Different LCD shutter glasses - And ghosting

Post Reply
flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Different LCD shutter glasses - And ghosting

Post by flexy »

Today i got the old-style ELSA revelators cheap off ebay with IR transmitter, and i hooked them up to my 3D ready Plasma and compared them to the ED glasses.

The IR emitter and frequency is basically the same - so from that point of view i got a VERY good deal on a set of wireless shutters + transmitter with 3 pin VESA plug.

But the Elsa revelator ghost A LOT more than my ED wireless - its also that the revelators seem to be "brighter".

I dont know whether the ghosting issue is just due the fact that those give a brighter picture (thus more ghosting seen on screen)...or if its an issue of slower LCDs with those glasses.

I would be very interested in a comparing list of currently available shutters - especially in regards how they ghost.

() Another idea i had.....given a typical system with a 3 Pin VESA sync output...be it wireless IR transmitter or using cables:
If there would be a simple way to build a "signal delay" device...something which would allow input of the sync signal, and using a variable resistor to somehow "delay" the output. We're dealing with simple TTL level signals here.

My idea is that such a device MIGHT enable one to adjust possible signals not 100% in sync (TV/Projector ---> Glasses) and therefore could POSSIBLY be used for eliminating ghosting and adjusting the 3D quality.
If such a "device" would allow a maximum delay/shift of a signal in the (time)range of ONE FRAME this should cover any possible situation.

If a signal, in theory, arrives 20ms "too late" (respective shutters sync 20ms "too late" and therefore ghosting appears)....it would be possible to just adjust/delay with this circuit so the glasses/3D display on screen are 100% in sync.

This based on my assumption that ghosting is a sync or timing-issue.

It might be the case that such a device might be very simple to make with some parts from RS?
Any idea?

The other theory is that common shutters just dont really always 100% shut...therefore there is always remains of the view visible..therefore ghosting?

Tril
Certif-Eyed!
Posts: 655
Joined: Tue Jul 31, 2007 6:52 am
Location: Canada

Post by Tril »

On a CRT monitor with VGA input (it's analog), as far as I know the monitor does not have any memory so it just draws as it receives the signal.

I've seen such a circuit or something very similar to what you want posted before by someone on another forum. I don't remember where I've seen it but I'm sure others will remember. Unfortunately, most of the time ghosting does not come from bad sync like what you described.
CPU : Intel i7-7700K
RAM : 32 GB ram
Video card : GeForce GTX 980 Ti
OS : Windows 10
Display : Samsung UN40JU7500 Curved 40-Inch UHD TV with shutter glasses
HMD : Oculus Rift

Image

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Re: Different LCD shutter glasses - And ghosting

Post by cybereality »

flexy wrote:The other theory is that common shutters just dont really always 100% shut...therefore there is always remains of the view visible..therefore ghosting?
There might be something to that. I don't see why they can't make an LCD-type shutter that would go from completely clear to completely black. Even those old Nintendo games on a watch could do that in the 80's, come on now. This stuff has to be dirt cheap by now.

Tril
Certif-Eyed!
Posts: 655
Joined: Tue Jul 31, 2007 6:52 am
Location: Canada

Post by Tril »

They use liquid crystal, like in LCD monitors. I have no idea what type they use. The speed of the LCD monitor has improved over the years. It might be possible now to make better shutter glasses with the technology developed for LCD monitors.

An LCD monitor has a backlight that's open all the time. If you show black on the screen, you don't see the backlight anymore because the light is stopped well. I don't think shutter glasses work that well.

It's not good for the glasses to apply a DC voltage for long but I did try it in the past. I manually shut one side for a few seconds with a battery and looked through it. From what I remember, it was not shutting the brightest light at 100%.

There's a test I did in the past. I made a jps picture with three colored bars (red, green, blue) in one eye and a black screen in the other eye. I then watched it with a jps viewer and closed the eye with the colors. If the screen does not ghost at all and if the glasses completely shut the light, the eye that is open will see a black screen. If the screen ghosts a bit or if the glasses don't completely shut off the light, you see some dark color bars. When I did that, I could see the three colored bars. They were much darker than in the other eye but they were clearly visible.
CPU : Intel i7-7700K
RAM : 32 GB ram
Video card : GeForce GTX 980 Ti
OS : Windows 10
Display : Samsung UN40JU7500 Curved 40-Inch UHD TV with shutter glasses
HMD : Oculus Rift

Image

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Ghosting with 3D shutterglasses

Post by budda »

I have noticed ghosting too and wish it wasnt there.

I play Lock On Modern Air Combat, and the 3D is fantastic for spatial awareness and gauging the terrain profile. I notice in the cockpit with very large contrast between dark interior and light exterior, that my instruments exhibit strong ghosting. Otherwise the ghosting is not really noticeable.

I am using ED glasses and did a simple test to examine the cause of the ghosting.

Using a pair of polarising sunglasses, I found the extinction to be very high in the "crossed position". This means the polarisers are very good quality in the ED shutterglasses. However it seems the shutter LCD itself, does not properly shut when active. It has an intrinsic leakage which causes the ghosting effect. The only solution is better LCD shutterglasses technology.

I am very tempted to modify some LCD shutters used in automatic welding helmets to acheive better viewing angle, shutter speeds and extinction factors than regular glasses can acheive.

Anyway, thats my pipe dream. :)

flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Post by flexy »

well i almost have no ghosting if i wear ED glasses AND revelators together :) Although its evry uncomfortbale and looks retarded :)

User avatar
Freke1
Certif-Eyable!
Posts: 1060
Joined: Wed Feb 28, 2007 6:40 pm
Location: Wake Island

Post by Freke1 »

I think most of the ghosting comes from the afterglow in the monitor. If You move the mouse arrow fast across a black background You can see a trail/afterglow. The eDimensional glasses shuts off well, but the CRT monitors doesn't. Samsung DLP's are much better I've heard - with no afterglow so much less ghosting, right?

flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Post by flexy »

well i have a plasma. I can tell you more if i'd see a DLP in comparison.

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Post by cybereality »

Freke1 wrote:I think most of the ghosting comes from the afterglow in the monitor.
I did have one idea of fixing this, but its not exactly DIY. If the line-blanking was done with an external screen overlay it should solve the issue of after-glow. So lets say the monitor always displays both images in an interlaced format. Then a separate monochrome-LCD overlay blackens every other line and swaps the signal in sync with the glasses. Assuming the LCD shutter device could turn completely opaque, I believe this would solve the issue of ghosting. Obviously this is not a practical idea, but maybe it will give someone else a spark as to how to fix this.

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Ghosting with Shutterglasses

Post by budda »

I performed some more tests with my CRT monitor, a Dell P1130, and the edimensional shutterglasses.

The apparent cause of the ghosting effect could be caused by the phosphor decay time or by the shutterglasses themselves.

[1]
I ran a standard openGL test program called RedBlue. It uses a stereo overlapped pair of rectangles, red and blue. If the openGL stereo works then a magenta overlap area will be seen. By viewing the images through the shutterglasses, I found that the red rectangle was completely extinguished in one eye whilst with the other eye the blue rectangle appeared as a ghost image and was not completely extinguished.

Therefore I assume that ghosting is colour dependent. Red is best, blue is worst.

[2]
I positioned an external polarising sheet up to the shutterglasses, and found the extinction to be excellent. The polarisers in the shutterglasses are extremely good.

[3]
I ran the LockOn Modern Air Combat program, and looked closely to see where the ghosting is most apparent. Firstly, there seems to be very little, if any ghosting in the red (warning and illumination lights) and the green lights (head up display). This in itself is remarkable. The ghosting appeared in the bluer colours and bright white high constrast areas. (I cannot understand why white shows up clearly considering it contains red and green colour components).

I lowered the colour temperature of the monitor to the lowest available, 5000K and this helped reduce ghosting overall.

UPDATE [4]
I modified the RedBlue openGL stereo test program to a RedGreen version, with a green colored rectangle replacing the blue version. The RGB values were pure : RED (127,0,0) , GREEN(0,127,0) and BLUE (0,0,127).

Using the modified program, I can isolate the ghosting caused by the green phosphors. There is significant green ghosting. The green ghosting appears slightly stronger than the blue ghosting, although similar in light intensity. I estimate the ghosting level of the green and blue phosphors to be 10% of their respective original light intensity. The red ghosting is non-existant.

I have taken pictures of my tests and plan to post them later in this thread.

UPDATE [5]
In general the stereo images appears to flicker in each eye, however the ghosting image does not appear to flicker.

UPDATED CONCLUSION
It seems to me the "standard P22" CRT green and blue phosphors show a noticeable and unacceptable level of ghosting with shutterglasses.

Despite shutterglasses appearing and feeling cheap, they perform very well. However the viewing screen technology lets their effectiveness down. Until this deficiency is fixed, then stereovision with shutterglasses will always be a second rate technology, with first rate potential.

IDEAS FOR THE FUTURE
Someone should develop a stereo test application which allow the user to visualise and estimate the ghosting levels for each red, green and blue phosphor of their screen.

A calibration intensity scale using red, green and blue greyscales could be aligned against the ghosting to gauge the ghosting levels (much like ordinary monitor calibration software methods).

Once manufacturers realise that users can rate screen performance objectively, then they may be pressured to produce better products for 3D shutterglasses.

MEMO TO MONITOR MANUFACTURERS.
Make your green and blue phosphors work as well as your existing red phosphors.


Thanks for your interest.
Last edited by budda on Wed Jun 11, 2008 8:43 am, edited 2 times in total.

EZVZ
One Eyed Hopeful
Posts: 1
Joined: Fri May 30, 2008 12:16 pm

Post by EZVZ »

Freke1 wrote: Samsung DLP's are much better I've heard - with no afterglow so much less ghosting, right?
I use shutter glasses with Mits DLP and have not noticed any ghosting at all when viewing all sorts of material.

User avatar
LukePC1
Golden Eyed Wiseman! (or woman!)
Posts: 1387
Joined: Wed May 16, 2007 11:30 am
Location: Europe
Contact:

Re: Different LCD shutter glasses - And ghosting

Post by LukePC1 »

flexy wrote: () Another idea i had.....given a typical system with a 3 Pin VESA sync output...be it wireless IR transmitter or using cables:
If there would be a simple way to build a "signal delay" device...something which would allow input of the sync signal, and using a variable resistor to somehow "delay" the output. We're dealing with simple TTL level signals here.

My idea is that such a device MIGHT enable one to adjust possible signals not 100% in sync (TV/Projector ---> Glasses) and therefore could POSSIBLY be used for eliminating ghosting and adjusting the 3D quality.
If such a "device" would allow a maximum delay/shift of a signal in the (time)range of ONE FRAME this should cover any possible situation.

If a signal, in theory, arrives 20ms "too late" (respective shutters sync 20ms "too late" and therefore ghosting appears)....it would be possible to just adjust/delay with this circuit so the glasses/3D display on screen are 100% in sync.

This based on my assumption that ghosting is a sync or timing-issue.

It might be the case that such a device might be very simple to make with some parts from RS?
Any idea?

The other theory is that common shutters just dont really always 100% shut...therefore there is always remains of the view visible..therefore ghosting?
@ signal delay: I don't think it will help much. Example:
You use 2 computers and play things in 3D at same frequency.
You have cable conection, so the signal is slightly shifted...
You look at the 'wrong' monitor. What do you see?

With CRT you see two pieces of the picture. A little like 'splitscreen'. The upper half is 3D, and the lower half is not 3D (because it was actually reversed 3D...)

This example shows, what a delay would do. You would have a little bar on the picture, which has no 3D, because it is 'reversed'. I don't think you can get much better picture with delay - at leat not on CRT...

@ shutting:
I put two glasses over each other, so the block out the light. It works as long, as you don't look into a lamp. If you look into the lamp the light becomes a little more blue/violet, so the theorie with blue having more ghosting might be true...
EZVZ wrote:
Freke1 wrote: Samsung DLP's are much better I've heard - with no afterglow so much less ghosting, right?
I use shutter glasses with Mits DLP and have not noticed any ghosting at all when viewing all sorts of material.
I noticed this on DLP, too. DLP screens 'trow' light on the screen, so there is nothing, that can glow longer than the picture takes - in contrast to CRT and aparently Plasma as well.

The Shutters shut away about 99% of the 'wrong' light, so compared to the other light this 1% is like nothing. They eye should not notice it at all...
Play Nations at WAR with this code to get 5.000$ as a Starterbonus:
ayqz1u0s
http://mtbs3d.com/naw/" onclick="window.open(this.href);return false;

AMD x2 4200+ 2gb Dualchannel
GF 7900gs for old CRT with Elsa Revelator SG's
currently 94.24 Forceware and 94.24 Stereo with XP sp2!

flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Post by flexy »

Well today i did some more extensive testing, using various modi on the Plasma (checkerboard, interleaved etc..)

I have to say it was a waste of money for "3D" since the ghosting is really unacceptable.

I saw some stereo slide shows - some pictures (depending on colors and similar) look great. But gaming and watching movies there is just too much ghosting, especially with high-contrast details of the picture, and it totally spoils the fun.
This is NOT something i want to play games for hours with.
It distracts and gives headaches if you all brighter things three times.

DLP was no option for me since i think we didnt have 3D-ready DLPs in Germany...so the Plasma was the first 3D ready screen here...and the ghosting is pretty bad.

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Post by cybereality »

flexy wrote:I have to say it was a waste of money for "3D" since the ghosting is really unacceptable.
I'm sorry to hear that. Is it really that much worse than the ghosting on CRTs?

flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Post by flexy »

i really never did a lot with CRTs..i still try to get real page-flip working on the plasma. According to a whitepaper the plasmas should defintly do pageflip. Then i could even adjust ghosting when i set the frequency lower, maybe. But the standard 3D is not that great, really.

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Post by cybereality »

flexy wrote:But the standard 3D is not that great, really.
Man, how long do we have to wait for a perfect solution? Maybe the 120Hz Samsung LCDs will be better...

flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Post by flexy »

in regards to DLPs....its unlikely that they dont have ANY coating...so i could see there being a certain latency also.
But if its the case that DLPs *really* dont show any ghosting..than they would be the clear winner.

this is a very good white paper: http://www.cmst.curtin.edu.au/publicat/2005-12.pdf
describing various display technologies in regards to their compatibility to S-3D, this is even better, in regards to the subject:

http://www.cmst.curtin.edu.au/publicat/2002-09.pdf

A very extensive paper on ghosting and what causes it.

The irony is that accoring to the first paper plasma is doing really well, and right now i am trying several methods of getting my glasses to sync using pageflip.

The ghosting is certainly a result of LCS quality PLUS plasma/phosphor lag. As for my set...it might really depend on the viewer's expectations. To give a very rough "idea" i could say the 3D effect is 75%-80% "ok"...highly depending on source.

The remaining 25%-20% is ghosting which is apparent mostly with bright picture details, eg. if i play a game and i see a shape/person actually throwing two white shadows left/right its just not something i can accept...it also totally destroys the 3D effect because focus is lost easily.

I am testing with elsa revelator, VGA output, ED glasses/dongle..and i can eg. sync my PC display to the plasma...and i can get the sync either out the PC using a variety of drivers...or directly get the 3D sync signal from the TVs jack..but i THINK the 120hz sync i am getting out from the set is just too fast. I dont have flicker....but ghosting.
I need a way to balance this out...less ghosting and lower freq, therefore maybe eliminitating "phosphor lag".

User avatar
LukePC1
Golden Eyed Wiseman! (or woman!)
Posts: 1387
Joined: Wed May 16, 2007 11:30 am
Location: Europe
Contact:

Post by LukePC1 »

I don't have a solution, but a little more objective method to mesure your ghosting. Maybe it helps you to compare your ghosting, to those of others.

I think the Plasma should be somewhere between CRT and DLP, because it should be at least better than old CRT...

1) the first PDF shows, which displays could work with paralax barriers... you know this DIY thing :D

2) acording to second PDF both things seem to be equally important for ghosting... if it's the case DLP should ghost - dependent on glasses...
I tried a DLP projector @ 85hz for some hours and it was not bad at least MUCH better than any CRT I tried...
Assuming DLP has no after glow, ther should be some ghosting left because of glasses...
Play Nations at WAR with this code to get 5.000$ as a Starterbonus:
ayqz1u0s
http://mtbs3d.com/naw/" onclick="window.open(this.href);return false;

AMD x2 4200+ 2gb Dualchannel
GF 7900gs for old CRT with Elsa Revelator SG's
currently 94.24 Forceware and 94.24 Stereo with XP sp2!

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Post by cybereality »

Thanks Flexy, that first paper was a great read! Gives me hope that the parallax barrier thing will work. Also, according to that paper, the 32" Samsung hdtv I already own may support field-sequential 3D! It has the "Game Mode"!!! I never knew really what that did, but it seemed to look better so I always had it on. Its almost too good to believe, I hope it works.

Also, have you tried the AnotherEye2000 yet? They're only $10, its worth a shot. I'm pretty sure they're better than the ELSAs.

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

CRT GHOSTING WITH ED SHUTTERGLASSES

Post by budda »

I have downloaded the ghosting photos on the forum album pages.

These photos show the color dependent ghosting effects using Edimensional shutterglasses and a Dell/Sony P1130 CRT monitor.

Red and blue normal CRT image
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1254

Red and blue stereo CRT image setup using edimensional glasses
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1252

Blue ghosting on red
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1251

Red ghosting on blue
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1253


RED and GREEN stereo CRT image setup using edimensional glasses
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1255

GREEN ghosting on RED
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1257

RED ghosting on GREEN
http://www.mtbs3d.com/gallery/displayim ... ?pos=-1256


Thanks for your interest :shock:

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Post by cybereality »

Ok, I just got my AE2K in the mail today (that was real quick, feels like I ordered it yesterday). I just got to try them a bit, but already they seem superior to the ELSAs and E-D in terms of the actual LCDs. The glasses are more clear than the ELSAs, they only barely tint your vision. The LCDs also seem bigger and don't have that tunnel vision from the other types. However they are super uncomfortable, I'm not sure I'd want to wear them for too long. Overall well worth the $10 I spent. The manual buttons are nice, still have to test more.

In terms of ghosting they are very good. In all honestly there was very little to no ghosting on a standard CRT monitor. I was using a 21" CRT @ 85Hz and the image looked good. There was flickering, of course, which was a bit distracting. But otherwise I didn't notice the ghosting at all. In extreme cases or when I looked for it, there was certainly some ghosting. But honestly a lot less than I remember the ELSAs being. So I think just for ghosting it would be AE2000 > ED > ELSA. However my old pairs of shutterglasses are broken so I can't do a back to back comparison. I tested with Unreal2004 and some random graphics demos. I tried to pick daylight maps with a lot of contrasting colors, but I didn't notice much ghosting. Not enough to distract me while playing at least. I could close one eye, and it would look like a 2D image, minimal crosstalk. I tried some videos too, looked alright. Very minimal ghosting (5% max). For sure they are better than ELSA in terms of ghosting. There was good sense of depth on the Nvidia test demo, the logo was popping out of the screen and did not ghost against the black background. I clearly remember how bad the ghosting looked with the ELSAs (it was about 10-15% ghost image) with the nvidia test application. The AE2000 was less than 5%, hardly noticeable. Not that they are perfect (they are not very comfortable to say the least), but good in terms of minimal ghosting.

Here is something interesting. I tried them on a regular Samsung 32" 60Hz LCD just for kicks. I just tested a few video files in interlaced mode. Ok, the flicker was horrible. Really bad. However there was a 3d effect! The only problem was the sync. It would get un-synced all the time, and I'd have to keep switching the glasses off and on to get it right. However, when it did sync up (for maybe 15-20 seconds at a time) it did look good in stereo 3d. Although there was heavy flickering, the ghosting really wasn't as bad as I thought. I mean, there was still a lot of crosstalk (because of the sync issues). But for the few seconds at a time it would appear to have decent depth. However the flicker and the sync problems make it unusable. So I guess I got my hopes up for nothing, although I actually got more than I was expecting maybe.
Last edited by cybereality on Sun Jun 15, 2008 1:11 am, edited 2 times in total.

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Post by cybereality »

I just did some tests with the ELSA Revelators VS. AnotherEye2000. I was initially just going to test the ghosting but the results surprised me. In terms of ghosting, the AE2000 are better, but maybe not by as much as I thought. However the overall picture quality was much better in general. Better colors, better depth, more contrast, etc. Somehow the AE2K's looked more "HD" than on the Revelators, even though I had both glasses hooked up at the same time running the same res obviously. I think this is just because the ELSAs totally kill the color. I've taken a number of comparison shots that I think illustrates the difference. These were all taken on a 21" NEC CRT @ 85Hz. I didn't know if you could take photos with shutter glasses, but you can. However photos make the ELSAs look worse (they are worse, but the photos make it look really bad). I think it has something to do with how the LCDs flicker, because I tried many times and the results were the same. But maybe it will give you guys an idea.
Shutter-Glasses-ELSA-AE2K_01.jpg
Shutter-Glasses-ELSA-AE2K_02.jpg
Shutter-Glasses-ELSA-AE2K_03.jpg
Shutter-Glasses-ELSA-AE2K_04.jpg
Shutter-Glasses-ELSA-AE2K_05.jpg
Shutter-Glasses-ELSA-AE2K_06.jpg
Shutter-Glasses-ELSA-AE2K_07.jpg
Shutter-Glasses-ELSA-AE2K_08.jpg
Shutter-Glasses-ELSA-AE2K_09.jpg
Shutter-Glasses-ELSA-AE2K_10.jpg
You do not have the required permissions to view the files attached to this post.
Last edited by cybereality on Sat Jan 01, 2011 11:25 am, edited 1 time in total.

flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Post by flexy »

that certainly has become an interesting thread...

the AE2000 would be interesting for me since they feature a "sync doubling" mode...althoug i dont know how this is implemented.

Why would this be interesting?

As owner of a 3D ready plasma, according to whitepapers plasma is suitable for page-flip/time sequential S3D.

So..i could try to "circumvent" the built-in checkerboard 3D (which ghosts too much, at least using ED and Revelators) and feed the system some output from the PC at 60hz.

The set internally seems to double the frequency and outputs 120hz....and if i set "3d mode" ON, while on the VGA output i get the 60hz sync signal...but the glasses wont sync with the set! If i could enable "sync doubling" now...there is a chance i can sync the glasses to normal VGA input on the plasma...using the PC output.

This COULD make it possible to achieve a real-pageflip at 120hz....with the option to actually turn it down to 50hz (== 100hz) and therefore have more control over ghosting...assuming it wouldnt ghost as much using 100hz since each frame would be a bit longer. (The point here is to actually skip the "internal" fixed 120hz output from the "3D Ready" to achieve less ghosting than "out of the box")

I think i'll order those just to check them out and see what results they give in conjuntion with a 3D plasma...

User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11402
Joined: Sat Apr 12, 2008 8:18 pm

Post by cybereality »

flexy wrote:This COULD make it possible to achieve a real-pageflip at 120hz....with the option to actually turn it down to 50hz (== 100hz) and therefore have more control over ghosting...assuming it wouldnt ghost as much using 100hz since each frame would be a bit longer. (The point here is to actually skip the "internal" fixed 120hz output from the "3D Ready" to achieve less ghosting than "out of the box")
That could theoretically work, but you'd still need a driver that could output in the method necessary for sync-doubling (over/under). You can't use the native checkerboard format for that or the nvidia driver. The manual suggests using the now defunct Wicked3D driver with a 120Hz monitor. There is some info here I found: http://www.stereo3d.com/wicked3d.htm

Also, I am not sure about this, but wouldn't page-flipping cause more ghosting? When using a format like checkerboard or interlaced, each pixel is dedicated to a particular eye. So there should be minimal cross-talk since no individual pixel would ever share information for both eyes. With page-flipping you get a cleaner, brighter image, but I don't see how that could improve ghosting much. Because now each pixel has to alternate between frames one after another. Unless the response time is perfect, there will be some bleeding in of colors into the next frame. At least thats what I would assume.

User avatar
LukePC1
Golden Eyed Wiseman! (or woman!)
Posts: 1387
Joined: Wed May 16, 2007 11:30 am
Location: Europe
Contact:

Post by LukePC1 »

@ ghosting and sync doubling...

what if you use interlaced mode? Wouldn't it double the sync, too?

I have concearns with refresh rate... if the Plasma can't take 100hz input why should it take the 100hz from the sync doubler?
Only possibility is, that it's only a software problem :D
Play Nations at WAR with this code to get 5.000$ as a Starterbonus:
ayqz1u0s
http://mtbs3d.com/naw/" onclick="window.open(this.href);return false;

AMD x2 4200+ 2gb Dualchannel
GF 7900gs for old CRT with Elsa Revelator SG's
currently 94.24 Forceware and 94.24 Stereo with XP sp2!

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Avoiding CRT Ghosting

Post by budda »

I have just tested another 21 inch class CRT - the Diamond Pro 2070 - made by NEC/Mitusbishi.

It behaves the same as the Dell/Sony P1130 with respect to color ghosting characteristics.

There is no ghosting with red, but green and blue show mild and noticable ghosting artifacts.

I suspect these, and other top-line CRT monitors using P22 type phosphors will demonstrate the same levels of color ghosting.

HOW CAN WE ENJOY GHOST FREE STEREOVISION WITH CRT & SHUTTERGLASSES ? :idea:

The answer is to have a RED & BLACK stereovision mode similar to black and white, but with the greyscale converted to an equivalent red scale. I do not mean just turning the green and blue off. I mean converting color images to black and white images and then converting the white components to red.

This would be better than anaglyph in red and blue for instance. The advantage is that typical P22 type CRT monitors with ghost free reds can display stereovision in a sharp ghost-free format, similar to black and white movies but without the garish effects of anaglyph color pairs.

I have just run Lock On Modern Air Combat with red only. :shock: FOR THE FIRST TIME EVER, I HAVE SEEN SHARP, GHOST-FREE STEREOVISION :shock: . I could live with it much better if the contrast effects of green and blue shades (not colors) were present. This is a welcome improvement.

This viewing method is easy to code, and provides excellent effects with existing shutterglass and games setups. Game producers and graphics card manufacturers should offer complimentary drivers or native modes which support this method.

Try it yourself !!! :shock:
Last edited by budda on Thu Jun 19, 2008 12:35 am, edited 1 time in total.

flexy
Cross Eyed!
Posts: 167
Joined: Tue May 06, 2008 2:12 am

Post by flexy »

uhm...stereo should be an enhancement and not taking something away.

I certainly dont want to have b/w images on my screen "just" to enjoy stereo...then i rather prefer 2D but in color:)

(Am i right you are talking about converting color into b/w ?)

Your method is interesting, nevertheless. Combining some sort of anaglyph and shutters. If you think about, GHOSTING is usually always more apparent the brighter the object....and some darker/less prominent shades/contours dont exhibit any ghosting at all. (Talking about shutters/plasma as observed by myself). Now...we could use the bright parts of an image, we can also use all gray parts....and apply some anaglyph method reducing ghosting while the smoother/less ghosting shades (the one with less contrast) can stay whatever method they are...eg. checkerboard/interlaced/etc...some intelligent form of combining shutters and anaglyph.

Implementation would be relatively easy...since output images just would need to go through a threshhold filter (eg. everything from brightness 200+) ....and the anaglypgh itself can be VERY subtle...since we are actually NOT needing total separation but just an additional "slight" anaglyph which is supposed to get rid of the remaining shadows (on top of the interlace/time sequential)...eg. combined shutter glasses with light green/cyan...maybe only at 30% intensity...should preserve colors pretty well.
Just a thought...dont know if i have any inconsistencies in there...it's early in the morning ;)

User avatar
LukePC1
Golden Eyed Wiseman! (or woman!)
Posts: 1387
Joined: Wed May 16, 2007 11:30 am
Location: Europe
Contact:

Post by LukePC1 »

If I understand right, it should reduce ghosting on CRT's, if I reduce Green and Blue color ... strenght but increase the red intensity, right?

I don't think you have to get rid of ALL ghosting, but reducing it to a minium wouldn't be bad.
Play Nations at WAR with this code to get 5.000$ as a Starterbonus:
ayqz1u0s
http://mtbs3d.com/naw/" onclick="window.open(this.href);return false;

AMD x2 4200+ 2gb Dualchannel
GF 7900gs for old CRT with Elsa Revelator SG's
currently 94.24 Forceware and 94.24 Stereo with XP sp2!

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

CRT "RED EYE" METHOD FOR SHUTTERGLASSES

Post by budda »

Hi, I have decided to call this method, the "RED EYE METHOD".

I tested pure red against pure black, in order to establish whether ghosting can be completely eliminated. I have demonstrated this.

What I did is turn blue and green down to zero using the supplied monitor calibration utilities. You may need to do this in the nvidia control panel. I am new to this too, so your set-up situation is likely to be different to mine.

In order to improve the red dynamic range, I increased the contrast level to 100%, and changed the brightness level to about 50% so that black is truly black (like the night). Any ghosting then would be clearly seen.

Importantly, adjust the gamma level in the control panel or elsewhere so that contrast levels are improved at the darker end of the greyscale. This adjustment is best done by eye, even without a calibration chart.

The setup used is : XP - nvidia - 91.31 - 6800GT - Diamond Pro 2070 CRT - ED wireless shutterglasses

I ran the RED EYE METHOD at 1280 x 960 at 85Hz and 120 Hz.

Results:

[1] No ghosting at the 120Hz scan rate

[2] No flickering at the 120Hz scan rate

[3] The screen image is bright with excellent contrast

[4] There is very sharp detail. Mainly because there is no ghosting or misconvergence of colors, and the effect of stereo AA.

[5] The red stereo screen images are easy on the eye. The view is improved in a darkened room.

[6] Green and blue shades need to be somehow better represented in the red color scale. Gamma correction makes a big improvement to shading.

[7] Edimensional shutterglasses work extremely well.


Remarkably there appears to be no flicker or ghosting at 120Hz. I did see light flicker at 85Hz but its easy to overlook because the stereovision is so immersive and the red color is easy on the eye. I am very impressed with the stereo effect using such simple methods.

There is a small 1 inch band of light green discoloration and light ghosting at the bottom of the screen. This is not the image itself because I can move the image. This artifact is not particularly noticeable, but is something to do with my monitor.

The results described should be reproduceable because these CRT monitors generally use the P22 type phosphors. So your stereovision experience should be just as good as mine.

Thanks. :D

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Sample High Contrast Pictures

Post by budda »

As no one has to contributed to this thread lately, I have included a supplimentary posting about the Shutterglasses Red Eye Method.

Please note the pictures I have linked are examples of HIGH CONTRAST images only to demonstrate any ghosting effects.

The double image photos are viewed without the shutterglasses, whilst the the single image photos are viewed through the shutterglass.

The real life impression is better. The small detail is revealed with zooming.

The actual photographic quality is limited due to the 1 to 2 second exposure times and no image stabilisation. The long exposure time I believe actually brings out any apparent ghosting seen, so please bear that in mind.

The photos were taken of an F14 mod in Lock-On Modern Air Combat. The set-up is as described in the previous post.


Photo Set F14 - 1

Double Image: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1674

Single Image: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1673


Photo Set F14 - 2

Double Image: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1672

Single Image: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1671


Photo Set F14 - 3

Double Image: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1675

Single Image: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1676

:)

vicx
One Eyed Hopeful
Posts: 5
Joined: Sat Apr 12, 2008 6:28 am

Re: Sample High Contrast Pictures

Post by vicx »

budda wrote:As no one has to contributed to this thread lately, I have included a supplimentary posting about the Shutterglasses Red Eye Method.

Please note the pictures I have linked are examples of HIGH CONTRAST images only to demonstrate any ghosting effects.

The double image photos are viewed without the shutterglasses, whilst the the single image photos are viewed through the shutterglass.

The real life impression is better. The small detail is revealed with zooming.

:)
Love your work. I'm going to give this a go.

HD Nintendo Virtual Boy ftw.

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Shutterglasses Red Eye Method with greyscale capability

Post by budda »

Hi,

I have built and tested a "simple" hardware mod for generating red monochrome with full RGB greyscale capability.

Watch this space for details in the coming weeks - if not sooner.

:idea: :)

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Red Eye Method - RGB to Red Monochrome Hardware Mod

Post by budda »

I discovered a technical paper which details a simple hardware method for realising the full greyscale capability of the Red Eye Shutterglasses Method.



The techical paper reference is:

Pelli, D. G., & Zhang, L. (1991) Accurate control of contrast on microcomputer displays. Vision Research 31, 1337-1350.

It is available online at: http://www.psych.nyu.edu/pelli/pubs/pel ... ntrast.pdf



Pelli and Zhang, describe a simple electrical circuit for converting three channel RGB video from a typical personal computer into a single monochrome video channel suitable for output to a monitor.

Their method requires an attenuator circuit composed of resistors to be incorporated in the video cabling between the computer and the display.

The monochrome video output from the attenuator is ideal for producing the red monochrome greyscale for the "Red Eye Shutterglasses Method".

For this implementation to work, the monochrome output channel from the attenuator is fed into the red input channel of the color display monitor.




Figure 4 from the Paper shows the general electrical schematic of the attenuator.

The normalised greyscale channel gains for G-red, G-green and G-blue are 0.30, 0.59, 0.11 respectively, in accordance with the accepted tristimulus values for the color sensitivity of the eye.

Only six resistances are needed to form the circuit. In order to determine the resistance values for the circuit, the Design Equations described in Appendix A were applied. Mathematica was used to solve for the required branch resistances.

The electrical schematic shows the six branch resistances being constructed from one or more standard resistor values for best-fit impedance matching.

Impedance matching of the cabling and connected devices is important for preventing signal reflections and interference.

The resistance in the circuit causes some minor attenuation of the video signals. However, the reduction in the overall gain can be readily compensated for by increasing the Contrast adjustment on the monitor.

The developed electrical schematic is as follows: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1770




I have built and tested this circuit using a VGA cable breakout board. I basically cut a spare 15-pin VGA cable into two and terminated the free wires to a breadboard for prototyping.

Only the Red-Green-Blue input, Red output and common ground wires are required.

The other wires are just passed through as normal.

The VGA cable break-out board with the Pelli-Zhang attenuator circuit is shown: http://www.mtbs3d.com/gallery/displayim ... ?pos=-1769




Photos of the resulting red monochrome video display will be posted sometime in the future .......


Thanks. :D

Bo_Fox
Two Eyed Hopeful
Posts: 73
Joined: Sun Oct 28, 2007 9:13 pm

Re: Different LCD shutter glasses - And ghosting

Post by Bo_Fox »

Hmmm, interesting results! I always hated the ghosting on my 24" Sony GDM-FW900 CRT, and knew that it had to do with the phosphor lag--the time it takes for the "excited" phosphors to cool down (which is an inherent problem with all CRT's, even on older TV's with movies running at only 24-30fps). Ahh, those infamous ghosting trails take like a whole 0.5 second to completely cool off when the contrast is high enough.

Are the DLP TV's really that good with ghosting?!?
8800GTX, 24" CRT (Sony GDM-FW900)
i7 920 @ 3.7GHz, Foxconn Bloodrage, WinXP-32 and Vista x64 SP2
3D shutters, 181.00 Forceware +162.50 S3D for WinXP

Other rig: 4870 1GB, i7 920 @ 4GHz, DFI T3eH8
24" LCD + IZ3D anaglyph, WinXP-32 and Win7 x64

budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Re: Different LCD shutter glasses - And ghosting

Post by budda »

Hi,

I am surprised this thread is still actively viewed.

Code: Select all

I always hated the ghosting on my 24" Sony GDM-FW900 CRT ...
It seems all CRT monitors share this characteristic ghosting effect with green and blue phosphors.

I believe the red europium phosphor makes all the difference because of its fast response and decay time.

The high-end CRT workstation monitor manufactures made a big mistake in not putting equivalently fast blue and green phoshors in these displays.

If they did, these old CRT monsters would still be superior to the latest S3D flat panel displays. These same manufacturers could have kept their CRT production lines open and still charged over $1000 for their high quality S3D monitors.

I have noticed the Sony FW900 line of "24 inch" (22.5 inch) 16:10 monitors still sell second hand for $900-$1000, whilst the slightly smaller Sony P1130 "21 inch" (20 inch) 4:3 monitors sell for around $50 second hand. There is alot of price sensitivity there, for small improvements.

I suppose there are many people who still like the performance of high end CRT's, particularly for its sharp image, contrast and true colour.


In respect of S3D viewing, the Sony P1130 has a superior scan rate of 130 kHz, to the 120kHz of the Sony FW900.

Refer also to this earlier thread regarding the Sony FW900 - http://www.mtbs3d.com/phpbb/viewtopic.php?f=5&t=2271" onclick="window.open(this.href);return false;


PS.

Old school CRT technology and S3D is not dead yet ......

I have a new CRT Red Eye Shutterglasses mod in the works which combines the advantages of anaglyph with the advantages of shutterglasses. More to come .... :wink:


Thanks.

Post Reply

Return to “General Stereoscopic 3D Discussion”