Test Your ghosting

[Freke1]

Here is an image for testing the amount of ghosting on Your system:
res 1024*768:
res 1680*1050:
res 800*600:


I have a CRT and eDim shutterglasses and get:
30% "white" ghosting
15% "black" ghosting
(don't know what else to call it)

Edit: After I calibrated my gamma and brightness I get
25% "white" ghosting
2% "black" ghosting

[crim3]

I don't have ghosting by design, but on the other side, I have the problem of having the same colors and brigtness in both eyes, and your test image comes very handy for that!

EDIT: I forgot to mention that I use a HMD

[LukePC1]

I got a different monitor lately, which ghosts more I think.
I don't know how to measure them, but I'd say I get
'black ghosting' about 10%
'white ghosting' about 30-40%

It might be more, or less
Both together let the picture appear even worse

[Jahun]

Hey Freke1, nice!

Will check this on polarized rig soon... great addition. I am very interested to see what comes up...

Using cross eyed viewing, I get 0 ghosting!!

[chilledsanity]

Hey maybe I'm doing something wrong, but for the bottom image, I get far more ghosting on the inside of the black square than I do on the outside. On the outside it's barely noticeable, on the inside it affects more than half of the black square.

[LukePC1]

Right. If you don't use crosseye, so mainly, if you use Shutters and CRT's you get both parts of ghosting. One at the sides of the square, and one (maybe stronger) at the 'center' or 'inside' of the square.
It kind of makes a ghost image above itself.
At least that is what I experienced on CRT's, too.

[Freke1]

Yes - the ghosting outside one square is similar to the ghosting inside the other square and vice versa.
Just estimate the ghosting outside the squares using the scales and post the result.

I'm really interested in seeing what the single/dual projector setups and the iZ3D monitor get...
I assume the Head Mounted Displays have zero ghosting?

[pixel67]

Really cool Freke1! I wish this were around years ago!

On my Sammy I get 0% ghosting.

PiXeL

[LukePC1]

Wow! I think, that'll be hard to beat.
But its logical: digital light projection (DLP) doesn't have phosphor to shine longer and the shutterglasses block about 99% of the 'wrong' light.

Please post someone with polarization system

[pixel67]

Wow! I think, that'll be hard to beat.

Its a trade off. No ghosting on supported titles, just need more supported titles!

PiXeL

[Likay]

Testing my polarized rig: Hitachi CP-X260 with max lumen 2500 but now running with 2000 lumens. Silverscreen from silverfabric and filters also from silverfabric (SPARS).


edit: had gamma at 1.31 and contrast at 95% at my first measure wich gave wrong measurevalue (10% white and no visuable black ghosting).
White ghosting: 15% with nvidia driver at factory settings.
Black ghosting: Still nothing with nvidia driver at factory settings.

Tested with the IZ3D and sure, it does ghost a bit. However the ghosting is not of a "greyscale" kind so it's really difficult to say how much it is.
But at an estimation it is about 10-30% white as well as 10-30% black ghosting. But it's more like a bluish tint really.

Curious of Jahun. Except for the beamers you have almost spot on same things as i have by now. Waiting for your datas as well!


cheers

[LukePC1]

What 10-30% on iZ3D? That's much, even compared to 'old' CRT's. Why buy a superb Rig, if I can have everything of that on my CRT's (except the drivers and a little lower Resolution). I save a lot of money - lower Graphicscard for less resolution and postprocessing, and the expencive Monitor.

I thought it'd be maybe 5-15 or maybe 20% ghosting.
On top of that you get it ALL THE TIME? On CRT's the ghosting is mainly at high contrast, which isn't realy prominent in normal gameplay.

IZ3D is built with polarizers, like your projectors. So if it is a little optimized the ghosting should be equal to Dual projectors - maybe less, because the screen might depolarize a little .

Question: How much do the colors ghost? On CRT's they ghost approximatly like the black-White ghosting.
On the white background about 5% - nearly invisible
On the black background about 30-40% clearly visible, except Red and Blue which have less ghosting.

[Likay]

The polarisation of the iz3d is not locked in any way to 45/135?. It's more a regulated circular polarisation. I think it's possible though to lock polarisation on the front panel in a checkerboard pattern. But then you'll only have half the resolution (either x or y) and you also loose quite a bit of light so i don't know if this could be an option.

And again: I can't say it is 10% or 30% or something in between because instead of ghosting i get a "colorghosting" kind. In my case a bluish tint on one eye and a bluish/purple tint on the other eye. So i simple can't compare it with the greyscale ghosting scheme. I'm going to try to take some screenshots with the camera so you get a hint though.

cheers

[Neil]

I don't think it's circular polarization, but the ghosting test has little relevance to iZ3D's mechanism of working.

I'm not a physicist, so I don't have the knowledge to explain how the light waves work the way they do, but I would compare it this way. You know how you can get a really good benchmark score in 3DMark or other benchmark utility, but the REAL score is how fast your games play? I think the same scenario happens here. The test in itself gives mediocre results, but the actual games look much. much better. I don't know why, they just do. I'm not talking about drivers, I'm talking about the visual quality. The part of the test that holds some relevance for me are the color patterns at the top and the word "ghosting" in red - very little ghosting for me.

Another factor to consider is your dominant eye isn't what is seeing the ghosting, the same way your brain only sees the dominant eye when it comes to aiming at the correct cross-hair (with the iZ3D driver).

Regards,
Neil

[Likay]

Of course i should have added it looks way better when playing games!
Comparing with the ghostrejection i had with e-d shutters and a philips crt-monitor it's 10-0 to the iz3d!
However i do have some ghosting in some places where it's really recognizable but most often the image and the ghosting rejection is of a very good quality. I have to add that i also am not a physicist but i believe that the front panel is the one "regulating" the polarisation between the eyes (like an adjustable retarder.. lol, therefore i mentioned cicular polarisation ). But please tell me if i'm wrong because i'm really curious about the precise function and the physic behind it. It's a brilliant invention though.

cheers

[Jahun]

Not entirely true Neil, when you mention the dominant eye thing and aiming, it refers to your brains being able to filter the image your dominant eye receives and use that for the aiming and sort of suppressing the image coming from the other eye.

With ghosting this is not possible, your dominant eye will see a part of what should have been just for the other eye, something that is not present in real life and can be annoying. Sure one can get sort of used to it.. but it is totally another thing imho.


I am not sure why the IZ3D has color ghosting though. If I think about it, it should be possible to get it close to 0.. since you can rotate the subpixels of a pixel individually... hmmm.. Would be nice if a IZ3D engineer can explain it.. I'd be very interested to know what is keeping them from 0% ghosting.

(not saying it is bad or so, just wondering where it is coming from. Hey I'm a physicist, we are supposed to question everything )

[Jahun]

Would that be possible Neil? A small interview with a IZ3D rep? Perhaps not front page since it is a bit technical....

[Neil]

Here is an iZ3D thread on the issue written by iZ3D.

http://iz3d.com/forum/viewtopic.php?t=6 ... t=ghosting

I think it will be helpful.

Regards,
Neil

[CraziFuzzy]

FWIW, I have the Samsung HL-T6189S, and I am showing zero ghosting. I have tested in all lighting conditions, and am very impressed.

[Freke1]

When I take a picture through the right lens with my 3.2Mb cameraphone I get this:

(CRT and shutterglasses)
There's a little more ghosting from the colored squares than the picture shows.

Here it is crosseyed 3D:

[Likay]

I took a picture on my polarized rig with my homemade stereocamera with polarizers attached to the cameras. I can do the same with the iz3d but need to fineadjust the driver for the monitor again due to "too much tinkering with drivers".



Format is side-by-side right image first so it can be viewed crosseyed. As far as i know you can't get any additional ghosting using that viewing method.

ps: The light from the projectors aren't as strong as it seems since the polarizer kills more of the lamp in the ceiling than with the already polarized light from the beamers. The light from this rig is more than enough to play in full daylight though.

edit rehosted by mtbs3d

cheers

[Freke1]

Great, less ghosting than my CRT.

[LukePC1]

Yeah that's preatty good, but I think a DLP rig would have even less ghosting - but a more flickering.
I've got to say that the ghosting isn't noticeable in crosseyed mode, only if I look closely on one picture and try to find ghosting

I'd like to see a pic of the IZ3D next - just to see, how it does in comparision 8)

[CraziFuzzy]

Yeah that's preatty good, but I think a DLP rig would have even less ghosting - but a more flickering.
I've got to say that the ghosting isn't noticeable in crosseyed mode, only if I look closely on one picture and try to find ghosting

I'd like to see a pic of the IZ3D next - just to see, how it does in comparision 8)

The DLP rig switches at 120Hz, so virtually no flickering either.

[LukePC1]

Sorry I had to make myself clearer. The DLP projector I tested had 'only' 85hz, so there was minor flickering. The new DLP TV's are 120hz, where no flickering should be noticeable

It makes no sense to test the ghosting of a DLP, because it has none. Other setups (e.g. Polarized projection) seem to be pretty good, too.
But shutterglasses are great, because of their low price. They work great with old, cheap CRT's

[Likay]

Agrees. The biggest advantage with dual passive projection is that you have a tremendous amount of light to each eye compared to other solutions. Another advantage is that several people can watch at the same time. If the cost is motivated?.. Well, I won't change now.
Can't advice anybody to get such a rig until the "driversituation" changes a bit.
By the way. Have taken screenshots of the iz3d and ghosttest. Tried to tinker a bit but for some reason it looks quite a bit worse with the camera than in real life. It gives a hint on how the ghosting and color difference behaves on the iz3d though. I'll try to add screenshots from games later on. It's difficult to take pictures of the screen when motives are moving so i have to take screenshots, then play the shots in sview, photographing the screen from there and finally make the stereopictures. To tired right now but maybe tomorrow (later this day).



edit: host changed to mtbs3d

Cheers

[Freke1]

I think all solutions show less ghosting in games because the colors are usually shades of gray, green or blue. Rarely is there pure colors and big contrasts.
I know there's no ghosting on some solutions, but it could still be interesting to see a ghost test pic, because You can compare the contrast, colors tinting if any, or just get an idea of what it looks like.

[wbloos]

Hello to all, helle Freke1

great idea to post a testpicture to deremine ghosting - unfortunately Freke1's test picture has some disadvantages:
- there is no possibility to calibrate your monitor
- it's hard to compare the ghosting result with the scale since it is placed away from the test sqare
- IMO the ghosting scale is not correct - it shoulg go from 0% ghosting to 100% ghosting - where 0 = no ghosting at all - 100% is full ghosting (same as you would not wear any glasses)

I made a new ghosting test picture where these disadvantages are not present
- First load the picture and calibrate your monitor so all grey scales are visible (change brightness and contrast if necessary)
- then loock at the test picture viewing it once with your left and then with your right eye - you will detct a white ghosting (white seems to be dominant at ghosting) as well as maybe also a black ghosting - so you'll get 4 results
-- left eye white ghosting (LW)
-- left eye black ghosting (LB)
-- right eye white ghosting (RW)
-- right eye black ghosting (RB)

the lower part of the test picture shows the same test only black/white reversed - the results campared to the upper test should be identical

you may download my test picture from:
http://www.3d-hdv.net/forum/ghosting_te ... wbloos.jpg

here are the results for my preproduction Zalman 22" monitor (I'll post the results of the newer model later) using my testpicture
LW: 10%, LB: 0%, RW: 10%, RB: 0%

greetings
Werner

[Freke1]

Great job
here's my result with CRT and shutterglasses:



(camera shooting through right lens of shooterglasses )
have You not switched left and right though?
The darkish ghosting is 40% on the dark scale.
The whiteish ghosting is 10% on the white scale.

[wbloos]

have You not switched left and right though?

oh correct - my testpicture is side-ba-side parallel - so left is on the left side

greetings
Werner

[wbloos]

ok I changed the test picture a little bit: added a 'L' and 'R' on left and right view and saved it as P and X versions:

http://www.3d-hdv.net/forum/ghosting_te ... loos_P.jpg (parallel view version)

http://www.3d-hdv.net/forum/ghosting_te ... loos_X.jpg (cross viewversion)

greetings
Werner

PS: I deleted the old test picture on the server!

[LukePC1]

Really cool Freke1! I wish this were around years ago!

On my Sammy I get 0% ghosting.

PiXeL

Hi pixel... I wan't to challange you on this old statement.

1) What glasses did you use? The cristalis3?

Could you test some other glasses? I want to know, if these things are valid for today, too.
http://www.cmst.curtin.edu.au/publicat/2002-09.pdf

Maybe you could test the cheapest glasses you can find. They could be used for other users as second choice in emergencies

[relaxman]

i cannot download werners test picture
Can somebody send me?

Even the 3d-hdv.net doesnt work now..
I like to hear more ghosting test of different polarizer filters (linear, circular)
who has dual projector setup with silverscreen, please share the experiences.
what about cheap, plastic filters? is it worth the money?
(for eg. from http://www.polarization.com/polarshop/p ... ucts_id=34" onclick="window.open(this.href);return false;
or from ultimate3dheaven.com the circular one) anyone have this?

thx!

[yuriythebest]

i cannot download werners test picture
Can somebody send me?

Even the 3d-hdv.net doesnt work now..
I like to hear more ghosting test of different polarizer filters (linear, circular)
who has dual projector setup with silverscreen, please share the experiences.
what about cheap, plastic filters? is it worth the money?
(for eg. from http://www.polarization.com/polarshop/p ... ucts_id=34" onclick="window.open(this.href);return false;
or from ultimate3dheaven.com the circular one) anyone have this?

thx!

alternate link
http://img62.imageshack.us/img62/7166/g ... 1050wb.jpg" onclick="window.open(this.href);return false;

[relaxman]

thanks!

[anotherFrench]

great test, I did try it on my zalman trimon 22"
I'd say:
white square = 25%
black square = 0% (with very little ghosting inside the black square)
I also tryed with realD glasses but it's difficult to say as the ghosting parts turn to purple
so...
white square = 35/40%
black = 45/50% inside the square
if I turn the reald glasses some degrees, then it get the same results as the zalman glasses

[Likay]

I put away my sparfilters for a while for testing how normal high quality polarizers from polarization.com performs with my lcd-projectors. I miss the high lumen of course but ghosting rejection so far is just as good as with the spars! Still 15% white, 0% black ghosting with Freke's image and 10% white, 0% black with WBloos image. I've been gaming for four hours with these filters and nothing have happened yet (burnout). I get back in the dual passive projection thread wether or not these filters are suitable or not. 8)

edit: Fixed mixing of colored ghosts.

[Rosomack]

alternate link
http://img62.imageshack.us/img62/7166/g ... 1050wb.jpg" onclick="window.open(this.href);return false;

That clearly shows the main issue with iz3d ghosting is color tinting. It's much worse for the left eye and is very distracting when coupled with ghosting. Waiting for new glasses before retesting .

[BlackShark]

Hey i just discovered Werner's black&white ghosting test.
It's excellent : much easier to use than the previous ghosting test I used.

On my Zalman 22" i get 40% white ghosting, and almost 0% black but there is a very very slight darkening the scale isn't fine enough to measure so i'd say aprox 1% black ghosting.

[relaxman]

"On my Zalman 22" i get 40% white ghosting, "

How do you test it on Zalman?
Load the picture to stereo photo maker and hit F4 full screen?
is it side by side in parallel format?