very short distance. If we had a lens with the same focus area but a longer throw and higher magnification, we'd just need a larger fresnel and use half of it.

[Rox]

Well when I asked brain just said 17 inches, I assume that means at the standard distance for the fresnel.

See that would work if only the throw distance was larger and the projection bigger. But putting it that far away, like you said, will result in a tiny projection at a very short distance. If we had a lens with the same focus area but a longer throw and higher magnification, we'd just need a larger fresnel and use half of it.

the problem with large panels is not the fresnells at all. The main problem is the projection lens. If you get a projection lens with focal and field angle enough then i would say the standar fresnells will do a good job for you. But if you have a lens with a extremelly small field angle, you will have a hard time doing it work.

The fresnell will work on a wide range just moving the lamp placemnt.

[Mikau]

I'm talking about for off setting. Lets say you had a 30 inch fresnel and had a triplet that could get a 30 inch lcd into focus and project at an adequete size at an adequete range. Then You could simplay place a 15 inch lcd on the that fresnel on the uper or lower half. This would result in a 50% shift.

[Rox]

i agree. But you would not find such a projection lens.

[Mikau]

True.

[charlie10]

Hiya Mikau!

Encouraged by this thread and the testing I've done with my halogen lamp (6mm filament) and the standard LL fresnels, I've ordered a pair of large (15"x15") fresnels (from diyprojectioncompany.com). My plan is to place a 9"x12" LCD in one corner of the large fresnel, hopefully achieving about 20 degrees of vertical and 5 degrees of horizontal shift. Will either of my triplets (Fujinon copy lens, Standard LL triplet) be able to focus this shifted image?

So I went ahead and placed the order last week. As of last update it is "Processing". It's not clear to me whether they have the lenses in stock or if there will be a longer delay. The suspense has been killin me!

If the image won't focus, the next step is to break apart the triplet. Would I use a dremel for this?

I'm pretty confident that by hook or crook we can get good results out of this, without needing a pro lens. My latest hope is to incorporate Minoten's panel, and make a nice, small, sharp short-throw coffee-table type setup. I will try to post results for longer-throw setups too.

[Mikau]

Hiya Mikau! 

Encouraged by this thread and the testing I've done with my halogen lamp (6mm filament) and the standard LL fresnels, I've ordered a pair of large (15"x15") fresnels (from diyprojectioncompany.com).  My plan is to place a 9"x12" LCD in one corner of the large fresnel, hopefully achieving about 20 degrees of vertical and 5 degrees of horizontal shift.  Will either of my triplets (Fujinon copy lens, Standard LL triplet) be able to focus this shifted image?

So I went ahead and placed the order last week.  As of last update it is "Processing".  It's not clear to me whether they have the lenses in stock or if there will be a longer delay.  The suspense has been killin me!

If the image won't focus, the next step is to break apart the triplet.  Would I use a dremel for this? 

I'm pretty confident that by hook or crook we can get good results out of this, without needing a pro lens.  My latest hope is to incorporate Minoten's panel, and make a nice, small, sharp short-throw coffee-table type setup.  I will try to post results for longer-throw setups too.

Well 20% isn't much really, though better then nothing and 20% means a larger shift the bigger the projection.

I think it would be better if you got the pro lens kit. The pro field fresnel will cover up to a 21 inch lcd. Thats six inches more then 15 inches. So you should be able to get at least a 25% shift.

However what I intend to do for my projector is, with the help of an N6, set it to widescreen letterbox display at all times, then I can shift down the fresnel 25%. Then what I believe I can do with the N6 is set full screen to display in letterbox with black bars on the left and right. While resolution is lost, this is still enough for 480p which is about as high as it gets for full screen formats. (Unless you play PC games, which I don't). Like I said, this will give about a 25% shift which will be about 16-18 inches up for a 9-10 foot projection which is what I plan to do. The fullscreen format will be about 1/3 the size this way, but I think I won't mind considering I'll have the projector out of my way.

Later I'll upgrade to a 17 inch panel, with the pro lens kit. The larger then needed fresnel will allow me to off set the panel about 25%. Furthermore I found that most 17 inch panels are actually a 5:4 ratio, (1280x1024) which means its even taller then normal so I can move the fresnel down even farther. So in the end I'll have about a 50% shift which means the projector centered on the top of the screen (which is all you need in my oppinion).

What I wish I could do is make 480p display in 480-640 pixels exactly regardless of the available pixels on my lcd. Then with a 17 inch I could have even higher off setting in fullscreen mode and there would be no distortion from scaling.

[Mikau]

Ok, I'm going to continue more talk about what needs to be done ignoring how we could do it.

The reason I don't think we could use a mirror or bend a fresnel is we need to 1, shrink the image on one side and stretch on another, and we also need to change the distance of the screen in differant spots. So in otherwords we need sort of a bent trapazoid. (Though not quite a trapazoid, more like a trapazoid with curved sides)

Though while analysis tells me that bending the fresnel won't work, I can't help but think it just might work. Someone has to try it.

Meanwhile I'm trying to think up cheap means of developing specialized equipment for keystone correction.

The first is a fresnel that is nothing but a bunch of the same triangles. If the projected beam were a square, this would work by placing this fresnel in front of the triplet. Like the picture below shows.

However tilting a triangle is not so easy. Its not exactly a triangle, more like a trapazoid. I need to test it on cad to see if passing through this type of fresnel will not produce a curve.

But even if it did work, redirecting the projector beam mid way through might really mess it up. Might not though.

Attached File(s)

 congfresnel.JPG ( 16.09K ) Number of downloads: 26

 

[Mikau]

The other idea is no one here has tried any keystoning experiments with glass wedges. Someone from another site built an annamorphic lens using two prisms, out of pieces of glass, one filled with water, one with oil.

I think if you can build an annamorphic lens you might be able to make a keystone correction device.

[DAZZZLA]

The other idea is no one here has tried any keystoning experiments with glass wedges. Someone from another site built an annamorphic lens using two prisms, out of pieces of glass, one filled with water, one with oil.

This is a simular thing that I was thinking about. I haven’t had time to test the idea yet. Seems to look like it would work.

DJ

[Mikau]

Yes I hope so. But optics is quite a delicate buisness.

[artificer]

The other idea is no one here has tried any keystoning experiments with glass wedges. Someone from another site built an annamorphic lens using two prisms, out of pieces of glass, one filled with water, one with oil.

I think if you can build an annamorphic lens you might be able to make a keystone correction device.

I made some 15deg prism wedges with lexan and water. A VERY quick atempt a keystone corection left me with some barreling. It wasn't a good test, since it was very quick. I probably was doing something wrong. You are supposed to be able to use 2 wedges to convert an eliptical light source into a circular one.

I need to get back to the experiment, but I did prove its fairly easy to make a test lens system with only some plastic, hot-melt glue gun, and water.

I've been toying with the idea of a varing focus lens. It would be a square with infinite focal length at the top (flat plate) and evenly progress to the desired focal length on the bottom. It would only be in one direction, basicly a cylindrical lens. My method of construction would be either two sheets of lexan with water filling them. One is flat, the other is curved. This would only be a test lens. A rigid lens would be made with a sheet of acrylic curve to shape, and filled with water clear acrylic casting liquid. Neither lens is perfect, but cheap, and easy to make to test out ideas.

Too many ideas... so little time. :-)

Do you have a link to where you found the person making the annamorphic lenses?

[Mikau]

Ask joe chevvy, he mentioned it on the inteligent video projector design forum a few weeks backs.

While I think it could possibly be correctable if you have acrilic wedges that you can bend, but bending it both ways would probably be impossible.

Unless perhaps you bent one side one way, and the other another. See picture below.

This post has been edited by Mikau: Apr 24 2005, 12:52 AM

Attached File(s)

 glasswedge.JPG ( 7.02K ) Number of downloads: 6

 

[Mikau]

But you should really try using two normal wedges and goofing around with them.

[Mikau]

I really think that the only way we can achieve good keystone corrrection, optically, is if professional equipment is professionally developed specifically for keystone correction/lens shitfing.

I think where we need to go in the future is hope companies develop cheaper smaller lcds with higher specs. So we can offset the lcd on the fresnel. At this point we can still use a 15 or 17 inch with a prolens kit to do this. Also we can use letterbox only if we wish to improve things.

[maddmike]

I really think that the only way we can achieve good keystone corrrection, optically, is if professional equipment is professionally developed specifically for keystone correction/lens shitfing.

I think where we need to go in the future is hope companies develop cheaper smaller lcds with higher specs. So we can offset the lcd on the fresnel. At this point we can still use a 15 or 17 inch with a prolens kit to do this. Also we can use letterbox only if we wish to improve things.

I think that we can do this with current fresnels from diylabs. the 15.5"x15.5" lens should give some shift on a 15" screen.

I also think that keystone is less of an issue with a longer throw. If we could think how to achieve a cheap way for a longer throw that would be a positive effort.

http://www.diyprojectorcompany.com/

[Squalish]

W/ longer throw we NEED large + expensive lenses, unfortunately, if we're going to use the same lamp.

If we switched to a smaller arc lamp (example OSRAM's 575W HSD lamps, w/ a 6000hr lifetime, 7mm arc, though I know not the cost), we could use a triplet or achro doublet and not have to worry as much about field angle or focus.

If you ignore the overall arrangement of the pixels for a second, the triplet's job, for each pixel, is to focus the 1" arc into a point in the general vicinity of the screen. With a smaller arc, this is much easier.

With long throws, divergence due to the arc size gets more and more powerful.

This post has been edited by Squalish: Apr 28 2005, 05:55 AM

[charlie10]

With long throws, divergence due to the arc size gets more and more powerful.

Is it true that using a 330mm rear fresnel, instead of a 220mm, will make the bulb arc image smaller?

If so then would such a setup improve focus (I guess at the expense of brightness / light gathered) ?

Apologies if this question has already been answered.

[Rox]

yes, the rate (fiel fresnell/rear fresnell) would give you the magnification of th arc lengh on the projection lens. If you have 330/330 then the arc would be the same size on the triplet (Care this is true only if you are not forcing the field focusing point placing the lamp closer to the rear fresnell, this would have more rate)

but there is other problem using longer rear fresnells, that the light taken from the bulb would be partially wasted because the solid angle will be less on a 330 focal than a 20 focal (the light cone is smaller on 330 than 220, so worst of light)

[charlie10]

It would be so great to not waste all that light from the bulb. A pre-condenser lens would help... but would the bulb arc image then become larger again?

[Rox]

thats it. We have then the same problem but on a larger box

[Mikau]

Just been thinking.

A normal convex lens will work as a projection lens but produces barreling. Triplets are a special set up of lens designed to correct this aberation. If a round focus plain can be converted to a flat one, why can't a normal focus plane be converted to a tilted focus plain? If glasses can be used to correct imperfections in our eyes, why can't we make a lens to optically correct for keystoning?

I've learned from projector building that optics are an extremly sensitive field and nothing works unless it is crafted with predesigned precalculated precision. Manually readjusting lenses in an attempt to modify its function will result in distortion, in 99.9999% of the cases. While making a keystoning triplet would be easier then offset projection I think, as it wouldn't require a lens with a very large area of focus, it may or may not be possible to design a keystoning triplet that works for every projection size. Then again it very well might, I don't see why the focus plain can't be converted to a tilted one, but what do I know about optics?

At any rate I think if a triplet was designed to only produce keystone correction to a fixed degree (like 40 degrees ) it would work for any projection size, but again thats just speculation.

But if it is indeed possible I don't see why anyone hasn't developed one for projectors of any form. Who does prefer their projector blocking their view? So maybe it is an impossibility. Or maybe no ones really bothered with it. Thanks to things like digital keystone correction.

I hate this. All I can do is speculate. :angry:

This post has been edited by Mikau: Apr 28 2005, 10:36 PM

[artificer]

But if it is indeed possible I don't see why anyone hasn't developed one for projectors of any form. Who does prefer their projector blocking their view? So maybe it is an impossibility. Or maybe no ones really bothered with it. Thanks to things like digital keystone correction.

All of the recent commercial projectors I use at work have what you describe. The projector sits on a table, and projects perpendicular to the screen. The bottom of the projected image is around the level of the projector, and the rest goes up (or down if inverted for ceiling mounted) Everything is in focus. Must be custom optics.

I think this is what you were getting at... now if I/we can only come up with the solution.

[Mikau]

Perpendicular to the screen would mean lens shift, not keystone correction.

Thats another possibility, once again I think it would be cheaper if it was shifted only to a specific distance, this would probably require a front lens being abnormally large.

Its definitly possible, but someone has to convince an optics company to develop one first, and even then it might be a costly sophisticated triplet.