Popular Myths About 3D - Page 5

I would like to hear a technical explanation of why polarized technology and FPR cannot be applied to plasma display tech, only LCD? All the (passive) polarized mfrs. are applying the tech exclusively to AM-LCD.

Is it because the (inside of the) screens must be coated with a polarizer and this conflicts in some way with the emissive nature of PDP?

Secondly, I've found that both polarized and active shutter eyewear cut light output significantly (as much as 70%). Is it true that AS reduces luminance MORE than passive glasses?

Many theaters misuse 3-D lenses to show 2-D films, squandering brightness, color

http://www.boston.com/ae/movies/arti...e_dark/?page=1


http://entertainment.slashdot.org/st...r-Local-Cinema



The inability to cam a theatrical 3D presentation is probably the biggest culprit in the use of 3D for too many films, IMO.

I had laser eye surgery, largely in part because I didn't want to wear 2 pair of glasses. One I'm great with. lol

Sorry this is quite long, but I wanted to try and be complete.

In general, most of the DLP installations are doing frame alternating 3D where the projector switches between showing the left eye and then the right eye, typically at triple frame rate. So the projector is showing 144 frames per second, 3 left and 3 right, times the 24 frames per second. There are 4 different systems in use presenting this triple flash 3D to the audience. Real-D which uses either a "Z-screen" or the "XL system". Dolby 3D which uses alternate RGB color filtering, Master Image which uses a spinning disk of alternating polarizing segments. and Xpand using shutter glasses. All 4 systems have some advantages and dissadvantages. Real D with a single Z-screen was the first system put into wide use. It is not very light efficient and suffers from a bit of image softness and ghosting/crosstalk. If used with a bright enough projector and carefuly setup, it does provide a very good image with only a minor loss in sharpness and color saturation due to projecting the image through the LCD polarizing Z-screen. Real-D XL is an improvement in many ways, but also adds some new problems. It's biggest improvement is in light output. Well over 40% more light gets to the screen comared to a single Z-screen system, but the light out of the projector is split into 2 paths and have to be recombined back at the screen. I have seen it line up perfectly with ease, but I also spent a solid hour tweaking one and was unable to get the image split to less then 4 pixels at al the edges. I think the box had been dropped and the prisms were mis aligned on that one. The XL box does have 2 images coming out of it, stacked vertically at the projection port, but they are both showing both eyes at full resolution. According to Real-D, 2D should NOT be projected through the box or Z-screen. When the projector is operating in a 3D mode, it is sending a sync signal to drive the LCD active polarizers. When running a 2D feature, this signal is not there and the panels do not behave correctly and could even be damaged. 2D contant such as trailers are often shown before a 3D feature in 3D mode and this is not a problemas it just shows the same image to both eyes. A full 2D feature could be shown in 3D mode this way, but is not done in any of the theatres I have been in. Unfortunately, the real-D screen or box on the front of DLP projectors is not normally motorized to move out of the way, so it does sometimes get forgotten and a few shows run through it. The effect on the picture is not horrible, but can be visible. A little dimmer and softer, and you may see the pulsing of the non synced LCD's. The XL boxes do fall back to a fairly fast mode so it will not damage the LCD's and the flicker is usually fast enough to not be visible. At least two other companies are making motor kits to push the box out of the way when not in a 3D mode, but theatre owners tend to be cheap. Master Image is very similar to Real D in that it is using alternate circular polarizing but instead of using an LCD polarizer it mechanically changes the filters by spinning them in front of the projector on a large motor. It uses 12 segments so the amount of tilt is fairly small. The light efficiency is right between the original Z-screen and the XL system from Real-D. Both systems can use the same glasses and require the silver screen to hold the polarization. Master Image does have one very nice feature, it comes with a motor system to move the spinning disk in and out of the projectors light path to switch in and out of 3D without human help. Dolby 3D is a bit different. Since it is using the shifted colors it works on white screens. The light efficiency is a bit less since it does not have the help of the high gain silver screen, but it also does not suffer the hot spotting and color shifts caused by the silver either. The color filtering is done in the projector before the DLP chips so the image from the chips to the screen is not going through anything extra. The left end right eyes are using different primary colors so the calibration is more involved and care has to be taken to get it right, but when done correctly, the colors are very accurate and vivid without the loss of sharpness and contrast suffered by going through the polarizers or other extra glass. The lower light in the center of the screen is easilly made up for with the much more even light right out to the corners. In many cases people think it is brighter overall when using the same projector and lamp size since shadow detail at the bottom and sides of the screen are lit better. All Dolby 3D systems using a single projector will use only one lens and the filter wheel is powered to move out of the light path during 2D operation. Xpand also works on white screens and has the more even light advantage. The glasses are LCD shutters so they suffer similar light loss to Dolby 3D or a single Real-D Z-Screen. There is nothing in the light path from the projector to the screen, the only thing effecting the image is the glasses on your face. Some people swear by the active glasses, other swear at them. They are heavier and need battteries. They are expensive and fragile compared to the passive glasses. Dolby glasses are also more expensive than the polarized ones and are therefore reused and washed like the Xpand ones, but are far more durable and don't need batteries.

Some larger screen theatres have gone to dual projectors. Digital IMAX is also a dual projector system. Real-D and Dolby can do dual projectors in 2 ways. The best is to dedicate one machine to each eye like IMAX does using static polarizers or fixed color shift filters in each machine. When running in 2D, it is common to then use only one machine, alternating on a show by show or day by day basis to keep the lamps at the same life. IMAX still overlays the two images and claims it increases resolution, but even if perfectly aligned, I see that as a tough act to prove. The DLP pixels do not leave much gap at all to fill with the other projectors pixel, but it does blend away the pixel edges a bit.

Xpand using 2 machines still has to have the machines in sync alternating eyes to match the glasses, so there is less advantage to 2 machines, you only get more light, nothing else. Being able to dedicate a machine to each eye eliminates the lag between the eyes where one eye is seeing the new frame first. This can cause a blurring of shuddering look in fast motion and even a loss of 3D depth on horizontal motion.

The SONY 4K projectors are using SXRD, Sony's own version of LCOS imagers. They are not able to alternate between the 2 eye images at 144 hz like the DLP's can. So they had to do 3D in a different way. After quite a bit of testing they came up with the dual lens system causing all the uproar in the Boston Globe. Basically, they went back to the way 3D was done on film in the 80's. They split the imager in half vertically. One half is always showing the right eye, the other half is always showing the left eye. This does give both eyes the new frame at the same time so it does work very well in moderate motion. The problem is that the 3D lens is a beast. It needs to be installed and aligned very carefully to get the two images perfectly overlayed. The 3D lens also does not have any provision for automated adjustments. So the images need to be scaled and fit to the screen depending on the format between flat and scope. Much of the imager ends up blanked and off the screen. This results in a lower resolution, and light output compared to using the full imager for 2D. Since the Boston Globe article came out, several theatres have started changing the lenses for bigger features. Maximum light output is cut to less than 1/2 when the 3D lens is installed versus the 2D only lens. Several theatres have also pulled fairly new SONY projectors out of larger screens as they just could not put out enough light. The DLP projectors not only hold a larger bulb, they put out more of the light the bulb creates. SONY claims 18,000 lumens out of the model 220 projector with a 4,000 watt bulb, the 320 may be a bit better, but I don't have the spec. The typical DLP with a 6,000 watt bulb is over 30,000 lumens. Barco is using a new 6,500 watt bulb and putting out close to 40,000 lumens.

TI is starting to ship 4K DLP projectors. Currently the DCI spec for cinema does not include a 4K 3D format, but one is going to come. For now, all 3D is only 2K resolution. IMAX does label their shows as 4K 3D, but using two 2K DLP projectors is not going to make a 4K image. True 4K projectors like the Sony and new 4K DLP's have 4 times te pixels, double in both directions.

If I did not answer a question you have, please post it or PM me and I will try to clarify anything I may have missed.

GXM

From

http://hardware.slashdot.org/story/1...y-Eye-Tracking

Showscan used 60 full frames per second to give your eyes a lot more information about an movement of an object. Moving objects give away their depth by the change in ange over time compared to other objects in view. Our brains have been taight over many years to understand all of the distance cues together. Chopping time into 24 frames per second is way too slow and we lose a lot of fine detail motion information. Larger motions and moderate speed motion still does give s some depth cues though, and this does cause a problem for stereo 3D. If the depth cue from stereo vision does not match the depth cue you get from motion paralax, your brain knows something is wrong. The size of the screen and the seating distance effects this and makes it so the contant is a compromise for "avaerage" theatres. Most likely it will be way off when put onto a tiny 50 inch screen in a home. At a theatre, the 2.5 inches between your eyes is a very small fraction of screen width, but at just 40 inches wide or so, the width between your eyes is a much larger fraction of screen width. I have seen the effect of the opposite, content made on a 40 inch plazma was ripping my eyes out on a 60 foot wide screen.

I made two new videos and this time they are in 3D side by side format, 720p resolution, left side first.

link

There is two videos, one is 23.976 fps per eye and the other is 47.952 fps per eye.
As before the goal is to see them move in the same line as the line moves from the left to the right.

I read some replies saying they had 240Hz or 120Hz TV's and this fixes any flickering, well I made a video that uses blu ray 3D speed and now you can see if your fast tv stops flickering.

Please try the test videos and reply with how you did as I find that interesting enough I did this work.

Stereographers get around this by adjusting the convergence ever so slightly going from theater to smaller screens, and vice versa.

In what way do we see 3D with a single eye in reality?

Very interesting. I thought that you had to have at least 2 eyes to see 3D in the real world. Ideally, 3 eyes would be better (one above the other two) but how could you wear glasses.

We've established that you are saying people can see in 3D with one eye. But what way do we see 3D in real life with one eye, and how is it different from a monoscopic movie?


Resolution, size and framerate don't increase depth perception. They increase immersion, which 3D is also about.

There are plenty of depth cues to give us a good idea of the third dimension in 2D photography. The only way real life is better than movies is that we are in the driver's seat controlling focal length and lateral head movements, whereas the director/DOP is controlling the movie's visuals.

We have two 2D cameras in our head. We can control the focal length (accommodation) and truck the cameras laterally by moving our body, and with both eyes open we can toe the cameras in and out to convergence on things and identify their Z axis location without the need of slow head movements or focal adjustments.


What do you base this on? Please explain. Use photography terms.

Huh? A camera can have shallow depth of field, deep depth of field, whichever it wants. It can focus on something far away or an inch away.


just contrast ratio? How about shallow depth of field? How about motion parallax? How about occlusion? Perspective?




Yes, you keep saying that. And I keep asking you to explain yourself. And you just keep repeating it. That's not explaining.



Shooting at higher fps means you have to use a faster shutter speed, leaving less time for exposure, reducing brightness, and requiring a wider aperture, causing a shallower depth of field. Showscan is about framerate (60fps) and film stock (65mm), improving clarity and eliminating flicker. Nothing about depth. Absolutely nothing. You've just made that up along with everything else.

Your opinions on 3D are strange, I'll be nice and put it like that.

No? I refuse to accept the explanation because it is full of holes.

You said, "Our eyes give us infinite focus, something a camera lens can't." Which is false. You then adjusted it to, "..with a single lens." Why should that matter? Switch lenses, take the shot. Next setup, choose the best lens for the job. Edited together everything is seamless. It should be a plus that cameras have that interchangeable lens flexibility, but you've made it into a negative somehow. Please explain.

Our eyes have a finite control over depth of field. A camera lens can really crank the depth of field very shallow or very deep, and can focus on very near things or very far. Our eyes can't focus on macro stuff, it can't control the depth of field, just focal length.


Please explain "the ability to increase focus as you increase depth of field." You mean distant objects becoming clear as you increase depth of field? That doesn't make sense to me in terms of why that's better for discerning one plane from the other-- shallower depth of field, not deeper is what queues depth.

The most popular myth about 3D is that it's all about things coming at your head, or the "I hate 3D but I haven't tried it lately" whiners. I just find it odd that people, especially those in the film industry, don't take the time to see what all the hype is truly about.


http://gizmodo.com/5830261/c+3po-hates-3dtv-too



So we have people hating 3D because a theme park ride or cheesy horror movie gave them the impression that popout is all they're about; and we have people hating 3D without even sampling the most game-changing 3D in film history.

No. All I know is what I've read, and it all conflicts with what you've said.

P.S. The depth "advantages" you cite are referred to in the "downsides" section of this article: http://www.barbeefilm.com/showscan%2...It%20Works.htm

The shallower depth of field is a result of opening the aperture to let more light in to compensate for the lower exposure caused by the faster shutter speeds required.

In other words, the depth-related effects of the Showscan system were not done for artistic reasons, and can easily be replicated in non-Showscan systems.

You really hurt your credibility when you said we see 3D with one eye, and then dodged my requests for a thorough explanation. The fact is you can't give a more thorough explanation because you just don't know what you're talking about.

"On the downside of the equation, the increased frame-rate takes light away from the negative. One and one quarter stops to be exact. Also in 65mm there is the approximate doubling of the focal length of the lenses, with the attendant shorter depth of field, for any given shot. This, coupled with the exposure loss can be troublesome, since you end up working at longer focal lengths with larger apertures than you might in 35mm."

It's obviously not all advantages.

I'm not saying you lost credibility for saying we see 3D with one eye; it was the sequence that it spawned, of me asking for more detail and you responding with non-answers and such. You continue to ignore everything I've said and asked of you. You're too childish to have a technical conversation with, as you never expand on anything I ask you to. I apparently caught you off-guard when you thought you could just say whatever you want and expect the gullible to nod in agreement. Stop pretending to be some hotshot expert on 3D. If you're such an expert, my requests for detailed technical explanations would be of no challenge to you and you'd be happy to share the wealth of knowledge and expertise. I inquired of you and you failed miserably.

No.


I don't need to Google how the eye sees. I already know. You just need to tell me what you find to be so drastically different in the human eye compared to a camera. Because frankly, the only advantage my eye has over a recorded image is the ability to control the muscles used for accommodation, which only has an effective range of a few feet, and requires the scene to be slow-moving or static so as not to add motion blur. Binocular vision is far more effective, as I can detect the dimensionality of a complex network of leaves and branches pretty much instantly. It's also why I can juggle 3 fast moving objects at a time. As a kid I'd close one eye and I'd always fumble the next ball that came down.

I kind of see what you're saying, as there are tons of depth clues that one eye can see, but calling that 3D is quite a stretch. Let's try this together: hold out your index fingers and point them at each other about a foot away. Close one eye. Now try to put the tips of your fingers together. Oops! I missed! But my one eye is 3D!!?!

Of course that one eye still has a number of depth clues so there's a good chance you were able to put your finger tips together (we're also aided by a non-visual sense, where our brain knows where our body parts are, so you can walk without watching your legs). Regardless, it's limited clues and there's a fair chance your finger tips will glide right by. Try it again with both eyes open. Try it 100 times, you'll never miss. This is (full) 3D. The addition of a second disparate image is a powerful clue. Don't cheapen it or suggest that one eye is just as good. We would not have evolved to have two front facing eyes if one eye could do the job.

Not a very good test, IMO. I could touch my figers every time. Try this (to eliminate "body part knowledge"). Place your remote control on the table in front of you. With one eye closed, choose a button, then try to touch it with your finger. Repeat with random buttons.

I never missed with that test either.

Now, go to a batting cage...