• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Canon demonstrates SED monitors

xtknight

Elite Member
http://www.behardware.com/articles/593-...-encounters-of-the-third-kind-sed.html

Wow, 400 cd./m² at white level and 0.004 cd./m² at black level, not bad! Can't wait to get my hands on one of these.

These are better than CRTs in a ton of ways:

[*]INSANE resolutions:
Prototype electron emitters have been
developed with diameters of a few nanometers (billionths of a meter).
[*]Much sharper text.
[*]Thin.
[*]Less weight.
[*]Much less power consumption.
* energy consumption that is roughly one-half that of a large-screen
CRT and about one-third that of a plasma display panel.
[*]Less electromagnetic radiation (no deflection yoke) and thus probably less lead. This might reduce headaches and eye strain, maybe, but I'm guessing on that one.
[*]I've heard they have a wider color gamut.
[*]Perfectly flat, individually-addressable matrix of pixels.

Better than LCDs:

[*]Much better black level than CCFL LCDs.
[*]0.002 ms response time(?)
[*]More accurate colors.
[*]Should cost less than LCDs (so I've heard).

Better than OLEDs:

[*]Should be available much sooner.
[*]Probably costs less.
[*]Better lifetime.
[*]Less size scaling issues.
[*]No special sealing required.

SED (Canon: Surface-conduction Electron-emitter Display)'s aliases: FED (Samsung: field effect display), NED (unknown)

Unfortunately, SEDs still flicker like CRTs (I think?) and are still stuck at a certain resolution, but we'll see what becomes of resolutions when Vista's vector-based stuff comes through.

More info:
http://engineering.dartmouth.edu/other/...94/student/tkerner/FED_talk/intro.html
http://www.meko.co.uk/sed.shtml
http://answers.google.com/answers/threadview?id=525928

For those that want to know everything from the ground up of SEDs, here's Canon's patent: http://patft.uspto.gov/netacgi/nph-Pars...2+AND+display)+AND+emitter)+AND+canon)

Probably some more patents here: http://patft.uspto.gov/netacgi/nph-Pars...D+display+AND+emitter+AND+canon&d=ptxt

Cliffs: SED monitors by 2006-2007, costing the same or less than LCDs.
 
...ooooh!

The question is, will this scale to larger screens, but cost less for the larger screen than competing techs?
 
Originally posted by: So
...ooooh!

The question is, will this scale to larger screens, but cost less for the larger screen than competing techs?

My prediction: Almost certainly, but not necessarily at launch time.
 
Originally posted by: So
...ooooh!

The question is, will this scale to larger screens, but cost less for the larger screen than competing techs?

The preliminary numbers were that SED will work very well for big (at least up to 50-60") screens, but it may not scale down very well. It's also going to be (relatively) heavy, since the front is a big glass panel like a CRT. And they suck power like a CRT (although this may get better).

If both SED and OLED work out as intended, we'll probably see SED take over for big HDTVs and computer monitors, and OLED used for portable and/or lightweight applications. Of course, this is a big if, and we're probably looking at at least a couple years before either of those technologies really starts to mature.
 
Originally posted by: Matthias99
Originally posted by: So
...ooooh!

The question is, will this scale to larger screens, but cost less for the larger screen than competing techs?

The preliminary numbers were that SED will work very well for big (at least up to 50-60") screens, but it may not scale down very well. It's also going to be (relatively) heavy, since the front is a big glass panel like a CRT. And they suck power like a CRT (although this may get better).

If both SED and OLED work out as intended, we'll probably see SED take over for big HDTVs and computer monitors, and OLED used for portable and/or lightweight applications. Of course, this is a big if, and we're probably looking at at least a couple years before either of those technologies really starts to mature.

It may be a while before LCD's die though. Too much manufacturing infrastructure. LCD's have come very far, very fast. Almost to the point of entrenchment that CRT's had up until the late 90's, IMHO.
 
Originally posted by: Matthias99
And they suck power like a CRT (although this may get better).

If both SED and OLED work out as intended, we'll probably see SED take over for big HDTVs and computer monitors, and OLED used for portable and/or lightweight applications. Of course, this is a big if, and we're probably looking at at least a couple years before either of those technologies really starts to mature.

FEDs are supposed to take 4x less power than LCDs.
http://engineering.dartmouth.edu/other/...94/student/tkerner/FED_talk/intro.html

Canon speaks of computer monitors end of 2006, early 2007.
I suppose 'computer monitors' are 23" or smaller? There have been 1" prototypes of FED. Assuming FED is the same as SED, that sounds damn good. Well, I'm not sure of the resolution though so that means nothing.

I wonder, will SEDs still flicker like CRTs?

I guess SEDs are quite not the same as FEDs but rather a type of them.
http://www.meko.co.uk/sed.shtml
 
Originally posted by: xtknight
I wonder, will SEDs still flicker like CRTs?

I guess SEDs are quite not the same as FEDs but rather a type of them.
http://www.meko.co.uk/sed.shtml

AFAIK, SED's are similar to CRT's in that respect, basically, they are CRT's with a micro (maybe nano) scale electron gun behind each pixel, as opposed to one gun firing in sequence at each pixel. While I'd hypothesize that it would be theoretically possible to make the gun 'on' or 'off' all the time, it'd probably put out a lot of heat and take a lot of energy...maybe too much to be practical. So, yeah, from what I've heard, SED/FED's will have the same 'flickering' that a CRT has, though possibly at a higher rate than a TV (say the rate of a CRT at 100Hz).
 
I can't wait for these things. On paper they appear to be something like the ultimate displays aside from the fixed resolution.
 
Being stuck at one resolution has been the #1 reason I have not gone to LCDs, although their scaling has much improved as of late. If these will be physically stuck at one resolution at launch, I'm not getting one. Other than that, they sound good.
 
Originally posted by: Avalon
Being stuck at one resolution has been the #1 reason I have not gone to LCDs, although their scaling has much improved as of late. If these will be physically stuck at one resolution at launch, I'm not getting one. Other than that, they sound good.

You might regret not getting one. 🙂 If text size is the reason you like different resolutions then you can knock that off your list immediately if you plan on getting Vista. If game performance is the reason, well, that is a problem.
 
Originally posted by: xtknight
Originally posted by: Avalon
Being stuck at one resolution has been the #1 reason I have not gone to LCDs, although their scaling has much improved as of late. If these will be physically stuck at one resolution at launch, I'm not getting one. Other than that, they sound good.

You might regret not getting one. 🙂 If text size is the reason you like different resolutions then you can knock that off your list immediately if you plan on getting Vista. If game performance is the reason, well, that is a problem.

Two more quick points:

1) Fixed-resolution displays (assuming the resolution is a reasonable one) are VASTLY superior for digital content that matches said resolution, or for content that can be dynamically generated at any desired resolution. If display resolutions stop increasing appreciably past 1080P, video hardware will eventually catch up to the point where 1900x1200ish displays are not difficult to render for even on relatively cheap cards.

2) CRTs still have a "fixed" resolution in terms of physical phosphor dots on the screen; it's just usually pretty high (on a good-quality 20/21" monitor, there are usually somewhere in the realm of ~2000x1500 phosphor triads). The electron guns in a CRT display effectively do a very good gaussian resampling of the digital signal onto the analog phosphor dots. If the scaling being done in a fixed-pixel display is good enough, and the physical display has a high enough resolution, it will not look any worse than on a CRT.
 
Originally posted by: Matthias99
2) CRTs still have a "fixed" resolution in terms of physical phosphor dots on the screen; it's just usually pretty high (on a good-quality 20/21" monitor, there are usually somewhere in the realm of ~2000x1500 phosphor triads). The electron guns in a CRT display effectively do a very good gaussian resampling of the digital signal onto the analog phosphor dots. If the scaling being done in a fixed-pixel display is good enough, and the physical display has a high enough resolution, it will not look any worse than on a CRT.

Wait a minute here...so the SEDs may be able to do the Gaussian blur too? And how are electron guns doing this effect by themselves? Is this like the canceling out of two audio waves or something inherent like that on the light-wave level?
 
Originally posted by: xtknight
Originally posted by: Matthias99
2) CRTs still have a "fixed" resolution in terms of physical phosphor dots on the screen; it's just usually pretty high (on a good-quality 20/21" monitor, there are usually somewhere in the realm of ~2000x1500 phosphor triads). The electron guns in a CRT display effectively do a very good gaussian resampling of the digital signal onto the analog phosphor dots. If the scaling being done in a fixed-pixel display is good enough, and the physical display has a high enough resolution, it will not look any worse than on a CRT.

Wait a minute here...so the SEDs may be able to do the Gaussian blur too? And how are electron guns doing this effect by themselves? Is this like the canceling out of two audio waves or something inherent like that on the light-wave level?

Theoretically, filtering hardware/software can do just as well as a CRT. It's just that in order to do it *fast*, most software and hardware takes serious shortcuts (like not taking enough samples, or doing linear sampling rather than quadratic or gaussian). It's no different than scaling an image by any other means; you can see similar effects by playing around with different resizing methods in Photoshop. Bilinear filtering looks pretty bad; Gaussian sampling looks pretty good.

Shine a flashlight at a wall. See how the middle of the spot is brightest, and the brightness falls off gradually towards the edges (unless you have some sort of fancy flashlight that carefully focuses the beam)? You get the same sort of effect (on a sub-millimeter scale) with an electron beam, and the scatter pattern can be modelled as a gaussian distribution around where the beam is aimed.

So, on a gross mathematical level, what a CRT display basically does is to do a very high-quality gaussian resampling of whatever signal you feed it to the physical resolution of its phosphor grid. Even with infinitely fast internals, a CRT monitor cannot truly display more pixels than it has phosphor triads.
 
Well, I created an image in Photoshop containing various text sizes of the infamous "The quick brown fox jumps over the lazy dog" phrase (in Windows' fonts). I created the image in 800x600 and scaled it to 1024x768 and 1280x1024. I was sure to rasterize and combine layers knowing that Photoshop is smart enough to resize using vectors. :Q

Photoshop's resizings look as bad as my LCD's. There is no Gaussian option, only:
[*]Nearest neighbor: ugh, why is this in here? 😛 This reminds me of Doom 1.
[*]Bilinear: a little worse than bicubic
[*]Bicubic: bad at 12x10, OK at 10x7, probably because of aspect ratios
[*]Bicubic smoother: worse than Bicubic.
[*]Bicubic sharper: worse than Bicubic.

There is a Gaussian blur filter if that has anything to do with what I want to do.

Oh, and I updated my OP a couple hours ago with a lot of new info. Canon says these SEDs do consume less power than CRTs.

What I wonder is if SEDs can do this analog Gaussian thing with no microchips' help and thus in real time.
 
Originally posted by: Polyfluoroethylene
Looks expensive.

No, not really. If that FED site is any indication, it says 40-60% cheaper than AMLCDs. Other sites say just a tad over a normal LCD, which it may very well be at production time. But I wouldn't mind paying for a display that is actually decent. This is up there (not in price, but with tech) with the $50,000 Brightside LCD.
 
Originally posted by: xtknight
Originally posted by: Avalon
Being stuck at one resolution has been the #1 reason I have not gone to LCDs, although their scaling has much improved as of late. If these will be physically stuck at one resolution at launch, I'm not getting one. Other than that, they sound good.

You might regret not getting one. 🙂 If text size is the reason you like different resolutions then you can knock that off your list immediately if you plan on getting Vista. If game performance is the reason, well, that is a problem.

I could resize my text, if that was even something that would bother me. I'm concerned about the very wide performance delta of the many games out there that I play. I play at anywhere from 800x600 to 1600x1200, depending on what game I'm in the mood for. If a SED can do multiple resolutions and look good at the same time, great. If not, like I said, I'm not getting one.
 
Doesn't a CRT relly on actualy physical effects to attain it's pixel resampling rather than computational methods? Doesn't it just switch the input signal to another logical pixel partway through the display of a physical pixel?

Also, for resampling resolutions, you would not want to use gaussian blur. Gaussian blur is for blurring, not resolution resampling. That's why you're not seeing the option in photoshop xtknight. In resolution resampling, you're never going to sample more than a handful of pixelx unless you're scaling from a high resolution to a rez much lower (rarely the case for flat panels and such).

xtknight, use photoshop to bilinearly rescale your image to a size that is ~1.7 times the original resolution of your picture to see an effect like what Matthias99 is talking about.
 
Originally posted by: So
Originally posted by: xtknight
I wonder, will SEDs still flicker like CRTs?

I guess SEDs are quite not the same as FEDs but rather a type of them.
http://www.meko.co.uk/sed.shtml

AFAIK, SED's are similar to CRT's in that respect, basically, they are CRT's with a micro (maybe nano) scale electron gun behind each pixel, as opposed to one gun firing in sequence at each pixel. While I'd hypothesize that it would be theoretically possible to make the gun 'on' or 'off' all the time, it'd probably put out a lot of heat and take a lot of energy...maybe too much to be practical. So, yeah, from what I've heard, SED/FED's will have the same 'flickering' that a CRT has, though possibly at a higher rate than a TV (say the rate of a CRT at 100Hz).
I think having the gun one all the the time would actually yield longer device life because the temperatures and plasma intensity would be more consistent over time. You seem to be thinking that the gun would be on all the time at the same wattage but if the gun is on longer, the instantaneous wattage should be dropped proportionally to the increase in the time it is on.

And yeah, I read some of the linked articles and the panel is passive matrix so it'd be susceptible to flicker. It would be possible to make it active matrix but that would kill it's cost advantage to plasma and lcd. I guess they could get around it if they had the screen run at 120hz or something like that.

 
Originally posted by: zephyrprime
Doesn't a CRT relly on actualy physical effects to attain it's pixel resampling rather than computational methods? Doesn't it just switch the input signal to another logical pixel partway through the display of a physical pixel?

Also, for resampling resolutions, you would not want to use gaussian blur. Gaussian blur is for blurring, not resolution resampling. That's why you're not seeing the option in photoshop xtknight. In resolution resampling, you're never going to sample more than a handful of pixelx unless you're scaling from a high resolution to a rez much lower (rarely the case for flat panels and such).

xtknight, use photoshop to bilinearly rescale your image to a size that is ~1.7 times the original resolution of your picture to see an effect like what Matthias99 is talking about.

lol, daylight savings time on this forum is really getting screwed. we are in a time warp! anyhow:

Ok, but why 1.7?

Edit: tried 170%. It looks just as bad. 🙁 wait, did you mean 170% on both height and width (ended up with 1360x1020), or did you mean 170% of the area of 800x600? I'm not thinking straight today. So 800x600 is 480000 square pixels, and 1.7x that is 816000. How would I scale that to a 4:3 resolution? 4a*3b=816000 or something? Someone figure it out a and b. 😛
 
Back
Top