How much crosstalk (ghosting) has your 3D TV? Test it! - Page 3

Greetings Nanite.

Any idea how DLP TVs and projectors seem to be able to achieve perfect score?

Best regards.

Just for giggles... CRT monitor, anaglyph red and cyan, dubois/optimized

Pro-Ana glasses

Left eye open/right closed -> top half screen smaller visible number -> 4
Left eye open/right closed -> bottom half screen smaller visible number -> 10
Right eye open/left closed -> top half screen smaller visible number -> 10
Right eye open/left closed -> bottom half screen smaller visible number -> 1

Pro-View glasses

Left eye open/right closed -> top half screen smaller visible number -> 4
Left eye open/right closed -> bottom half screen smaller visible number -> 10
Right eye open/left closed -> top half screen smaller visible number -> 9
Right eye open/left closed -> bottom half screen smaller visible number -> 1

I have a PN58C7000 that i just received today with the SSG-2100AB/ZD glasses. The TV came running the 1032 firmware.

The results I get for this test depend entirely on what vertical rotation the glasses are in. Only the horizontal line across the center is capable of decently obscuring the opposite eye image. It's very easy to verify by simply tilting your head up and down to look at the various numbers with either eye closed.

In my opinion, these glasses are completely defective, or the sync signal is so aggressively timed that the glasses have insufficient time to darken the lens properly.

Can anyone please verify? Would it be worth trying the 1016 firmware as some others have reported?

Having used the Mitsubishi starter kit glasses (they are the SSG's just with different branding) on my DLP, I can confirm that yes, they do lose their effectiveness at certain angles while other glasses like the UltraClears have a slightly large effective hot spot. The Samsung/Mitsus can block the 80 to pitch black iat the right angle, but I can't get the glasses to block 1-80 simultaneously because the coverage isn't big enough around. The UltraClears juuust barely do cover all 1-80 at the right angle.

In real world content, I saw crosstalk only once or twice watching The Mad Magician with the Samsung/Mitsu glasses, and any time I do notice ghosting extensively it's while gaming (which is very rough on 3DTV's) I switch to the UltraClears and they show absolutely zero ghosting.

Of course, DLP is sort of the holy grail of crosstalk performance; if you're seeing more crosstalk more often than I am, blame the TV, not the glasses. The best advice I can give is that those TV's have been reported to do better after they've warmed up for like an hour.

Thank you for the information. This kind of performance is simply appalling. I agree that gaming is harder on the TV, but harder does not make such horrifying 3D performance acceptable.

I have had the nVidia 3d vision kit for a long time and spent a lot of time profiling the performance characteristics. The Samsung 2100AB's seem to retain a lot of brightness ... more than is possible. Now I know why.

The hotspot is very apparent on the glasses if you run this test. You really have to point your glasses at exactly the right spot on the TV in order to get anything resembling effective crosstalk cancellation. Meanwhile the rest of the TV is showing ALL of the numbers, and in your other eye as well.

Given what I'm seeing here and knowing the performance characteristics of plasma (which should be *better* than LCD), it makes me suspect that they did so at the expense of severe crosstalk.

I'm going to try the ultraclears (which seem to be nearly identical to the 3d vision glasses, so that may be promising if the shutter time is superior to the 2100AB's) and see if they help. Right now this kind of 3D performance is an utter sham being perpetrated on us.

It's not a display defect per se. What I'm referring to is rotating my head up and down to look near the top or bottom of the screen. Look at one part, the other part goes bad. And vice versa. This is due to the hotspot in the glasses.

rkuo:

I'm not seeing "horrifying" at all like you are. In my experience (with a DLP RPTV, mind you), crosstalk from the SSG's to the UltraClears is like going from seeing crosstalk .1% of the time, to seeing it .001% of the time.

What's the highest number the glasses cancel out within the hotspot on your TV? If the TV is... innocent... you should be able to block the 80.

cakefoo:

I'm on a plasma, so my results are going to be entirely different from yours. DLP generally has excellent switching times from my understanding, so your lenses may have a lot of time to kick in, relatively speaking.

Using the optimal hotspot of the glasses, I would say the place where the numbers tend to get a little too visible for my tastes is around the 52-60 range. 70-80 is flat out not good. Even in the better cases there is still a dark red crosstalk image of the number which probably should be mentioned because I really could read that image if I wanted to.

Keep in mind that the hotspot is so tiny that simply looking at the center of the screen is enough to reveal the numbers down to 30 or so.

Well, it turns out that Sears stocks Xpand 103 universals in-store, so I was able to go in and purchase a pair to try out on the spot.

Suffice it to say that the vast majority of my crosstalk problems are now gone.
Games are very playable. Crosstalk cancellation is largely independent of viewing angle.

I'm not sure if I got a bad run of my Samsung glasses or if all Samsung's 2100AB glasses are defective by design, but I recommend to anyone and everyone to consider the very real possibility that if you are using the Samsung 3D glasses you may be using inherently broken hardware.

EDIT: I'm going to say a lot of the crosstalk problems are now gone but there's still a fair amount more than I would like. Games are playable tho.

3D TV model: Samsung PN59D6500 59" Plasma
3D glasses model: Samsung 3D Active Glasses 2011 (bluetooth?)

Left eye open/right closed -> top half screen smaller visible number -> 1
Left eye open/right closed -> bottom half screen smaller visible number -> 10
Right eye open/left closed -> top half screen smaller visible number -> 1
Right eye open/left closed -> bottom half screen smaller visible number -> 10

LG 47lw5600 passive 1-38, 1-38 and the 38 is very dim.

Ouh, that is good ! Very good ! They made tremendous progress in FPR technology : it seems almost too good to be true.

There are hints of a few lower numbers but you can't make them out.

Hey, no cheating

When watching, there is no crosstalk on this set. In fact the 3D looks sharper than the 2D. Not just on this set but compared to all the other 2D HDTVs in the house.

My tv is 8605 Philips

Hello I have made a test with almoust 50 balanced settings , in severall stuff , I took off the perfect contrast and I saw made a significant difference turnng on light sensor ....
then I set the image setting to standard , and the following are the results

Warm screen :
Crosstalk test movie:
Left eye open/right closed -> top half screen smaller visible number -> 27
Left eye open/right closed -> bottom half screen smaller visible number -> 60
Right eye open/left closed -> top half screen smaller visible number -> 60
Right eye open/left closed -> bottom half screen smaller visible number -> 27

now I must say I am slight miopy on left eye anyway , this is for a fully readable number , but I still saw slight number image gray slight brighter than the dark and this I classify as barely seable , I cant read the number but I can write after I open also the other eye and the 60 become a 47 ...

so does my tv has a problem?

The weird thing is that after all this fixing when I started watching ocean wonderland 3d the image was awfull and I jus thad to switch to normal and scrap all the fixing I did .... anyway ghosting was present often ....

With the lowest number visible being 27, that tells me you had the picture settings way too dark.

Indeed I guess I tuned it badly , anyway I have launched Ocean wonderland and no way what I try in some scenes some of the fishes always appear with a double shade and some other give a little of headache or eye confusion/stress ....

I am wondering if the problem is more with my tv or specifically with this 3d documentary , I just bought it couse was cheap and wanted to test the 3d but I amstarting to think eventually to get another movie ( more serious one ) to test it and see if it is a problem of tv or the movie itself ....


anyonelse tried that documentary?

Well I bought another 3d movie called Oceanworld and I could see it on a samsung last series and looks great , I then checked wwith my philips and I must say looks Awfull , tons of double images ...

Here's part of a review that confirms crosstalk with the TV with all kinds of content.


http://www.trustedreviews.com/Philip...eatures_Page-2

So it's just philips that suck at 3d or it can be an occurring case on the sold ones that can be fixed in future eventually?
I love the philips tv is great and I would just like it to work fine with 3d as well as other tvs , I have seen the samsung and the same movie and on mine looked ghosted on samsung a slight effect not even noticeable if u dont really go looking for it ...
So since the Samsung I watched was the latest version of 55" , and was perfect 3d to watch with the same movie , also has a calibration system to fix the ghosting if not wrong , ok I coul dstill see some ghosting but this is so minimal and so light that you don't really notice unless you go look for it and I didn't had any headache feeling , with the philips instead I saw fishes doubled and not on the boundary , I mean really 2 fishes instead of one as sshould be , so the ghosting is really severe, I called a technician now and I hope it can be fixed at least to a samsung level ghosting , because right now it's totally unuseable ...

Since I like the Phjilips tv and ambilight I was considering moving to the superior series hoping to fix my 3d issues , anyone knows how much improves passing from a 8605 to a 9705? also the upcoming 9706 is better in 3d than the 9705?

and finally how does that performs compared to the new flagships of samsung in 3d ? I alwys consider led tv , plasma not an option ....

the 8605 has awefull 3d totally unwatchable without getting sick after few secs and need to vomit , at least on my tv tha tI tought was broken , in the assistance told me the serie 9000 is way better and doesnt cause any sickness and is comparable if not better than samsungs and other same lcd led tvs....

what you say about that?

Googled Philips 9705 Crosstalk. It also has huge issues.

Bill, thanks for posting. I've got the 55" version of the same set, and was wondering how it would measure on this test. The FPR film, if aligned properly, should be basically ghost-free, to the limits of its extinction ratio (I guess it is actually the combined extinction ratio of the film and the glasses?)

In practice, you get ghosting with these sets if you are too much above or below the perpendicular, as this creates a partial misalignment of the FPR film to the display lines. Did you try the test with some non-trivial vertical angles?

I am actually considering the 9706 upcoming in agoust , it should have improved ative 3d , and superresolution , the panels of the 9705 are definetly better than the 8605 as that was told in the assistence of philips , and severall reviews around , if I am not wrong the crasstalk in philips 9705 should be comparable to the one of the other brands of leds and stuff....

woudl be nice to have tough a decent list of performances of all the lcd/led tvs on 3d ....

I have been also told that there is a big difference in the results in 3d among a 40 " and 46 " , I was told actually that the more it costs the better elements are used for the construction and is not just a matter of different size ... but of quality of the components ...

comments on that?

Don't need to run the test, too high or to low causes bad crosstalk.

Sony NX711:

1, 1, 1, 1

The middle section of the screen however has much, much less crosstalk. Since there is so much variation, i wish we had a test that gave a bit more information about the whole screen, but was still numerically based.