Question about 3D and anamorphic lens... - Page 2

The text states dual projectors, so dual anamorphic lenses too?

One lens vertically stretches, the other horizontally stretches? Combine them into one lens and you have a spherical wide angle adapter. Same result as zooming.

Save your money and wait until the processor chips get fast enough.

So there's a name for it huh? Well it's not something I want to do. I think I have a solution for myself anyway. I am just going to be a two projector theater for the time being. I figured up the projector throw and while I can't zoom it, I can MOVE it further back from the screen to achieve my desired size for scope. This is not ideal either, of course. I'll keep my current projector and ISCO III for all else (which is like 99.9999999% of my viewing) and just employ the 3D projector when I want to experience some 3D. When prices go down or a better solution comes along, I'll just move the projector to another room.

Actually, I guess you could also get something from Navitar? Is that what their lenses are? I know they sell lenses that change a projector's throw distance, but again, that is not the best way to achieve the desired results.

Today I installed a new JVC X3 with the 3D kit and MK4 lens. This projector replaced a SONY WV60 which was previously used with the MK4 lens at a very short throw of just 1.4:1.



After a few adjustments, I managed to get the image perfectly squared (thanks to the H lens shift) and re-aligned the MK4 to deliver perfect focus corner to corner. It worked perfectly even though the JVC has a recessed lens, though it is boarder lining on vignetting due to the fact that glass moves inside the case. If the case was like an ISCO where it became shorter as the glass was wound in, this would not be a problem at all. Guess I am going back to the designing a new case now

So whilst the JVC won't Vertical Stretch 3D, I still wanted to test the anamorphic lens because of the recent Grid Distortion talk in the C5E thread. The MK4 uses multi-aspheric grinding in its lenses and is therefore corrected for GD (at most throws). However regardless of correction, GD just gets worse at shorter throws, so therefore at 1.4:1, there is no doubt going to be some present. I measured some single 1080 display lines both at the centre and at the edges and there was an increase and this would happen with ANY anamorphic lens at this short a throw.

So what I was concerned about was if this GD would mess with the 3D or not. Because 3D images are aligned vertically and it is their horizontal mis-alignment that provides depth, I was curious to see if the very short throw would affect the 3D. It appeared not, and I only wish I could have vertically stretched the image because 3D at that size (when compared to the letter boxed version) is very impressive.

Of course the HDMI lead was tidied up and I also blackened out the edges of the glass before re-assembling the lens. My camera batteries went flat, so I was unable to shoot the final install.

Looks nice Mark! There is talk that the Oppo 93 is in fact able to vertically stretch 3d, as long as the disc doesn't want to run java durring playback. This is just what I've read, no first hand experience. Oh, and you are able to use the MK4 with the X3 at a 1.4 TR? That's pretty sweet!

-Sean

It would sure be nice if that were the case. I am going to be supplementing my room with a separate low cost 3D ready projector and Optoma 3D-XL. If I could use the Oppo to vertically stretch the scope stuff, I could just get a good horizontal expansion anamorphic lens to get CIH! One day, one day, I want a single projector, 3 Chip DLP solution that doesn't cost more than my car, then I will be a happy camper. I hope someone can confirm the stretch feature in the Oppo soon.

Yeah it looked very good. The answer would be for the owner to get the latest EDGE which now supports 3D. If he was able to rip the 3D to his PC, he'd able to VS using that (maybe?).

My understanding is that both the Edge and Duo support passthru only, but will no perform any sort of processing on 3D material. That may be old info, but I don't think the ABT2010 chip in either machine has the bandwidth to process 3D.

Because a thread about my design has been invoked, I feel I must come in and correct Mark on some of his claims made here.


More than "talk". Actual numbers.

Furthermore, the discussion on the C5E thread was about eliminating distortion on a flat screen. Mark often omits to mention in such discussions that he uses a tailored, curved screen, which of course goes a long way to eliminating GD. It will with any lens, at any throw ratio.


There is no "therefore" about it. This is just voodoo optics.

Cylindrical lenses are not classed as true aspherics.

While cylindrical lenses are, at a pedantic, technical level, non-spherical in rotational profile, they are a trivial case of non-sphericality and are not commonly referred to in the industry as "aspherics". To do so would be to ascribe far too much sophistication to cylindrical lenses, which are a commonplace lens form.

An "aspheric" surface curvature is generally taken to mean either non-spherical or, in the special case of cylindrical lenses, non-circular in profile, e.g. parabolic, hyperbolic or a multi-term cubic function, to nominate three possibilities.

The cylindrical equivalent of an "aspheric" surface (i.e. one which has a non-circular, instead of a non-spherical profile) is probably better termed "atoric". If Mark is claiming his lenses are truly atoric, this would be a first, for him and for the industry.

Hence, an anamorphic adpater that uses cylindrical lenses is not truly "multi aspheric" (or if it is, then so is every other anamorphic adapter on the planet, bar none). Put bluntly, Mark's adapter does not "correct" for distortion. He uses a curved screen to do it, which is a whole lot simpler than all the "multi-aspheric" waffle.

If he has given up using his curved screen, and has "corrected for" GD even on flat screens, let him be the first to officially advise us all of this fact in plain English.

I emailed Oppo. Surprisingly, I got a reply just before New Year! If a title is not BD-JAVA enhanced, then it CAN be zoomed vertically. If a disc IS enhanced, it cannot be zoomed. Again, this is from Oppo customer support. So I guess it depends on just how many scope 3D titles are Java enhanced and how many are not. Time to do some research! Happy New Year.

Voodoo optics?

Sorry the images are not sharper.
Two different cylindrical anamorphic lenses.
Same 1080 DLP projector.
Same Throw Ratio (2.1:1)
SAME CURVED screen.
Same camera settings.





Sorry for the slight vignetting on the 2nd lens test pattern. I did not have a proper mount for it. You will note that the two out circles are not "round" in the 2nd photo. This lens uses a simpler design than the MK4.

This is really going to bug the hell out of those wanting 3D CIH.

Two different cylindrical lenses? Do you mean the Mk3 is the bottom test pattern? It sure looks like a prism lens to me as it has astigmatism. Care to complete the info list and tell us what it is please? What C lens does not have astigmatism correction??

Prism lenses have worse GD than all C-lenses. You and I both measured the mk3 GD over a year ago, it was about a 11% difference in width from centre to the outside edge from memory.

Scott has measured the 5E for GD up against the isco3, and the 5E has quite a bit less distortion from centre to edge, more even light distribution across the screen too.

This will take the thread way off topic. I posted initially because even though I could not vertically stretch the image, the MK4 lens did not disrupt the convergence/divergence used to make 3D work.


No, the other lens is a cylindrical lens, not a prisms/cylindrical hybrid like MK3+C. It has a neat rotating threaded sleeve system that moves the front lens in or out. I could not actually get this lens to focus properly from corner to corner and I do wonder if maybe if this lens has been dropped causing one of the lenses to have rotated slightly preventing it from focusing.

Anyway, the point being, under the same conditions, it produces an image with GD, where the MK4 does not appear to. And yes, I know that my curved screen does take care of some GD. I've since imported my images into paint and placed a single pixel line over the circles to measure the distortion. My centre circle appears to be compressed by just 1.5%.

I will re-test this using a flat surface and post the results soon.

Back on topic - went to see TRON LEGACY in 3D and whilst the image was Scope, they did not use an A-Lens. Disappointing...

Anyone know how to tell if a blu-ray is BD-JAVA enhanced before purchasing it?

OT: Mark what are your thoughts on the X3? Both 2d and 3d performance? Buy it or wait for a next gen?

-Sean

So it was a curved screen, and not the "multi aspheric" (i.e. ordinary cylindrical) lenses after all.

An appropriately curved screen is usually used to eliminate pincushion. A little-discussed by-product of this is that it should also pretty-well get rid of GD, much less than "some". A figure of < -0.3% GD right across the screen (at a TR of 2.1) should be readily achieveable.

This is only the 2nd time I've seen one of these JVC 3D projectors in action. Overall for what these systems cost, I am quite impressed. I just wish it was able to VS.

Flat screen shots are coming soon.


And I fully understand what your saying. So tell me, if the screen is doing all the "magic", why does it only do so for the MK4? What about the oval-ness of the outer circles of the OTHER lens?

The screen is indeed correcting the GD with the mk4. I set it up last night at a TR of 2.45:1, dialed it in perfectly and there was quite a bit more GD than with the C5 at the exact same throw.

You comparison was done on a curved screen, which corrects GD vs a lens which remains un named and may have been 'dropped'.....Not what you'd call a fair test.

I'll post screen shots tomorrow as it's sunday. Rest assured the GD is there and as you note will be even worse at shorter throws. My TR is quite long at 2.45:1.

I'm still not sure why this being debated. You're basically saying Aussie Bobs new design is no better in any area including GD. You're going to be very surprised!

I'm still not sure why this being debated. You're basically saying Aussie Bobs new design is no better in any area including GD. You're going to be very surprised![/quote]

That is not what I am saying at all. However feel free to read into it any way you want.

That is not what I am saying at all. However feel free to read into it any way you want.[/quote]

Will post some photos tomorrow. It's Sunday, go be with your family.

Are you serious?

Correcting GD and pincushion with an appropriately curved screen is one of the most trivial exercies in projection optics. They've been doing it with Cinemascope for 60 years. Once you get the screen curve right you can fix almost anything, much easier than you can by warping lens shapes, especially only two of them (as the test of the C5E against the Isco that you referred to demonstrates).

Here's a link to the official Cinemascope curved screen manual: http://www.widescreenmuseum.com/widescreen/table2.gif . It's about 58 years old.

If the GD with the other lens is still apparent there are at least three possibilities... probably many more...

1. On that particular screen, at that particular throw (and throw ratio) the MK4 works better than another, namless lens we know nothing about. No wonder... this screen was tailored for the MK4, wasn't it?

2. The other lens is most likely not expanding at the same ratio as the MK4. It may be closer to 1.33x, or it may be further away from 1.33x (although looks to be further away). This will drastically affect GD if you just swap lenses. We have no numbers and hence no details as to which lens is "right". We don't know where it came from, whether it was a discarded design, or a commercial lens you got cheap or were given as a present. Once again this other lens is completely anonymous, as are the test conditions.

3. You speculated that the other lens may have been dropped or otherwise misaligned. Who knows what damage or otherwise has been done to it?

If someone told you that standard cylindrical lenses are "multi aspheric" they were pulling your leg.

I'm confused about why there is any debate about any lens going on here at all? That's not what this thread is about, nor did Mark initially claim his lens to be better than another; all he said was that the GD talk got him curious and he wanted to test it mainly to see if it had any negative affect on 3d material.

Mark, you should pick up a 93 and check out the X3 in 3d using the full panel and a lens.

Back to my other question, does anyone know how to tell if a blu-ray is BD-JAVA enhanced before you buy it? If there is a way to tell, we should start a thread keeping track of which ones are (assuming there are fewer enhanced than not) so people don't buy a movie thinking it can be used in CIH when in fact it can't.

-Sean

You have to read between the lines, Sean.

The O.P. was about anamorphic lenses and 3D.

3D will work with any anamorphic lens, whether it distorts or not, whether it has CA or not. Aberrations will apply (and cancel each other out) to both phases of the equation - left eye and right eye. All an anamorphic lens really does is enlarge an image. The 3D metrics embedded in the screen content still apply. That's the short answer.

The next question is "How well will it work?"

The answer does not involve "multi aspherical" lenses, curved screens or allegedly perfect geometry. That is all voodoo optics, the "between the lines" stuff. I'm not surprised you don't get it, because it doesn't make sense, even to me. And I'm the only one around here who's gone to the trouble to personally design a new form of lens and make it work, verifiably, even up against the mighty Isco.

It's an old debate, for which facts and figures somehow never get supplied. Only assertions are made and never backed up, except with screen shots, which prove very little.

I'm as tired of it as you seem to be.

No. It was originally tailored for the MK2 at a much shorter throw. When I met you in person in 2008, I'd just got my BenQ W5000 and was still using the old MK2 lens (the very one that came to the test day) and the screen was then set to match the pincushion at the new longer throw of approx 2.1:1. If you can remember, I even called you asking about why the prisms needed more rotation to fill the screen at the longer throw. Since then, I've upgrade my lens twice (MK3 and now MK4) and yet I've never felt the need to change the radius of the screen.


You've actually tested this lens and if I remember correctly, you said it was more than 1.33x. I simply used it as an example of another design. I've not bothered to post its overlay simply because I am excepting the fact that it is irreverent here. What I am out to test and prove is the amount of GD with the MK4 and the MK4 alone. Nothing else matters here.


And why it is now irrelevant.


So they all lied to me and just wanted my money?

Are you going to suggest that an ISCO III has simple single radii designs for their lenses?

Are you going to suggest that it is not possible to design a lens with multi radii even though the optical engineers I met from WaveLength (a company that distributes and teaches optics using ZEMAX) say it can be done, just expensive to do.

And am I not to believe the optical designer who designed the lenses for the MK4 either?

What about the Australian company BAE who wanted over $6K for a design that used multi radii designs? Or what the another guy in the UK who wanted even more for a similar design? All described the design as multi asperic or multi radii.

So are you telling me that ALL of these people either don't know what they are on about or they just straight out lied?

Thank you Sean. The reason I posted my initial findings in this thread was simply because I had been able to run a 3D projector through an anamorphic lens and it appeared not to prevent the 3D from working which was pretty much why this thread was started in the first place.


Might have to import one. They are twice the listed price locally.

Heres a tip guys stick with 2D and everything will be fine

Yeah, except I've seen some good 3D and I want it now.

Which movies would you recommend from there large catalog Mark?