AVS Forum banner
Status
Not open for further replies.
1 - 20 of 30 Posts

·
Registered
Joined
·
260 Posts
Discussion Starter · #1 ·
WARNING/GUARANTEE: This is just me blowing smoke, not some tech I have found, but...

The color wheel in DLPs (see www.dlp.com for background) seems to be one of the major limits for HT specifically. It seems silly to have such an "analog" doohickey as a wheel of filters turned by a precision(?) motor in a digital display device. I'll bet that many of the "rainbow" artifacts aren't always due to the observer; ANY imprecision in the wheel coordination would easily screw up the overlap of colors. The "Thumper mods" for the UP1100 seem to include tuning up the wheel - this assembly must be a pain for mfgs, but "good enough for business use".


What you need is a better timed-filter unit. Why not a one-cell LCD, for instance? (If any such beast exists.) This would make it very simple to adjust colors, too.


DLP is a 2-stage system, with the TI DMD modulating the brightness per pixel and pixel color determined by the filtered light bouncing off of it at the time. Is a mechanical color filter the best idea that TI could come up with, or is it a $$ thing?


Just to blither away some time with wild theories, while we wait for Infocomm... http://www.avsforum.com/ubb/tongue.gif
 

·
Registered
Joined
·
260 Posts
Discussion Starter · #3 ·
The 3-DMD $$ question has been asked before, with no real answer. Probably the optical assembiles to split and recombine the light 3 ways - the optics are critical, and the alignment more so. Precision = $$$. Single-element FPs seem to be the only way to get the cost within reason.


I just noticed a comment in the thread about the SID conference that some geek may have stolen my idea already - I have to publish faster!

** Sorry - it was this thread: http://www.avsforum.com/ubb/Forum10/HTML/006520.html


[This message has been edited by bbordner (edited 06-06-2001).]
 

·
Registered
Joined
·
148 Posts
I agree completely, the system is archaic, but let's include the arc welder that is used to produce light. You should be able to replace the entire bulb, colorwheel contraption with a modulated array of high intensity red green and blue leds, and get rid of the heat, fan noise, dust infiltration, and possibly a great deal of the rainbow. From what I read, a high intensity led can put out about 30 lumens at a limited angle. The entire array required might dissipate less than 10 watts, allowing for a completely sealed projector.


Do we have any physics students out there that can poke holes in this?


------------------
Bob F.


"He thrusts his fists against the p-p-posts" Bill



[This message has been edited by REF (edited 06-06-2001).]
 

·
Premium Member
Joined
·
2,564 Posts

·
Registered
Joined
·
23 Posts
I like the L.E.D. idea-- out here in Northern California they have been replacing a lot of the traffic lights with the new models that use L.E.D.s-- apparently as sort of energy-saving measure. I will say that they seem *significantly brighter* than the average lightbulb-based traffic signal despite the energy effectiveness of the design. It would be neat to see if a projector could be built around this technology.
 

·
Premium Member
Joined
·
25,550 Posts
"Single-element FPs seem to be the only way to get the cost within reason."


3-chip LCD projectors have these optics, so at this point, it must be that a DLP chip costs more than an LCD chip.


------------------

Noah
 

·
Registered
Joined
·
578 Posts
Lots of LCoS (liquid crystal on silicon) microdisplay systems are using new types of color techniques. No idea why a DLP set up could do the same yet all the one chip designs are using goofy color wheels. JVC uses a holographic filter made by Dupont. Philips is using rotating prisims to filter RGB and scan them across the chip at the same time to eliminate the 'rainbow' effect of the color wheel. Other systems just use 3 sepp. RGB LCD filters.

About using LEDs for the light source and color element, I think that'd by far be the best approach. Very simple, very compact, basically no heat, long lasting, but I think the one catch is that it's not nearly bright enough. Maybe I'm totally wrong on that, but I think that's the only reason it's not being attemped (at least in something 'near' market that's been announced).

If anything it'd first hit the brightness level needed for rear projection before front proj. A product that looks like it'll be incredible is the SCRAMscreen (from www.scramtech.com -check it out!). sheets of fiber optics stacked to create a screen that drastically improves contrast and brightness of any microdisplay system. My first HDTV looks like it's gonna be rear projection when I had thought it'd be front till I read about this killer simple system!!
 

·
Premium Member
Joined
·
25,550 Posts
Unless I'm missing something, this is just a method of transporting the light from the imaging device to a rear screen, it doesn't address the issue under discussion.


I don't see how they can make it as inexpensive as a simple rear diffusion screen.


------------------

Noah
 

·
Registered
Joined
·
578 Posts
The SCRAMscreen is the screen itself, and not the process of getting the image 'to the screen', though that part of the display system was something I did also mention.

And so what if the technology I mentioned wasn't DIRECTLY related to the original post!?

The first part of my reply post WAS directly related to the original question of color wheel/color filters. Mentioning the SCRAMscreen was related to something I thought the original post was also "getting at".... How some parts of a display are very "high tech", yet some like a color wheel AND (as I added to your dismay?) rear projection screens aren't. These all need to work together to form a 'picture' so why are you bothered that I mentioned it!? I personally love to hear about new technology that I didn't know about, so I mentioned something that I thought others might be interested in, just as someone posted something on this thread about LEDs to replace the bulb and color wheel.

You're pretty rude to make your little attack blurb... which by the way had nothing to do with the original post. And how ironic that that was what your complain with my post was about.
 

·
Registered
Joined
·
578 Posts
And since you mentioned it.... the screen I'm sure can't be cheaper to make than a simple diffusion screen, but it eliminates a large mirror, complicated expensive optics, reduces the strain on the light source (=cheap probably), and makes a thinner, simpler (again =cheaper probably) cabinet.

Mostly it just shouldn't end up being much more costly, and should make vastly improved picture.

But like you said this isn't what the original post was about. I had already started a diff. thread about this new screen, so don't complain again about my posts.
 

·
Registered
Joined
·
265 Posts
the rotating prism solution will not work for DLP since that solution is based on frequency around 180 to 200 Hz.

LCOS manages that frequency only just, but DLP is yet too slow unfortunately.
 

·
Registered
Joined
·
260 Posts
Discussion Starter · #14 ·
Wow - I go away for a few hours, and come back to all this! Please remember that this topic is strictly wanking, and don't get offended/excited by anything...


Work permit - I forgot about color wheels being used in early color TVs - good one! The evolution of color broadcast TV was insanely convoluted (to make it B&W compatible) - I'm still amazed that it works. In general, what I was trying to say is "eliminate moving parts".


The LED idea is cool, as there have been great improvements lately (which I know no details). I've seen some new traffic signals which seem painfully bright - maybe LED? If anyone has substantive lumen/cost info, please post... The bulb is another expensive "clunker" which definitely holds back FPs! Most are designed for a 2-3 year business-use life, rather than the 7-10 years we would hope for in a HT system, so these "quickie" solutions persist.

"But that's just my opinion; I could be wrong..."
 

·
Registered
Joined
·
148 Posts
Unless I've missed something!


Hi Intensity leds can be had in small quantity for less than $2.50 each. The Light output is 2500 mcd or greater, or 2.5 candela x 4 pie lumens per candela = 30 lumens per led.


In an array of 30 leds per color, that's 900 lumens for each color array and $225 total. The rest of the components are peanuts.


Isn't that 2700 lumens of white light, or only 900 lumens?

This would also last the life of the projector.


------------------
Bob F.


"He thrusts his fists against the p-p-posts" Bill



[This message has been edited by REF (edited 06-07-2001).]


[This message has been edited by REF (edited 06-07-2001).]
 

·
Registered
Joined
·
203 Posts
The timing of this thread is coincidental in that Texas Instruments has just announced a new type of color wheel that will yeild much more brightness and/or color saturation.


"Sequential Color Recapture - SCR - is a technology that allows all three colors to be present simultaneously, such that a single panel DLP system can now be as efficient as - potentially more efficient than - a three panel system, but without the additional parts, size, weight and expense of three panel systems."

http://biz.yahoo.com/bw/010607/2208.html



------------------

Skyhawk

- Infocus LP350
 

·
Registered
Joined
·
203 Posts
Oh I should add that since this new technology allows all the colors to be filtered simultaneously, I suspect this technology will also take care of our little "rainbow" problem as well.




------------------

Skyhawk

- Infocus LP350
 

·
Registered
Joined
·
5,211 Posts
Remember the application. Always crucial when talking about any product. WHat was it designed for.


THe one chip DLP device was designed for small business portable projectors. Yes they have gotten better, yes they have gotten smaller and brighter, but they were only ever designed for business portable.


Higher quality was always intended for the two and three chip beasties.


When trying to squeeze more performance out of something you always have to remember that the original design will always be a limiting factor for you.
 

·
Registered
Joined
·
260 Posts
Discussion Starter · #19 ·
The SCR method sounds great; relatively cheap, as it's just a new wheel and timing electronics. It still leaves us with a whirling dervish filter and expensive, fragile bulb.


I love the LED idea, but after thinking more, it ain't gonna happen. The real killer is color - you get only what the LED puts out, and I would bet that none now produced have the right RGB needed for video. Also totally non-adjustable. From what little I know, anyway... nutz!


We've asked about 3-chip DLPs before; never got a satisfactory answer, but apparently it will remain expensive for quite a while. DMD prices may drop like a rock, but precision optics costs won't change much.


Back to one of my previous ?s: is there such a thing as a 1" LCD that could be used to replace the color wheel? It would seem that you could have much better control over the timing and color adjustments, at least. With control of individual cells in the LCD, you could do other things (beyond my tech knowledge).


[This message has been edited by bbordner (edited 06-07-2001).]
 

·
Registered
Joined
·
203 Posts
bbordner,


SCR must involve a little more than a new wheel and electronics, as the press release also mentions that changes in the DMD itself are necessary. It sounds like a fairly radical change (how can 3 colors be filtered through a color wheel at once?), and one that we probably wont see on the market anytime soon. But I figure my current projector will last me a couple years at least anyway.


If Dennis Fritsche isn't exaggerating when he says this new technology will in all respects "match three-panel modulators with a single DMD", without the cost, size, alignment, and complexity of three-chippers - it pretty much makes the discussion about 3-chip versus color wheel a bit obsolete for the use which we put our projectors. And I suspect the "fragile" bulb at least not get any more expensive or fragile in order to pump out 40% more brightness out of these units. I suspect the big 3-chippers used in digital cinema are a bit expensive too. I really don't see the difference in bulbs with this new color wheel concept verus 3-chip designs if the press release is accurate.



------------------

Skyhawk

- Infocus LP350
 
1 - 20 of 30 Posts
Status
Not open for further replies.
Top