How LED Projectors Will Affect JVC/Sony Sales - Page 3

What does that have to do w/LED's?

So far I haven't seen a jump in CR associated w/LED.

Is there somewhere where I can read a technical explanation how local dimming LED on a front projector will work? I have the samsung 8500 flatscreen with local dimming LED, and the picture is nothing short of amazing. I'm totally sold on local dimming LED tech. However with a 1" dlp chip, how can they cram enough LEDs with local dimming to light such a small chip? I also have a two year old optima hd80 in the dedicated HT room which I love, but the new flatscreen in the living room picture is vastly superior to the PJ's image, but of course it's half the size. I will upgrade when I can buy an LED DLP PJ that is also 3D ready and under $5000. I hope this will be within 3 to 4 years, or am I dreaming?

-Eric

An array of light tunnels that only allows light from an array of LEDs to illuminate a specific areas of the chip? I've asked the same question myself many times, and I'm excited to see what kind of improvements in CR that the tech brings to front projection.

Light tunnels, eh? Interesting! This does sound very exciting. I'm looking forward to following this technology, and hoping one day it will be affordable for the DIYer.

-Eric

I'm fairly confident (based on what we're hearing) that we haven't seen what LED tech can really do. Most of the examples we've seen so far have just been lamp-designed machiens with LEDs dropped in for light sources (most notably the Vivitek that's essentially a Planar 8150 with LEDs).

I think we're beginning to see it with ground up LED designs like the Mico 50 (read the thread over in the $20k forum, it's very interesting).

But what I'm really thinking of is Local Dimming brought to us by the likes of Sim2, Marantz or Runco/Planar, someone who will have it setup to be completely transparent.

LEDs are already showing us they're capable of producing machines with effectively infinite sequential CR. Pair that with local diming and we'll have machines with DLP's sharpness, punch, MTF, plus LEDs "solid" colors, CRT-besting/DiLA-crushing intrascene, simultaneous contrast, and no RBE.

Wouldn't such a machine change the ballgame?

Local LED dimming isn't practical (or very beneficial even if implemented somehow, because of the light scatter) in a LED lamp based projection device. It works well on LCD flat panel TVs because they are direct view, not projected.

I'm not sure if (or when) OLED will get bright enough to allow the modern day equivalent of CRT projector, but that's the closest thing to local dimming that we're likely to see hit the market (and even then not for a long while.)

While a 2nd (or 4th) panel isn't local dimming, it would have the projector-equivalent effect. Some companies have demoed R&D prototypes, but it doesn't seem like that will hit market anytime soon either.

+1 on what sethk stated. Just take the whole local dimming off the table for projection -it does not apply.

How can you be so sure? Have you tried it, do you have inside information that it's not possible?

What light scatter? Are you talking about the washout effect described partially by ANSI CR now, where an LED projector can do over 700:1 ANSI CR? Or are you talking about light scatter around the room? Either way it doesn't seem like it leads to your conclusion. There are many scenes where the light scatter from those is not the dominant factor in intra-image CR and where a lower black floor for part of the image would improve the intra-image CR significantly. There are of course issues with how many zones you have, but that applies to flat panels also.

Given that ANSI CR reduction from a room isn't very much a local effect (it tends to wash out the whole screen) I think front projection may have some advantage with local dimming over current flat panels as far as how bad the halos are with x number of zones. There are likely multiple factors that go into this effect though and so I'm not sure how they would all play out in the end.

--Darin

local dimming on flat panels does not need any real light focusing to work. It also does not have much size constraints. It simply has one LED (or 3 R/G/B LEDs) behind several pixels, which for flat panels is a relatively large area for that LED. There are a total of about 1000 LEDs for this to work (so about one per 20 pixels on the flat panel). How are you going to cram 1000 LEDs behind a 0.65 inch square LCD, DLP, or LCOS chip? If you don't put them directly behind the chip (or in front for DLP and LCOS), where do you put them, and then since they are not in direct proximity, how does one focus their output to go not just directly to the LCD/DLP/LCOS panel, but each LED only to a set of maybe 20 unique pixels on that panel? IMO, it is simply not a technology which is practical for projection. If anyone has any alternatives to make this work, I'd love to hear about it.

I think he may be talking about the light scatter from each LED (you will need like a thousand of them). Each LED needs to send light to a group of pixels, and only that group. Imagine a 40" diagonal circuit board with nothing but LEDs on it. Now make sure the top left LED only shines light on to the 20 mirrors on a DLP chip, only on the top left side. Now do this with the other 999 LEDs (to different respective areas on the DLP chip).

I guess it can be done, by first ensuring every single LED sends light out on a straight path, and then having a massive lens to capture contract all the LEDs' output, then another lens to straighten out the light path, forming a beam of light about the size of the DLP chip. But is it practical?

Interesting to see people dogmatically state that local dimming is not possible with FP. The notion that local dimming would happen was first posted on this forum a while back, so that aspect isn't new.

Not only is the question regarding the possibility of local dimming redundant, as it can be done, 2 companies are already making very serious head way.

Im not at liberty to go any further atm, but there are some people here who are also aware of the situation, and can verify that, if they so choose.

I need to be elsewhere, so have a great Christmas.

A fat plumber told me.

The local areas can overlap. This is the way Brightside did it with their flat panels. While an LED does need to send much more light to one area it doesn't have to be perfect. For instance, even if an LED in the upper left corner lit up the lower right corner 1% as much as its focus area of pixels the local dimming could still improve intra-image CR in many images by a significant amount. With a DLP with about 4000:1 native on/off CR for instance, if the lower right corner of the chip was getting 5% of the light it would get without a local dimming area the black floor for that area would be about 20 times better than without local dimming before considering the washout effect largely described by ANSI CR. And since an image could have just a little bit of bright stuff in the upper right corner without that much washout (say 700:1 on the extreme ANSI CR checkboards and so way less washout than that with just a little bit of light in on corner compared to the far corner) the intra-image CR could go way beyond the ~4000:1 it would have been limited to without local dimming. I could see it being able to do 40k:1 or higher in an image with a little bit of light in one corner and the rest black or near black, even if the native on/off CR for the whole thing without the local dimming factor was only 4k:1.

--Darin

How well does the local LED technique work in LCD tvs of today? (I haven't really seen any more than fleetingly.) Can they produce a pitch black starfield with bright pinpoints of light? Without haloing? Do you think an LED-DLP with local dimming would be able to do as well in in a similar situation?

I suppose it will be guess work since they're not here yet, but qualified guesses would do.

I take it the number of LEDs will make a big impact. Anyone knows what kind of arrays they are working with?

I agree that local/dynamic dimming of LED light source is when we'll see the technology begin to really shine to its best advantage. However, even without this feature, it still can have many advantages over bulb-based technolgies.

Not having to replace a bulb (and buy a new bulb) is one advantage. So once projector prices fall for LED, that will be a strong incentive for consumers. And even if you can't take advantage of the wider color space with today's software, the fact that the color space is so pure and so predictable means you'll be able to dial-in close-to-perfect-color without color balance shifting over time as a bulb ages.

LED is the future, and as the kinks are worked out it will benefit all projector technologies. However, the fact that it can erradiate single-chip's most bothersome artifact, RBE, will allow the advantages of single-chip designs (perfect pixel alignment and no inter-pixel interation) to be enjoyed by those who previously wouldn't have been able to enjoy DLP.


As for people suggesting that the price delta with LED will somehow negate its advantages for the consumer...

please remind yourselves that only a few short years ago all of the "experts" on this forum were telling us that we'd never see affordable 1080p projectors for the home-market. Never. And then we were told that we'd never have 1080p source material anyway. Never.

You can search out those long-winded debates fanned by experts on this board who, in all apparent honesty, tried to convince themselves and others that the human eye couldn't even see a difference between 720p and 1080p.

The fact that just a few short years later we not only have affordable 1080p projectors, but that those projectors cost far less and perform far better than any 720p machine they owned at the time, should indicate that the price of new and developing technology today says nothing about its widespread cost tomorrow.

LED will become the ubiquitous light source for all lamp-based video display products in a few short years. Really.

I agree with most of the above comments except that LCOS is an anlog medium even though it is modulated digitally. analog is a big advantage because the single chip dlp is not fast enough to produce an 8 bit grayscale.

WE are in 2013 and nothing yet :(

The issue was brightness. The LED technology hasn't grown enough where they're able to produce more than 600 or 700 D65 lumens. This is why you're starting to see laser/LED hybrid light engines come out. Hopefully that will make up for the lack of lumens and dive the technology into popularity. Time will tell...

Optoma HD91 due in September.

Over at AVF a very experienced and knowledgeable member did a pretty rigorous comparison of the HD87 and one of the JVCs and found they were difficult to choose between - one's strength was the others weakness, and vice versa.

The 91 is said to have a brightness of I think 1300 or 1400 lumens - the 87 was said to be 1700 lumens and extremely bright. And it's often said that LEDs look brighter than the stats suggest, meaning the difference to the human eye my be negligible.

If Optoma have cracked it - if the 91 is an LED 87 - with all that model's performance plus the advantages of LED, then it'll be an exceptional projector.

But there are a lot of ifs, buts and maybes in there.

Steve W

And its an Optoma. To me not a favorable thing.

That's pretty much what you get from a JVC except with a newish lamp.

The issues are cost and black level, both of which are lower with JVC.

It is not working good at all. There will always be problems with clouding and halos, because the grids are too large. That is why most of the current LCD TVs do not use it anymore.
You should forget about LD, even if it would be worki with PJs it is a dead end. In terms of native contrast LEDs do no better than lamps.
Edited by steak2 - 3/11/13 at 2:00pm

Naively that's definitely true. But when I enable the LED modulation (Dynamic Iris) the black level is actually better than a JVC the LEDs can literally shut off on a pure black back ground. The cost is coming down as there are a number of units that street cheaper than the top of the line JVCs.

That would be fabulous yes LED and lasers can be completely off with no light.  The lamp does not turn off :(

You're absolutely right.

Manufacturers exaggerate their figures and specs, but I tend to work on the presumption that they exaggerated pretty evenly between different models in their product range.

If Optoma says the 87 was 1,700 and the 91 will be 1,400, and given that LEDs generally appear to look brighter than the specs suggest, I'd be surprised if they didn't look quite similar.

Of course, we have the added factor of dimming lamps. The 87 would be down to 1,400 (exaggerated) within a few hundred hours.

As ever, time will tell.

Steve W

I can't believe how often and important frames that are completely totally at 0 are and how important it is to have a projector that shuts its LED light sources is for those multitude of instances. This is a tremendouse thing for say an AVS meet. Everybody says OOOOOOO. Then some ungrateful notices that anything slightly above black has well its not right, yea its not linear but a quantum leap. I tell him to shut up and just hit the back button to that black frame. A LED DLP DC4 machine is a nice thing in lots of ways but I really think turn off at 0 is very infrequent and definately not a real reason to buy one of these.
Edited by mark haflich - 3/12/13 at 8:18am

I didn't mean to imply that a totally black background was the reason to buy one. The LED modulation makes a substantial difference on darker scenes and the actual black level is actually lower than any JVC. That was the point I was trying to make. Most of the time on very dark sequences the JVC will have the advantage, I've seen it many times, but there's just something about DLP that the advantage in contrast on some scenes doesn't justify my purchase because I think they're lacking on a lot of other places.

I have a Sony SXRD machine. There are obviously performance differeces depending on the chip technology involved. Then there are light source differences. Some deep thinking here by me in writing this. I like them all. Too bad we can't get the strengths of all at once.
Edited by mark haflich - 3/12/13 at 5:06pm

That's true and useful when there aren't bright objects in the scene, but not otherwise the brightness of those objects is reduced.
Edited by noah katz - 3/12/13 at 5:22pm