Official AVS Crosstalk Topic

Yes they all show the ghosting in the same place and the scenes where the ghosting occurs contain bright light colored or white content. And if the glases do are slow in responding or can only make the lens for the inactive eye are only 99% opaque and not 100% opaque then ghosting will occur.
If you think the ghosting is part of the actual content what do you think the makers of the content are doing wrong to make it occur.

Now, I'm no professional on the matter but the way I see it, 3D simulates depth but actually stimulates the eyes like naked sight. What we call crosstalk happens through the naked eye has well but we are used to it and so I'm guessing that brain waves adapts these with the depth calculation that stereoscopic vision provides...

Put your hand about 10 centimeters in front of your eye, you will "see through part of your hand" since your other eye sees what's behind. That would be the same as crosstalk and becomes noticeable since the naked eye doesn't normally process at such distance.

My point is, the more you watch a 3D TV, the more your eye adapts and processes the crosstalk since it learns and works muscles to refocus quickly so that you don't see it as much. But there is no standards as of now so it may take some time and probably training before the technical capabilities of 3D come out as crisp as it really is.

That would also explain the few reports of people fixing vision problems with 3D viewing and also why 3D in the theater begins as uncomfortable but after a few minutes becomes more natural and lifelike.

Keep in mind that this is all based on my logic, I'm no scientist!

I don't remember much cross-talk with the Panny VT20 and Wipeout HD. I don't have that Panny anymore (too much eye-fatigue from ASG for my liking), but I played a lot of Wipeout HD back when I had it and didn't notice any cross-talk. There might have been some minor cross-talk here and there, but nothing that made me sit up and notice it.

Sorry, guys, but I just don't see how anything that's been said here confirms that content is the cause of crosstalk. Content is the source of ghosting if the content creates opportunities for the content of one eye to be perceived by the other eye - such as in high contrast subject matter that overlaps bright and dark parts of the two images.

The real culprit, though, is not content; it's the inability of current active shutter glasses to block 100% of the light from the other eye. If the glasses are properly synced (that is, if one eye of the shutter glasses is as closed as it can be while the opposite image is on the screen), and we still see a ghost image, then the problem is obviously the glasses. They just don't block out enough light.

Content causes ghosting in the sense that current shutter glass technology is inadequate to the task of allowing each eye to see only what it's supposed to see. I suppose we can blame content creators for producing movies with a lot of ghosting, but we should really blame the manufacturers of the glasses. One of these days, there will be a technology that blocks enough of the light that we won't see any ghosting. Until then, we just have to live with it if we want to enjoy 3D.

If you doubt this, check out AVSer Frank's test images with your own shutter glasses (Samsung, Panasonic, Sony, whatever). Frank has created some still images in side by side format. One side has Frank's dog, while the other side is black. When this side by side is turned into 3D by the television's processor, one eye should see nothing. Close the eye with the image, though, and you can see how much ghosting there is in the "black" frame. It's startlingly obvious how ineffective shutter glasses can be. Sure, you could blame Frank for creating content that has crosstalk, but the test shows quite clearly that the glasses are at fault.

Right now, it seems to me, content producers can lower contrast and create less dramatic shots, thus reducing ghosting, but that's a compromise, too. Even though I appreciate the lack of ghosting in "Cloudy with a Chance of Meatballs," I like the contrast and color scheme better in "Monsters vs Aliens," even though the ghosting is much worse.

I too was puzzled reading the proclamation of confirmation in the title, then not finding it in the thread. But in the same vein, while many many are claiming that the issue is how much light the glasses can (or can't) block in the "closed" state, it is possible that it's still a response time issue as opposed to an absolute "opaqueness" issue. DLP and plasma both use PWM to control light output. Both the mirrors, and the electrodes (respectively) are on/off devices. Brightness is determined by how long they are "on". If a dim pixel is produced by a very short "on" time in the middle of the frame, while 100% brightness is generated by being on for the duration of the frame, you can have differences in light leakage through the lens between dim and bright scenes due to slow response in the glasses, even if they block 100% when closed. If the glasses are transitioning during the beginning and ending parts of the frame, they could effectively block "dim" output that is only generated during the middle of the frame, but allow some leakage on bright pixels where the pixel is already (and still) on while the glasses are transitioning.

To the best of my knowledge, LCDs aren't PWM designs, they produce dim/bright in an analog voltage sense. But LCD displays have their own lag time (as do phosphors in plasma).

I'm not suggesting this is the issue, simply that there may be multiple possible causes. But I do agree it's not a "content" issue. It's a technology maturity issue. Like most undesirable traits in technology, some content can expose weaknesses better than others.

Certainly, a lot of things can contribute to ghosting, but content is the last thing we should single out. One day, hopefully, ghosting will be a thing of the past. We shouldn't ask content producers to compromise their work now because the technology isn't where it should be. If the technology worked the way it should, there would be no thought of placing limitations on how content is created.

Although ghosting is, at this point in the history of 3D, the thing that bothers me most about the first generation products, I'm going to make every effort not to stress about it. It will get better. I'm just happy 3D is here at all.

Not to get caught up on semantics here, but I think what folks are saying (myself included) is that the content is not the cause. The cause is the display technologies themselves and, to a lesser extent, the glasses.

Certain types of content expose deficiencies in the display and glasses technologies. But it's unfair to say the content is the culprit.

That's like saying that the fast chaotic movement of splashing water is the culprit in macroblocking in broadcast HD. The culprit is that the bandwidth is insufficient for accurate reproduction of that type of content. The solution isn't to stop brodcasting swimming events. The proper solution is to stop bit-starving HD broadcasts.

The lack of sufficient bandwidth is the culprit. Certain content simply exposes that weakness. When hardware is incapable of accurately reproducing real content, the hardware is the problem, not the content. If the weakness could only be observed by artificial test content, you could make the argument that the hardware is "good enough". But when dealing with real content which is trying to portray real life images, you don't want to dumb the content down to what the hardware can handle. The hardware needs to be improved.

Some users have reported that they have been able to watch MvA without any ghoisting.
Both slow response time and inability to get 100% opaque are ways the glaswes cause ghosting. In an anouncment about the new Xpand 103 Xpand stated that these are the fastest glases they have ever made and this will reduce ghosting.

Good analogy.

Content is definitely not the culprit. If we could fix the parts of the 3D technology that don't work as well as they should, would ANY content be a cause of crosstalk?

If shutter glasses blocked 100% of the light from the eye that wasn't supposed to see it...

If sync timing were perfect, so that LCD shutter glasses transitioned only at the precise point in time when they were supposed to...

If the 3D display could refresh itself so that no part of the other eye's image remained on the screen when it wasn't supposed to be there...

And if every other part of 3D technology prevented light leakage...

then, nothing in any 3D content could cause crosstalk. If what one eye sees is complete and utter darkness, while the other eye sees exactly what it should see, crosstalk is zero. Ghosting is impossible. Blame glasses, blame synchronization, blame LCD lag, blame slow phosphors. Blame any other part of the technology itself that contributes to the crosstalk, but don't blame the content.

You're off base. There are a number of factors that could cause or contribute to crosstalk, including the content (or how it's stored and played back), display, glasses, sync, etc. Trying to simplify a complex subject does not help.

Topic title edited.

wipeout hd is a high framerate game. It was dumbed down for 3d they say but it may still be high framerate.

In it's current state, full frame sequential 3d acts like interpolation: interpolation lag. I'm saying there's lag.

A frame is shown on the screen, and it is going at a framerate that does not slow down.
And the display is able to show a different realtime framerate without crosstalk.

If the glasses has to go at a framerate of 30fps, ~ 33ms (milliseconds), and it took the glasses 41ms then the framerate of the game would be too fast for the glasses to keep up with.

As it is movie content is 24fps, ~ 41ms, and the shutter glasses are showing crosstalk, so of course 33ms will show crosstalk too.

And the game wipeout hd 3d is 30 fps, or ~ 33ms. So there, now you see why.

So the Wipeout HD game is not designed to be run on hardware/software and with glases that cannot refresh at only 30fps per eye since it as you stated a highframe rate game.
It is not the content of the game that is wrong since it will work on systems for which it is designed.

Whoa, wait.... 30fps per eye?!? The framerate movies are encoded at isn't the same as the framerate of the glasses. I'm not so sure about the current state of other 3d displays, but DLP is always 60hz per eye. People would know (and complain) if they were at 30hz. Some of the glasses I've seen say they are good up to 120hz.

The reason your idea has come under such close scrutiny is your bold claims:

"Source of crosstalk confirmed!!"

And "Crosstalk IMO is entirely content based."

Those are sweeping generalizations that simply aren't true. We all know that certain content exhibits ghosting more than other content. Personally, I have every confidence that one day 3D technology will be refined to a point where ghosting is a thing of the past, at which point most (if not all) of the content that you are blaming for crosstalk will be exonerated.

The Nvidia 3D game softare tranmits frame sequenial content in 720p/60 format which is therefore 30 fps per eye. So a 120Hz DLP system will just display each 30fps frame twice.

In that case, the glasses would still be operating at 60hz. So it wouldn't be an issue of the glasses "not being able to keep up".

I will rephrase what I said.

The glasses go 60 fps or 16ms g2g, per eye.
24p per eye uses a g2g of 41ms.
30fps per eye uses a g2g of 33ms.

The fast scenes in 24p when shown in 60fps may show blurriness or afterimage/crosstalk.
Then look at the 30fps content compared to the slow moving 24p content.
The 30p content behaves similarly to 24p fast content.

Now maybe this isn't always true. maybe on slow panning scenes the 30p behaves like slow 24p content, but the game wipeoutHD is a reflex game with fast moving scenery.

Why is 30p content behaving like fast 24p content? Because the g2g of the 30p is much faster than the g2g of the 24p content by default. Right now the g2g isn't showing 24p content properly, so now it needs a better performance standard than even 24p it is insufficient.

And perhaps I should too, as I don't think I made my point clear. I think it's easier if we break this down into two separate issues: the response of the glasses, and the framerate of the display.

The response of the glasses is... whatever it is. I thought in my recent research of glasses, I saw some mention of them going up to 120hz (each eye). I'll have to look back again and see where I read that. If they DO have fast enough response for 120hz, then 60hz should be no problem. But of course, there's manufacturer's ratings, and real world. But regardless of what the response of the glasses are, my point was simply that, at least in the case of DLP, the screen refresh is always 120hz, so the glasses are always 60hz per eye. So regardless of what the glasses can do, they should do it consistently regardless of the content. Yes, 60hz (full) framerate can be considered slow for fast action games. But that's what you're limited to with console games anyway. You need PC games and a good monitor to get faster. So whether or not you're watching in 3d mode, whatever the content is, it's going to get cut up into 60 snapshots per second. Film can show blurriness due to framerate, but that blurriness is created at the point of capture. Once it's transferred to video and displayed on our screens, whatever blurriness there is is recreated as a static (blurry) snapshot, regardless of the displayed framerate because the blurriness was captured, not created during display. With video games, there's not typically blurriness (unless it's intentionally created for realism). You might have tearing if v-sync isn't being used, but not blurriness.

So 3d glasses or not, you have a series of snapshots being displayed at 60fps (with 120hz screen refresh). Since the framerate is constant, regardless of content (at least on DLP), I just don't follow how the glasses can be causing rainbows based on whether or not it's a "fast" game. With a fast game, the action might get chopped up into too few frames for good response (from a gaming reaction perspective), but as far as the glasses and our eyes are concerned, it's still just a series of static snapshots.

I would still contend that it's more about the image content (coloring, brightness) than the framerate, since the framerate is always the same.

EDIT: there is a 3rd possibility... I don't know what 3d format these games are using, but it seems like it could be possible that the 3d conversion process is a factor. For example, SBS encoding requires a horizontaly compressed image (half the horizontal resolution of 2d) which must be stretched with interpolation. Image scaling can cause moire, and different patterns can make moire more or less apparent. This could be compounded on DLP systems, since they also have compromised resolution in 3d mode, but it's a different compromise than SBS (the former missing every other pixel in a checkerboard pattern, the latter missing every other vertical line). So there's kind of a double loss there when viewing SBS on a checkerboard display. But IF that's a factor, again, that's related to image content, not framerate.

A very good technical paper on crosstalk can be found here.

The causes of crosstalk are broken down by display type, covering CRTs, PDPs, DLPs, and LCDs (both time-sequential and polarized).

I'm not speculating. I stated facts. Again....

There are a number of factors that could cause or contribute to crosstalk, including the content (or how it's stored and played back), display, glasses, sync, etc.

Thanks. That's a good read.

There are two kinds of receptors in the eye: 1 bit Monochrome, or many bits of Color.

The concept of Colors is that there is a lot of light, so much in fact that the color receptors in the eye are activated.
The concept of black and white on the other hand is the light is low, so low the color receptors go blind.
So if the eyes don't see color because there's not enough light, it's got blind color receptors.

The concept of active polarized glasses is that we give the color receptor light, then we remove the light from the color receptor in the eye so it see's only 1 bit of color.

If we lit up one eye of the shutter glasses, then dimmed the other, maybe we could blind the color receptors better.
In other words, design the glasses so the color receptors see color better or worse.: When the one eyes receptors use color, get the other eyes receptors to see only black and white.
If we can replace color for monochrome, this acts like blinding the one eye, and this may close the crosstalk.

The solution is, make the visible color the glasses can see - either rgb or monochrome, as they open or shut.
Then the speed of the glasses isn't so important, or the only factor that gives positive results.

Light sends electrical impulses when it hits the eyes receptors.
We are using frames in a movie to send these electrical impulses to our brain: the more images per second, the more electrical impulses are hitting our eyes.

When the shutter is closing and the g2g is too slow so it's not totally black when the other eye lights up.
The shutter is showing black monochrome to the eyes receptor by closing, but since the shutters partly open, the eyes color receptor is still being hit.
And so the eye is seeing more than the black and white receptor, it is seeing the color receptor too, and sending the electricity hitting the color receptor to the brain; then both eyes send electricity from the color receptor to it, and this confuses the brain.

Frame sequential 3D - shows the brain electricity from one receptor that see's color, and electricity from one receptor which sees black and white.
If both eyes color receptors have electricity hitting it, the brain has no receptor showing it monochrome - and so it has to create one artificially.
When the brain gets tired of creating the illusion of monochrome electricity that enables it to separate the color receptors electrical signal, then the viewer feels queasy or ill.

As a result of seeing more color and less monochrome, it's easier to see crosstalk in the eye with a slightly open shutter when the other eye has a image shown to it. This is because the brain can't separate the two images.

In summary ,the solution is to change the color bit capability of the closing shutter to 1 bit, or monochrome, and make the eye that's open to see more bits of color (8 bits or more), then the brain distinguishes the receptors electricity easily and blinds the eye giving it a 1 bit monochrome electrical signal.

Audio of this post so it's easier to get through, in case reading it is difficult.

Topic title changed. Topics merged.

?