Nvidia drivers allow you to rotate the image, reverse the image, all sorts of stuff. I bet it wouldn't be hard (for Nvidia) to write a keystoning driver.

I don't have experience with any other (current) manufacturer, but I am sure they could also do a driver keystone routine.

Perhaps a coordinated mass emailing would get enough attention to actually get keystoning incorporated into the next driver release?

Of course, if you are using a Mac, you're screwed from the start... heh heh. Just kidding big guy.

This is not an issue of computer software! I want a reproduceable solution that is independent of a computer. There are component-to-VGA converters out there.

Thinking about it real quick guys.... I don't think this occurred to anyone yet, even brainchild. Keystone adjustment wouldn't work if done at the software anyway. The other issue besides the shape of the picture is the projected image being in focus. The distance between the projector and the top and bottom of the screen is different. Having an angled second fresnel is the only way to take care of this. Sotware doesn't and can't adjust for that. Come to think of it.... Maybe the reason I can't get my image totally focused with keystone adjustment...??? Well the angle of the second fresnel turned out ot be a bit different than that of the mounted projector, so that may explain the focus...maybe my lower drage fresnels...
You guys feeling me on this?

I have to disagree. Brainchild says somewhere that if you can find a monitor with trapizopid adjustment than this can be used to adjust keystoning. The only problem is that LCD's don't seem to come with that adjustment.

Now if software can reduce the width of either the top or bottom than that is going to provide the same effect. Not sure... but the software may be required to add black to the edges of the parts which are shrunk to stop the LCD snapping it back to full width....

It seems that software to do this isn't availible... (I had thought of trying to find a uni student who could do an nVidia driver hack...) but that dosen't change the point that keystoning could be corrected at a software/driver level.

dougy, jo had a valid point. If you are hanging a projector off the ceiling and pointing down at a 45 degree angle (an extremity but just say), and you use a trapezoid software function to accomplish the keystoning, then, while the ANGLE is corrected, there will be a distance / throw matter to be resolved, because the top of the screen is closer than the bottom of the screen. Therefore, the bottom will be out of focus while the top is in focus, or the top will be out of focus while the bottom is in focus, or the top and bottom will be out of focus while the middle is in focus.

The fresnel lens angling resolves the FOCUS issue, which is half the purpose of keystoning. Keystoning resolves FOCUS and ANGLE, not just angle.

yeah i just caught on, sorry.


tonytemplin, the fastening idea is good, but u could also use a weight on the bottom of the assembly so that it would tend to auto set itself when loosened.

We are talking dowel rod in hole here, not a ball-bearing-zero-friction mount. But yes, certainly the weight is a good possibility. I was thinking of having the adjustment slot marked off in degrees by using a 50 cent protractor, then using said tool to measure from the table/ceiling the angle the projector is at, to get a precise alignment.

Or, hang a weighted string from the pivot axis to the adjustment stay, to tell you where to move it to. What do builders call that, a plumb? A plumb-bob?

What we (also) really need is a lens that can allow us to move the projector back farther without the image being 12 feet tall, not necessarily a zoom... Then the angles wouldn't be so bad. Any keystoning past 10-15 degrees has got to hurt the picture quality...

******************

Oh and Xiaoth, I am sorry to be thinking of my unique situation, rather than yours.

quote: This is not an issue of computer software! I want a reproduceable solution that is independent of a computer. There are component-to-VGA converters out there.

You did bring up a good point to my short-sightedness however: the focus, not just the 'squareness' of the picture, is important. Thank you for that.

You are correct about the keystoning getting hurt after 15 degrees. The image just starts to get a little.. well... weird after that. You shouldn't have to go much past that. My box is at about 14 degrees with the rear of it less than an inch from the ceiling projecting a HUGE image from a couple inches below the ceiling to 2 ft off the ground. I think the weighted mechanism you guys are talking about sounds great, but it may be more work than necessary if you are semi-permenantly to permenantly.

Here is a quick schematic with clearance for ceiling mount 15" projector in a standard 8 ft. ceiling - not sure if I did the keystoning corectly, but you get the point that it can really screw up the projection with an off screen center mount and no keystone coreection!

You can use software to correct keystone but you'll need to add minor angular focus correction with the pj lens. Using a field lens is better in one regard because you don't lose resolution.

Dang, Rorshach, I wish my quick diagrams looked that good.


Excellent work!

brain.
i have decided to make my projector a floor model.
i know that i have to place the field fres in front of the lcd to correct for keystoning, and i am going to use the center piviot opticly weighted design.
my annoying questions are,

what is the distance from the lcd to the closest point of the field fres.?
should it be the same distance as the rear fres. to lcd- 10mm?

also will i have to compensate the projection lens and move it out
say the same 10mm?

or will i have to measure from the lcd to the center of the field fres.
and move the projection lens out that amount?

or should the projection lens stay in the same place as in the guide
so it can focus on the lcd?

and should i cut the field lens a little taller if it is to be tilted?

any other ideas in relation to the lens set up i might not have thought of
are always welcome
thanks again

You will not have to change the distance of the projection lens. The projection lens to LCD is the most critical measurement. You should cut your field lens to 9.75" which should give you enough movement for about 16º keystone. The fresnel will have to be slightly farther away from the panel so that when it is tilted it will not come so close to the panel to be in the projection lenses focus. I do advise you to set up your rig in a non permanent way to test everything first.

Hopefully I've answered them all, if not just ask.

Ok if I want to have keystoning adjustment. Do I need to have the fesnail that is in from of the LCD be say 10mm or so foward of the LCD which now places it closer then then 290mm as the guide says. So Now I have to move everything back 10mm. to get the 290 mm. Is this right?

Yes, but don't worry about the guide measurements too much, they will change to reflect the new lenses and everybody's set up will differ. Plan to put your pj lens 300-340mm away from the panel. As always, set everything up before you finalize anything so you can test it thoroughly. You will need to space off the panel with the field lens about 10mm.

So I need the LCD 300mm to 340mm from the PJ lens and a 10mm space between the LCD and the field lens. (AKA the keystone fixing lens) Is that right?

So which way makes the image bigger, move lcd toward PJ lens or back toward lamp?

Yes that's right. As far as bigger goes..the farther away the projector is from the screen the closer to the panel the pj lens moves to focus. Moving the lens toward the panel makes the image bigger but it doesn't matter because the focus is what's important.

I'll need to tackle keystoning when I build my projector as it will be ceiling mounted.

I am totally useless at math therefore I always look for an easier / practical solution. Wouldn't it be possible to mount the lens on a hinge similar to the one used on the lid, then adjust the angle by hand until you get the keystoning out of the projected picture, then fix the lens somehow at that point.

I did think of running a threaded rod through the projector and attaching the lens to a nut on the rod, in such a way that turning the rod by a handle at one end, adjusted the angle of the lens. That may be a little complicated however.

I think the basic idea of a hinged lens would work though. Also this would have the advantage of being adjustable for different projecting situations. I don't know if this may already have been suggested... It seems so simple, it probably has. Or more likey I'm missing some point, and it wouldn't work anyway !!!

What do you think ?????

advance tech,
optically weighted lens...(fresnel)
should work to keep the lens parallel to the projection screen.
using a piviot point from the sides or top to allow the fres. to automatically
adjust to the angle you need using gravity. you may need to be able to move the
lens front to back however to prevent reaching the min distance from lens to lcd(10mm), since this will change with each angle.

Ok, I've confused myself again...
Wait - I answered my own question -- sorry

Thought I would add this graphic to illustrate top hinge vs. center hinge. I thought that the fresnel would move too far upward in the arc in a top hinge design, but that doesn't appear to be the case.

It looks like a top hinge would be perfect for ceiling mount, and the center hinge ideal for a wider range as the fresnel can be closer to the lcd panel when in the vertical position (doesn't swing as far).

This is definitely one of those things that really experimenting with it doesn't compare to diagrams. Find your best placement by watching the results as you experiment. You guys really are not going to want to go much past 10 degrees. It starts to fisheye and get out of focus where the fresnel is farthers from the LCD after 10 degrees. 15 is a no go.

My projector will have to be about 6-7 feet from the wall. Do you think that since it is closer that the usual 10 feet, that I will have problems while trying to keystone ceiling mount. I say this because I'm thinking that since it is closer to the wall that it will have to be tilted down more degrees.

It depends on how far down from the ceiling you are dealing with. I would try to get the screen closest to the mounting surface as possible. You will lose clarity on the bottom corners of the image (ceiling mounted) at around 10 degrees. You should experiment to find what is acceptable for you.

I had only placed the 15 degree swing as I thought that Brain had mentioned that somewhere. But the main point of the diagram was to illustrate any differences between center hinging or top hinging. I was worried that the top hinge would swing the fresnel upward in the arc and pulling it off center from the projection line. This doesn't seem to be the case as the arc is quite large.

The dimensions may be off on the projection size and distance but it gets the general idea across. Looks like it would be beneficial to lower the projector a bit to reduce the angle.

Great diagram and thinking ahead, but it doesn't matter where the real or perceived pivot or hinge is. The only thing that matters is that the angle is correct compared to the LCD and the screen, that it is the correct distance from the LCD, and that it captures the entire LCD image. This is definitely something you are going to want to do hands on experimenting with. I have seen the 15 degree deal thrown around a bit, but it absolutely doesn't work for me.

You got it! Lower it as far as possible without obstructing view to provide for the smallest frez angle. I am shotting for 10 degrees on my final PJ. The box will also be shaped to provide the lowest profile to less obstruct the view.

Once I will build my pj, assuming we will have a working zoom-out feature by then, this pj keystoning issue is one problem I hope will be behind me (pun intended).

Has anyone constructed a weighted optical keystoning system with top hinge?

Also, does the keystone correction blur any corners or worsen the picture when compared to straight on?

I was just going to use small eye bolts suspended from a threaded rod with nuts on the rod to adjust/hold side to side position. I think gravity would do just fine to maintain your fresnel parallel to the wall.

You can get more than 10º by changing the angle of the panel. You would also need to angle the pj lens to match the panel which is relatively easy.

Any thoughts on just tilting the screen 5 to 10 deg to correct keystone.when I say "screen" i DO NOT mean the lcd.
My projector may be about 1 1/2 feet off the floor so I dont think i will have to angle the projector that much to get it centered.
I just dont think putting the screen at a 5 deg angle will be that noticable.

I may wrong on this I guess trying it out will tell
Peter

Brain -
How should (i.e. which direction) should the pannel and the PJ lens be tilted to assist with increasing the keystone correction capacity? (perhaps a diagram?)

Everything would tilt in the same direction as the field lens. You could only get away with maybe a few degrees of tilt in the pj lens/panel until you ran into other problems.

Good call homie Why didn't think of that !?

Got the hemisphere mold today, NOT polished. It is heavy weight stainless, very nice piece. Now I am looking into polishing, notching and equiping these for use in our gear. I will make these available through the store. More soon.

ah, good man.

Brain, what is the smallest hemisphere mold you've come across, is it 6.25"?

Hmmm they go very small, down to an inch or less I believe.

If you come across a 4 or 5" one let me know, that would be perfect for my design at the moment.

When you provide the molds, will you allow submissions of custom designs? (In my case for example, where I'm using a longer lamp etc, and a lightbox where a smaller (less-deep) bowl is required)

I'd happily pay for the extra service (and plus I trust you, and I was impressed with the lenses, and the speedy overseas delivery)

Let me know what you think

Thanx

Sure I'd be happy to. Need to find out if these things can be polished in a big way. Going to check on that tomorrow.

Great, thanks. I'll keep my fingers crossed.

Hey guys, just a quick theory check:

It seems to me it should be possible to adjust for horizontal keystoning using the same methods except having the fresnel pivoting on the Y axis, no?

A setup like the one pictured below should work, right?

(Not using it in my current house, but still curious. I think it should work, but I'd like to hear from others...)

It does work, I tried it my prelims. I tilted my LCD slightly also. Your diagram is almost identical to a quick sketch I made, as this is how the pj will sit in my room.

Awesome!



They say great minds think alike...

It takes one to know one

Haven't been keeping up with all nine hundred posts made in the last week, have you? This was posted on Joe's new design thread.

What was that tony? Can you make the font any bigger, I can't quite read it from across the room!

Keystone can go both ways, after all, or does the name change if it is side angle adjustment you need?