or Connect
AVS › AVS Forum › Display Devices › 2.35:1 Constant Image Height Chat › Cylindrical A-Lens Owner's Thread
New Posts  All Forums:Forum Nav:

Cylindrical A-Lens Owner's Thread - Page 11

post #301 of 623
Quote:
Originally Posted by coolrda View Post

My point is simply this. Until a native 2.35 projector based on masking the corresponding 16x9 chips horizontal resolution by 25%, the lens will always be equal too or superior to a native 2.35.

I'm having trouble understanding your point. You say:

"Until (projector A), the lens will always be equal too (sic) or superior to (projector B)"

where

Projector A = "native 2.35 projector based on masking the corresponding 16x9 chips horizontal resolution by 25%"

Projector B = "native 2.35"

What's the difference between A and B?

Starting with a source of 1920x800 pixels, it sounds like you feel that a 2.35 image consisting of 1920x1080 rectangular pixels (vertically upscaled then optically stretched by a-lens) is superior to 2530x1080 square pixels (vertically and horizontally upscaled).
post #302 of 623
Quote:
Originally Posted by ilsiu View Post


Projector A = "native 2.35 projector based on masking the corresponding 16x9 chips horizontal resolution by 25%"

Projector B = "native 2.35"

What's the difference between A and B?

Starting with a source of 1920x800 pixels, it sounds like you feel that a 2.35 image consisting of 1920x1080 rectangular pixels (vertically upscaled then optically stretched by a-lens) is superior to 2530x1080 square pixels (vertically and horizontally upscaled).

Projector A: Masked 1920 x 1080 panel down to 1920 x 810.

If your making down an existing 1920 x 1080 chip to 1920 x 810, then you still get the benifit of 1:1 pixel mapping for Scope, however your 16:9 image suffers with a re-scale by down rezing the 1080 back to 810 and down rezing the 1920 back to 1440.

Projector B: Native 2560 x 1080.

This projector will upscale both the H and V rez being currently 1920 x 810 to 2560 x 1080. The quality of the scaling remains to be seen.

This projector does however have the ability to map 16:9 program 1:1.

projector C: Native 1920 x 1080 + A-Lens

So at this stage, a 16:9 + A-Lens is still the best option because whilst it still scales Scope, it only does so in one direction leaving the 1920 pixels mapped 1:1, and if you move the lens you still get 1:1 mapping of 16:9.

In the cases like myself where I leave the lens in the light path, the only down rezing I have is for 16:9 where the 1920 is scaled back to 1440, however it retains the full 1080 rez so still has more pixels than the masked 16:9 chip.
post #303 of 623
Thread Starter 
Quote:
Originally Posted by ilsiu View Post

I'm having trouble understanding your point. You say:

"Until (projector A), the lens will always be equal too (sic) or superior to (projector B)"

where

Projector A = "native 2.35 projector based on masking the corresponding 16x9 chips horizontal resolution by 25%"

Projector B = "native 2.35"

What's the difference between A and B?

Starting with a source of 1920x800 pixels, it sounds like you feel that a 2.35 image consisting of 1920x1080 rectangular pixels (vertically upscaled then optically stretched by a-lens) is superior to 2530x1080 square pixels (vertically and horizontally upscaled).

ilsui,

That post was another iPhone casualty. Let me expand my thought and explain it better. Mark kind of covered it in another thread. Unless I've misread it, at this time 2.35 ratio DLP projector's, such as PD new 2560x1080 unit are using a masked WQXGA chip(16x10), which has a native res equals 2560x1600. PD's own F35 is a WQXGA projector used in the commercial market, even post-production. My point was running an F35 in 16x9 mode would yield a res of 2560x1440. Using this with and A-Lens would give you a 33% gain in horizontal res over the 2560x1080. Actually we've come full circle and are back to the 800 lines zoomed vs. 1080 + A-lens. We've already answered that. By native 2.35 I mean just that. At this time all DMD's are 16x10, 16x9 and maybe a few 4x3's. If a 3840x1600/2.40 chip is made than there's no benefit to using a lens as the max res is 1600. That's funny as 3840 is the res of a 4K chip, 2160x3840. In that case, here again the lens wins out, as it's 16x9.
post #304 of 623
Quote:
Originally Posted by coolrda View Post

At the last two CES they had a display set up to answer that very question. Overwhelmingly 1080 sources/2160 display was night and day difference over a pixel mapped 1080/1080. It goes beyond pixel fill and differences. Having viewed several 2160 displays, both 2160x3840'and 2160x4096, one thing stood. The complete transparency of the picture. The term organic is often used. The best analogy I can thing of is its like viewing a studio quality 16x20 photo, annabsolute facsimile without playback degradation. Until you've experience this you have know idea. My point is simply this. Until a native 2.35 projector based on masking the corresponding 16x9 chips horizontal resolution by 25%, the lens will always be equal too or superior to a native 2.35.

Sounds like a 'Monster Cable' type demo!!

I think we should agree to disagree.

There is no way image integrity can be maintained as image perfect as the original image displayed at it's original native capture ie 1920 x 1080 capture displayed at 1920 x 1080. Viewed at the appropriate distance for ones visual acuity.

I'm not saying that scaling cannot look good, it can, but it won't be as good as the original. Sure, if the resultant image is not large enough for the viewer and the size produced by zooming makes the pixel structure interfere with the desired quality of the projected image, scaling and using more pixels and an A-Lens will produce a better result under those circumstances.

But to say an A-Lens will under all circumstances produce a superior image is false, zooming at times can be superior to scaling and using an A-Lens and vise versa. Think MTF & ANSI etc. Think of UMR's observations.

If image size is not the deciding factor, viewed from the appropriate distance based on ones visual acuity, a masked 2.35:1 CIW will produce the best image quality, better than both zooming or scaling and using an A-Lens.

My next spend is to buy another motorised 16:9 screen, slightly larger than the current, mount it a bit closer to the viewing area, both screens will have horizontal masking, the larger screen masked to 2.35:1 will have the same area and vertical viewing angle as the non masked smaller screen(16:9), essentially will be a CIA, however as it's mounted closer, the setup will also be CVIA (Constant Vertical Image Angle) between both screens (larger screen masked to 2.35:1 and the smaller screen displaying 16:9). The plan is 17.5 degrees vertical for both.

The larger screen image will be dimmer, but that is a plus as it will reduce unwanted artifacts from being seen due to the image being bigger. If the encode is far from optimum, I can watch it on the smaller screen be it 16:9 or 2.35:1 masked...zero black bars on either screen, irrespective of the AR, 1.78:1 or larger that is.

Should be interesting.
post #305 of 623
Quote:
Originally Posted by rboster View Post

I think this is a great idea for a thread topic.


What would not be appropriate for this thread is a debate on A-lens vs no lens or vs zooming etc. That type of discussion should be considered off topic.

post #306 of 623
Thread Starter 
HighJinx,
Monster Cable Demo. Hardly. What struck me with all the 4k projector and panels I've viewed is the crossover to reality or the "Looking through the Window" look of the source material when compared to the same material on a 1080 display. It's clear to me that you haven't experienced this for yourself. Everyone that has seen it is absolutely amazed. My argument was clear. I prefer, as do most to all here, a Cylindrical A-Lens and full panel full resolution(from 16x9 projector) 2.35 image to a zoomed image. How many times have we read posts here of people moving to a lens and scaling from the zoom method and talk about the superiority of the picture. I don't think I've ever read anyone prefering zoom over scaling+A-Lens, which makes 1-1 scaling a moot point. With today's advances in scaling, one to one pixel mapping is irrelevant. I can pretty much guarantee you I can put you in a room with half a dozens different scaled and non scaled panels (one to one) and you couldn't tell which was which, which picture has dot for dot scaling and which is scaled. But I'm sure you could tell the diffence between a 2k sourced 2k picture and a 2k sourced 4k picture. Another thing, I base my statements on what I've seen in person which extensive and not based on someone else's findings. It's pretty easy to spot a person that deals in theory and one that has had hands-on experience. Having been to six or seven demo's of Joe Kane over the years with all types of equipment, I must say I may not agree with everything he says, but, atleast I've experience it. I'm not using someone else's opinion or posts. Just to wrap up in conclusion, and to clarify as clear as I can make it, I prefer, as do most here, a higher res'd picture with A lens to a 25% les res 2.35 chip for displaying 2.35/2.40 material. The only ones that don't prefer this is those that don't have an A-Lens, for whatever the reason, typically cost. Once again, the current 2.35 chips are sliced (dmd's turned off) 16:10 chips. Good night to you sir.
post #307 of 623
Thread Starter 
Quote:
Originally Posted by Highjinx View Post

Sounds like a 'Monster Cable' type demo!!

I think we should agree to disagree.

There is no way image integrity can be maintained as image perfect as the original image displayed at it's original native capture ie 1920 x 1080 capture displayed at 1920 x 1080. Viewed at the appropriate distance for ones visual acuity.

I'm not saying that scaling cannot look good, it can, but it won't be as good as the original. Sure, if the resultant image is not large enough for the viewer and the size produced by zooming makes the pixel structure interfere with the desired quality of the projected image, scaling and using more pixels and an A-Lens will produce a better result under those circumstances.

But to say an A-Lens will under all circumstances produce a superior image is false, zooming at times can be superior to scaling and using an A-Lens and vise versa. Think MTF & ANSI etc. Think of UMR's observations.

If image size is not the deciding factor, viewed from the appropriate distance based on ones visual acuity, a masked 2.35:1 CIW will produce the best image quality, better than both zooming or scaling and using an A-Lens.

My next spend is to buy another motorised 16:9 screen, slightly larger than the current, mount it a bit closer to the viewing area, both screens will have horizontal masking, the larger screen masked to 2.35:1 will have the same area and vertical viewing angle as the non masked smaller screen(16:9), essentially will be a CIA, however as it's mounted closer, the setup will also be CVIA (Constant Vertical Image Angle) between both screens (larger screen masked to 2.35:1 and the smaller screen displaying 16:9). The plan is 17.5 degrees vertical for both.

The larger screen image will be dimmer, but that is a plus as it will reduce unwanted artifacts from being seen due to the image being bigger. If the encode is far from optimum, I can watch it on the smaller screen be it 16:9 or 2.35:1 masked...zero black bars on either screen, irrespective of the AR, 1.78:1 or larger that is.

Should be interesting.

The one common factor in all mega-buck cinemascope home theaters is the use of a A-Lens, commonly a Isco IIIL. Most of the manufacturer endorse and even sell them as an add on kit. It makes sense as they what the best option for cinemascope. If this wasn't the best option the trade show demo's wouldn't be littered with these lenses. Mark's new lens is a CAVX MK4 and not a MK3a. Its different and better in every way.
post #308 of 623
Quote:
Originally Posted by coolrda View Post

The one common factor in all mega-buck cinemascope home theaters is the use of a A-Lens, commonly a Isco IIIL. Most of the manufacturer endorse and even sell them as an add on kit. It makes sense as they what the best option for cinemascope. If this wasn't the best option the trade show demo's wouldn't be littered with these lenses. Mark's new lens is a CAVX MK4 and not a MK3a. Its different and better in every way.

Ive seen some mega bucks use schneider
post #309 of 623
Quote:
Originally Posted by Franin View Post

Ive seen some mega bucks use schneider

Yeah like that monster XL prototype that has been shown at the Aussie CEDIA since 2008.
post #310 of 623
Thread Starter 
Quote:
Originally Posted by oztheatre View Post

Like they say Frank, 'nobody builds homes like they do in WA'

So many double brick homes over there.

Quote:
Originally Posted by Franin View Post

Ive seen some mega bucks use schneider

Ah, of course. Schneider is the Godfather of the cylindrical.
post #311 of 623
Schneider's "XL" is in production, has been for a while. I've sold several of them. It is an Isco IIIL in a different case. Identical (literally) glass. Isco (now Isco Division of Schneider) makes them both. You probably meant the MF Premiere. That's a semi-prototype *really* big lens. Can be Isco or Schneider branded. The one that shipped around to the shows is branded Schneider. They also show their great big 1.25x lens. That's a monster, too. But not applicable to our (normal) applications. History wise, Isco and Schnieder were the same company a log long tiem ago.

re Grandfather, I don't rememebr the details, but the President (strictly speaking - Managing Director) of Isco explained the path of divergence to me a couple of times. For large anamorphic history, Isco gets the nod. They were the only producer of a large format 1.33 lens for a long time.
post #312 of 623

Here is the that XL Schneider

LL
post #313 of 623
I would say, that the Isco 3L looks WAY better than the XL of Schneider
post #314 of 623
And I'd have to agree


LL
post #315 of 623
Quote:
Originally Posted by Widlarizer View Post

I would say, that the Isco 3L looks WAY better than the XL of Schneider

Quote:
Originally Posted by CAVX View Post

And I'd have to agree


Anti schneiders
post #316 of 623
The "XL" Mark posted a photo of is NOT the XL in production. That's a old prototype, never sold to my knowledge. The production XL looks very different from that lens. I think I have a photo somewhere, I'll post it if I do.
post #317 of 623
Thread Starter 
GG, I remember you gave me the scoop on Schneider/Isco timeline, leave it to me to screw that up. Those 1.25's are for the DCI 2k units with their 1080x2048 res. (2048/1080x1.25 = 1920/1080x1.33). I must agree as well. The Isco is the Victoria Secret Model of the A-lenses.
post #318 of 623
Quote:
Originally Posted by Alan Gouger View Post

Another cylinder lens is about to enter the market. Will be able to share details soon.
Top performance. Should put a dent in the market.

Quote:
Originally Posted by Aussie Bob View Post

Should read "alignment is highly important to equalizing pincushion top and bottom of screen". Pincushion with cylindricals is about the same as with prism lenses. Pincushion improvements require much more complex optics than just using a couple of cylindrical lenses.



Correct.



Should read "aligning the lens so that its optical axis is completely normal (i.e. perpendicular) to the light path and centered on the beam is just as critical for some lenses."

The more elements in the design that are in the light path, the less critical this alignment (especially horizontal alignment) is. This is because the distribution of curvatures is spread over more glass surfaces.

For example a lens with 4 glass-air surfaces (i.e. not counting cemented surfaces) will have more flexibility in alignment if all four surfaces are curved.

A 4 glass-air surface lens with two flat surfaces will only have the possibility of 2 curved surfaces. These curvatures must, of necessity, be quite steep (as the other two of the four are flat), and alignment problems can ensue with them.

A lens with, say, 6 glass-air surfaces has even more degrees of freedom, and thus more flexibility in alignment.



Wrongly put. Using a lens should "theoretically" be 33% brighter than zooming

There will always be a light loss with an A-lens. You are expanding the image in the horizontal direction by one-third. There's 33.3% light loss straight away. Then there's loss through the lens (due to internal reflection and absorption by the glass surfaces).

On the other hand, with zooming you're expanding the image bioth vertically and hosizontally by one third. The Zoom Method arithmetic is 4/3 x 4/3 = a 16/9 increase in image area. Inverting this number you get the relative zoomed brightness (compared to pre-zoomed brightness): 9/16 (56.25% of the pre-zoomed brightness). However you're also enlarging the image which, with projector zoom lenses, makes the aperture of the lens effectively larger (relative to the shorter focal length of a wide angle zoomed lens). This is equivalent to reducing the f/number, so there less loss than you might think... the f/number of an optical system is the focal length (shorter when zoomed) divided by the (fixed) physical aperture.

Compare this greater light transmission efficiency of a zoomed lens with the light loss from using an anamorphic lens and the final 'scope brightnesses are closer than you might think. An A-lens should still deliver greater bruightness than a zoomed lens, but not by 33.3%... more like 15%-20% (and sometimes they're about line-ball, depending on the projector, throw ratio etc. used). On the other hand, an A-lens will most likely never be dimmer thanthe Zoom Method...



Make that "another two" lenses to be available soon.

I understand that at least one these two new lenses will be a lightweight, low-distortion, 5-element design, arranged in 3 groups, which is a much-needed improvement on the traditional Schenider/Isco (and other) 4-element, 2-group designs.


Have we reached a point in time where any new information about these lenses can be released?

Pricing? Availability?
post #319 of 623
Hi all,

Any news about the Schneider Cinedigitar 1.33x LE new model to be launched late Sep - Oct early?
post #320 of 623
Quote:
Originally Posted by eummagic View Post

Hi all,

Any news about the Schneider Cinedigitar 1.33x LE new model to be launched late Sep - Oct early?

Apparently its been delayed - again.

plastic - good for 720, not so good for 1080 or so i was told
post #321 of 623
Quote:
Originally Posted by CAVX View Post


Apparently its been delayed - again.

plastic - good for 720, not so good for 1080 or so i was told

Huh?
post #322 of 623
I assume that means the cheaper LE was to be using plastic elements rather than glass.

Gary
post #323 of 623
Quote:
Originally Posted by Gary Lightfoot View Post

I assume that means the cheaper LE was to be using plastic elements rather than glass.

Gary

Yeah but Gary why it's only good for 720p?
post #324 of 623
Probably because plastic doesn't have the same optical properties as glass, so doesn't pass light as well - maybe more refraction or less MTF capable? That would mean smaller pixel sizes wouldn't resolve as well as larger ones

Gary
post #325 of 623
Quote:
Originally Posted by Gary Lightfoot View Post

Probably because plastic doesn't have the same optical properties as glass, so doesn't pass light as well - maybe more refraction or less MTF capable? That would mean smaller pixel sizes wouldn't resolve as well as larger ones

Gary

Oh ok, thanks!
post #326 of 623
I'm sure Aussie Bob or Mark will give us a more accurate answer though
post #327 of 623
Guys, I've a question about throw ratio. I've seen mentioned that A-Lenses shouldn't be used for throw ratios smaller than 1.5x. But what throw ratio are you refering to here? Is it the throw ratio with or without using the A-Lens to fill the Cinemascope screen?

CIH screen width: 280cm
distance to projector: 392cm
throw ratio for 2.35:1 movies: 1.40x
throw ratio for 1.87:1 movies: 1.76x

Is my throw ratio ok for an A-Lens or is it too small? Would any A-Lens do, or would it have to be an Isco IIIL?

To be honest, one major reason for my considering of using an A-Lens is that it might allow me to use projectors which I couldn't use otherwise. E.g. the TruVue Vango only goes down to 1.5x. Using zoom only the Vango isn't able to fill my screen. But with an (HE) A-Lens it would fit. But again that makes me wonder if my throw ratio is too small for an A-Lens or not, respectively if I could do with one of the upcoming bang-for-the-buck A-Lenses or if I needed the very expensive Isco IIIL.
post #328 of 623
Smaller throw ratios are not recommended because the size of the image coming out of the projector could be too large for the A lens, so the sides of the image can be obstructed by the A lens casing. In photography they call it vignetting.

With larger lenses, a smaller throw can be accommodated, but because you're using more of the A lens' glass area, you can get more pincushion. With a smaller beam, you use less curve on the A lens glass, so get less pincushion.

Gary
post #329 of 623
Gary, I understand that. But my question still stands: Which throw ratio is "too small"? And especially: When you say a number (like 1.5x), do you mean the throw ratio before or after horizontal expansion?
post #330 of 623
It's the throw before expansion, and will vary depending on the A lens.

With an ISCO IIIL for instance, you might be able to use a 1.5 throw because it has a larger aperture, but you won't get away with that on something like one of the small Schneider lenses, or the smaller prism lenses. The beam exiting the pj will be too large.

Sometimes a pj will have a recessed lens, and that will mean a longer throw will be needed compared to a similar pj with a flush or protruding lens.

Gary
New Posts  All Forums:Forum Nav:
  Return Home
AVS › AVS Forum › Display Devices › 2.35:1 Constant Image Height Chat › Cylindrical A-Lens Owner's Thread