I'm intimately familiar with this topic because I was an early adapter of DLP projectors for home theater use. 1-chip DLP has a serious flaw, in that, at any given moment, there is only one of the primary color being projected onto the entire screen. Some people are much more sensitive to the so called "rainbow effect." Moving your head quickly can exacerbate the issue. They reduced the problem by either spinning the RGB wheel faster, or doubling up the colors into RGBRGB wheel, so that one color stays on the screen for a shorter amount of time, each. Technically, even 3-chip DLP will have many pixels displaying just one color at some moments, but these are per-pixel effect, not whole screen effects. Fundamentally, what is projected onto the screen is a pulse-code-modulation pixels. Each pixel is either ON or OFF at any given time, and in case of 1-chip DLP, in one of the Red/Green/Blue. There is no grayscale. It's just ON or OFF. Each pixels blinks at incredibly fast rate, you can't normally tell, and the brightness is controlled by the rate of ON time, vs. OFF time (hence PCM). For example, 50% brightness would be represented as ON/OFF/ON/OFF/ON/OFF and so on. 1/3 brightness would be ON/OFF/OFF/ON/OFF/OFF, although they could throw in some randomness to making it look more natural to human brains.
a fun detail about the rainbow effect is that it gets significantly worse if you're not directly looking at it. The human eye is a lot more sensitive to motion at the fringes of vision (higher density of cones) than at its center and so if you look askance at it you see the flicker much more easily. Once you notice this it's really interesting how many other things you can see flicker--fluorescent lights, LCDs, and more!
@@SalmanHusainGplus Helps when you squint a bit and move your head quickly from side to side! I've always had some fun when going to the theater looking for the effect but never knew what caused it. Amazing!
OK but why would you even need a projector now that you can get a 77" OLED for $2500 and the tech has been making meaningful improvements past few years? And if you can afford a movie theatre in your house, wouldn't you go for a micro LED or something, projectors just look washed out in comparison and probably can't satisfy a modern nit whore anyway.
Yeah, I really hated 1-chip DLP projectors we had in school because of that rainbow effect. But it seems most people don't see or notice it, lucky for them.
@OlafWillocx I imagine they were able to control brightness with the original DEFORMING-Mirror-Device, but they just couldn't manufacture a device with all those hundreds of thousands (to millions) of mirrors that deformed at the same rate. I doubt it's practical even with current technology to map each pixel to different analog algorithm to get a consistent image across the whole device. DIGITAL gets rid of this problem. Slight variance among individual mirrors won't matter at all. I also remember reading about how the molecular fatigue behavior is entirely different at nano scale. If macro scale physics numbers were applied, there's no way the hinges would last long enough to make the DMD chips viable, but they apparently don't behave that way, and are incredibly long lasting at nano scale.
@@ggboss8502 MEMS and semiconductors are related since many MEMS devices include control circuits and they share similar manufacturing processes. Photonics is a very specific branch of light manipulation using technologies similar to semiconductors (you can create voltage controlled light switches, for example), but in general these do not employ MEMS techniques. Having said that, MEMS devices such as the DMD described in the video can be used within photonic systems but only for relatively low speed switching compared to true photonic switches.
@@ggboss8502 I think the main difference is one is mechanical, and the other is electronic? But mems also depend on semiconductors for their functioning. So I consider mems to be a subset of semiconductors, perhaps at the extreme going into nanotech.
@@ggboss8502 sorry I just realized you mentioned photonics and not necessarily MEMS or even optical MEMS tech. But consider this: the semiconductor fabrication process relies to a large extent on optical technology. It's essentially a reverse microscope. So I consider optics and semiconductor tech to be 2 sides of the same coin. It's basically at the same size for one to be manipulated by the other
Soo im wondering if projector mems technology has been considered for use with vr/ar technology? The way how the eyes works is that light is bouncing around in a room so maybe u could shoot laser inside of eyeball to make hologram? They should use this and try to make a technology that totally shoots lasers directly into your eyes and you hallucinate holograms and 4dimensional VR/AR games and slowly burn your eyes out of your skull. That sounds to me like great idea to further drive the technocratic societies attention spans down the shitter. Then no more ‘mental illness’ because everyone will be surviving in a scarce environment.
When you make normal semiconductors mems is huge dimensionally. Getting the hinging to work really is as difficult as you think. It was the final yield limiter to really get ironed out.
@@rydplrs71 I'm excited to see mems speakers finally break onto the scene and see a similar revolution in the audio world to match the growing mems microphone market, it's been a long time coming. Some aspiring vendors like Australian Audiopixels have been laying it on thick claiming to be able to fully replace current loudspeakers using flat mems chip arrays - I have to say I'm skeptical but more than willing to be surprised if I could get these into my room which doesn't have anywhere near the space for a decent 5.1 speaker set up.
DLP stuff is always fascinating to me. I had the privilege to work at TI in the UK from the late 90s to mid 00s, although not as part of the team supporting the DLP business. The DLP guys were generous with their time in explaining stuff and TI also produced quite a few internal papers on how the technology worked and the development history. I still have my copy of the TI Technical Journal which was dedicated to this topic.
Some theaters use DLP projectors, but they don't have the contrast range that LCOS (liquid crystal on silicon) projectors do. You didn't mention the pulse width modulation used on the mirrors to generate shades of gray in each color, but it was a good overview. Thanks for all you do.
My understanding was that the contrast ratio of DLP was effectively infinite, because to generate black, the DMD just needs to dump 100% of the incoming signal into the void rather than onscreen. Are there other factors I’m missing?
@@PaulFisherere are several sources of flare in DMD. 1) the material below the mirror itself does show through a bit since it has to be exposed for a mirror to be free of its adjacent companions. While this is typically coated in a material selected to be as black as possible over the desired wavelength range, it’s not perfect. So even if 100% of light directed at the mirror is diverted, there will be some reflected light from the substrate. 2) the mirror can only tilt so far. Ideally you’d direct all the light onto a light dump orthogonal to the exit pupil, but the mirrors can’t flex 90 degrees, only about 10-12. So some light will be scattered from the mirror edges and the anchor via structure, letting a little light through even in the static full off position. 3) ideally, your light source would have total spatial coherence (or equivalently, be a point source at infinity). Real light sources are only partially coherent and have finite size, so some photons impinge on the mirror at a small angle. When these are reflected in the off state, some small amount of light will still go through the exit pupil. 4) the light dump material is chosen to be as black as possible and actively cooled from behind. However in practice, a true black body does not exist, so there will be a tiny amount of reflection which will bounce around the DMD cavity. Some of this will escape through the pupil.
@@Grak70 We had a Christie 10,000 lumen 3-chip DLP projector which was replaced by a Sony LCOS-type projector. The DLP unit had noticeably higher black level than the Sony. I don't recall the light output rating of the Sony, but it used a xenon lamp where the DLP projector used two mercury lamps - so it wasn't exactly a fair comparison. However, in addition to higher light output, the Sony had a much lower black level than the DLP unit from Christie.
@@curtlundgren6867 sounds about right. Off state black levels are a major reason DLP became obsolete as LC black levels improved and then LED off-current improved. Then came active LED, OLED, and then quantum dots with ever better black levels. There was a lot of room at the bottom for solid state light solutions, but DLP could never quite keep up. MEMS in general are much more finicky and lower yielding to fab, not to mention they face very different challenges to CMOS. TI repurposed a lot of silicon equipment to make DMDs and I think the ROI of using apples to make orange juice was difficult to justify in the long term, especially with such stiff competition from Japan, Korea and later China. Once China started pumping out decent quality LED displays that kicked even the best DLP’s ass, the margins in display evaporated.
Always love when another underrated UA-camr I've been following for a long time gets a mention somewhere. Mikeselectricstuff always brings the goods, just like Asianometry.
The amount of time you invested to research this must be astounding... great job. I find it very odd how some companies with just one major product line that costs tens or hundreds of thousands of dollars simply fail horribly once a new product arrives, rather than trying to adapt.
If the competitor patented their tech and the out dated tech is unable to complete there's not a company can do. Even if it's not patented if the new technology is fundamentally different the old company won't have any experience with the new technology and would struggle to compete. That's why R&D is so important.
15:00 Whoever came up with this mirror configuration must have been a genius. Three beams of almost identical frequency light splitting in entirely separate directions, then coming back perfectly in line with each other. On the diagram it looks like chaos, yet every angle of that configuration is a careful design choice. How do you even approach that? Looks like magic to me.
My home theater system uses this along with a laser light source, and the video quality, the color reproduction, the resolution and the contrasts are outstanding!
You've taught me so much stuff that I wouldn't even know to ask about and you make things that seem "boring" absolutely fascinating bro. I get so excited whenever you post a video that when I see a new thumbnail I make noises lol.
Thanks, fascinating story. I worked with a GE Talaria projector in the early ‘80s, it gave a sharp and extremely bright image (which was essential because it was being used as a back projector in a well-lit TV studio). The thing did have drawbacks, slow to warm up, lousy colour purity and an interruption to the video feed would generate a line of bubbles in the oil which would take a couple of minutes to clear as the disc slowly rotated. As I recall, the potential for the oil to contaminate the vacuum in the light valve was circumvented by a catalytic device that ‘cracked’ large molecules to smaller ones that could be absorbed by a getter.
I remember reading about MEMs as far back as the 80s, in EDN. Saw my first TI DLP projector, a consumer grade one, at the CES trade show, either in the late 90s, or early 2000s. I, too, once worked for a company that won an Oscar (Moviola). Though they won that Oscar decades before I was born.
I love projectors! Especially projection mapping. I learned so much and this one I had never imagine before. Super cool. Everyone raves about those Panasonic three chip DLP laser projectors, but I’m not sure how many industry folks know the nitty-gritty like you illustrated there.
Nearly all laser projectors in cinemas still use DMDs/DLP for their imaging, the lasers simply replace the xenon arc lamps as the light source and can improve contrast due to the narrow light band of the lasers
@@moki123g generally they dont, shutting the lasers off between frames keeps them from producing stable colours and in cinema units colour accuracy is paramount. blanking is more something done by the DMDs themselves by simply putting all pixels to the off position and holding them there. It is used fairly extensively even with arc-lamp based units, principally its used to keep you from seeing the image go hilariously terrible while the rest of the imaging system re-adjusts for a new raster timing/resolution. the main benefit to the lasers is their narrow colour bands means less light that doesn't go to the DMDs, which increases contrast because there is less 'unwanted' light coming out of the projector
I absolutely LOVE your channel. I love how you deliver information on topics I personally know very little about in such a concise and to-the-point way. You make me want to learn as much as I can, thanks for that
Excellent presentation. I’ve been working on “light valve” computing for well over two decades. It’s particularly useful in new areas including quantum computing.
I had a single chip DLP retroprojector around year 2000. Although it was only SD, predating HD by a couple of years the image was great. If you examined it close up the pixels didn’t have black around them, unlike the plasma screens that were popular at the time.
Amazing doc, I had a series of Mitsubishi CRT based TV's in the 1980's, and the last one I had was a 65-inch and had three 9-inch glass focusing lenses. The innovation of the home versions of projection TV ran parallel in R and D and must have helped fund design of the new, huge screens we take for granted. In the 1980's, people looked at 65 inches of TV and were like, bro, you like TV, or what? Smirks sometimes followed. We took the snide comments in stride, and continued to support the early expensive tech. We were pioneers!
Yeah. The US military spends fortunes to be the most technologically advanced organization on Earth. DARPA tends to be the first adopter for many new technologies, because the DOD cares a lot about gaining any possible advantage and has the money to singlehandedly pay for the massive development costs of new technologies that don’t have other uses yet.
15 years ago, a company called Glimmerglass made 3D MEMS capable of directing 512-1024 frequencies of light for switching in optical networks. Cost of each unit to manufacturers incorporating it into their products was over $200,000, as I recall!
I was working with Synelec and Eyevis for 25 years with DLP projectors for large format video walls for conference rooms, security centers, power plants, and so on.
DLPs are in automotive headlights now. As ridiculously expensive as they are, it's one of the most exciting modern semiconductor features in cars to me.
I just use aftermarket LED light bulbs to blind the people in front of me. The key is to use purplish and blue color which really distracts them and makes them turn their heads away from the blinding light.
@@megalonoobiacinc4863 Anything to slow them down would work. Thanks for the strobe idea. A nice multi colored strobe led would make them crawl on their knees.
15:00 I would argue the more significant advantage to braking out the RGB to dedicated chips, is that you get rid of the color disintegration effect (as the colors are no longer using time based sequencing, and the time base is instead confided merely to the luminance component of the image) one experiences when you pan your eyes over a single DLP projection, where you see individual color trails (instead of the final intended composited color) streak through your vision, a very noticeable and irritating effect of single chip DLP. Also, sometimes, commercial projectors use dedicated RGB light sources, instead of a prism, to separate out the color channels.
I rarely comment on any video. But after many videos and months following you, and after discovering some other scientific divulgation channel I must say, I appreciate your research, very well done; but I also have to say I really like your voice for narration. Other channels, despite having a well researched topic, don't have the skill/knowleadge to make ther voice a pleasent tone to be listened. Keep the good work, love your videos, would love to see a future video about China's nuclear power and the state of nuclear power generation in Asia overall after fukushima and russia-ukarine war
Love my home single chip DLP. It uses 4 color RGBY variable speed wheel, can do true 24hz and 48hz, its amazing. Only downside some rainbow effects common to wheel projectors and chip resolution is 1024x576p down-converted from source HD, looks way better than that implies though, basically cinema quality on a 2 meter (80 inch) wide screen
There are several sources of flare in DMD. 1) the material below the mirror itself does show through a bit since it has to be exposed for a mirror to be free of its adjacent companions. While this is typically coated in a material selected to be as black as possible over the desired wavelength range, it’s not perfect. So even if 100% of light directed at the mirror is diverted, there will be some reflected light from the substrate. 2) the mirror can only tilt so far. Ideally you’d direct all the light onto a light dump orthogonal to the exit pupil, but the mirrors can’t flex 90 degrees, only about 10-12. So some light will be scattered from the mirror edges and the anchor via structure, letting a little light through even in the static full off position. 3) ideally, your light source would have total spatial coherence (or equivalently, be a point source at infinity). Real light sources are only partially coherent and have finite size, so some photons impinge on the mirror at a small angle. When these are reflected in the off state, some small amount of light will still go through the exit pupil. 4) the light dump material is chosen to be as black as possible and actively cooled from behind. However in practice, a true black body does not exist, so there will be a tiny amount of reflection which will bounce around the DMD cavity. Some of this will escape through the pupil.
When I was in Chemistry grad school, I proposed using digital micro mirrors to make a Hadamard transform near infrared spectrometer. It would have some advantages over a Fourier transform device in terms of noise elimination but was wavelength limited by the sizes of the mirrors.
I didn’t actually make one. I was just required to make a research proposal as part of getting a PhD. The problem was that if you went into the infrared, the wavelengths would start getting longer than the width of the mirror, and if you went into the visible there would no longer be a noise advantage from multiplexing the signal due to the sources of noise found in visible vs infrared detectors. (It’s been over 25 years so I’m shaky on the details).
The last time I used a paper airline ticket was about 2015 when an intercontinental flight was cancelled and the service desk put me on a different flight. When they handed the paper ticket I was actually worried that it would not be honored for my connecting flight but it worked.
In 1999 I was living in New York, I drove over to New Jersey with a friend of mine that owned a movie theater, to see the Phantom Menace digitally displayed, having seen it already in a regular theater. I did have some issues with small swirling artifacts, but it seemed more like a software compression issue (additive/subtractive images) than a projector issue. My friend got us into the projection booth after the movie for a tour. We had to go through security and sign some papers, but it was worth it. What I found more interesting than the projector, was the computer system it was using, if I remember correctly, it was called a 'Pluto Box'. It had redundant drives that it could switch to incase of a problem, but you could program it a month in advance for showtimes, where it would also work the lights, sound, switch languages, and even pull back the curtain (if it had one). Seems like the norm now but was really impressive 24 years ago! =)
I’m surprised you only compared DLP to other video technologies with no mention of the 35mm (and 70mm) film that was THE standard (and only option) for cinema before DLP. 35mm was largely in use up until 2008/9 and slightly beyond.
This episode had more to do with the projector rather than the storage medium. It is a big part of the story, but maybe it is an aspect told later/elsewhere.
This leaves me to wonder if we could utilize DMD tech in an analog/digital highbred processor unit design. I think one of the prayers for optical processors is the fact it can minimize the capacitance, resistance, and magnetic issues of traditional electrical pulse based processors. I can imagine an array of these being able to do basic arithmetic on a near continuous basis. Although larger in footprint than tradition semiconductors, they may be able to reclaim some of that space with cross beam refraction within a near vacuumed package and the use of being able to cross beam paths… something traditional traces can’t do. Just something I often think about.
I've been curious for how circuits are validated in general... do you think a video about validation techniques would be good? It's one of the unique areas in the industry where most the methods are standardized in some way or another
Car manufacturers are experimenting with DMDs for beamshaping in headlights. For example, check out Mercedes-Benz's "DIGITAL LIGHT" demonstration at the Geneva motor show in 2018.
Just to be clear, when you only have digital switching of the brightness, you have to have much finer control for when to turn each pixel on an off so that the right amount of light gets through. The control circuits essentially have make adjustments thousands of times a second, so the "refresh" rate for modern LCD and DMD are in the thousands of Hz.
Around 1993 my employer, Alphatronics, made probe cards for testing IC dies while still on the wafer, and one customer was TI. A common error for the probe tester was for the lift device to fail before the probes were translated to the next die. This was generally catastrophic for the probe card and the wafer. We got one probe card back from TI with bent tips and a gray smudge on the underside. Under a microscope the smudge resolved into itty bitty glitter- someone at TI had about seven million years of bad luck coming.
I'm pleased to see that the team at TI working on the project near the time I left in 1993 were, it seems, very successful. That line about TI building a fab to make the DLP is not what happened. The DMOS IV fab had been built to produce the .5um products. First was the 16MB DRAM. The DLP development team under Hornbeck were officed in that building and their lots ran on that line, but the driving force for the fab was the DRAM.
One effect Optical MEMs and DMDs had on my life experience was I mostly stopped going to the movies - while good they don’t have the brightness I remember from film projectors. A shot of the sky used to almost hurt your eyes and now is a bright grey.
1/60 of a second is enough to perceive the flickering of the colorwheel. It was the major drawback of DLP projectors. It was noticeable especially when our eyes make saccadic movements.
wow, that eidophor is just crazy complicated. I never even imagined that theater projection was so high tech that long ago. I have known about the DLP chips for a while though, very cool tech.
The ability to project movies from digital stream rather than physical media opened up opportunities for the studios to have better control over leaking new releases before an official premiere presentation. The high profile productions are streamed into the projector’s buffer remotely from a trusted data center. Local operators cannot copy or access the movie stored in the projector, they can only play it at agreed times. The only bootlegging option is to record the new release with a camcorder. But thanks to the brightness and resolution of the DLP projector the quality of the bootleg is often good enough.
15:45 Huh??? You seem to have a very small world. Everyone i know in Asia has seen The Phantom Menace. I've watched it myself in the theaters back when it came out here in Asia.
Asianometry: Please look into doing a video on Dahon and Tern bicycles. Its may be a bit off the beaten path for you but I suspect will be a topic you might be interested in.
Since Asionimetry did such a nice series on the semiconductor world, here is another application for MEMS in the world of lithography: for some systems this method was used to shape the illumination beam. See: ASML Flexray...
I think my favourite use for optical MEMS would have to be Sam Zeloof's modification of a DLP projector for doing maskless photolithography for DIY silicon chip manufacture.
I have a Panasonic laser DLP projector. It is quite large but has great sharpness and brightness. You can get these used with few hours on the laser for $300 for a 1080p model.
Ah you just reminded me of the old DLP projection TV my parents bought way back in 2005 from Samsung... It looked good while it worked, but it was a total money pit (it's primary push was that it used the then novel mems to achieve brightness and contrast ratios that put plasma and later many early OLED to shame). It's MSRP was $5k USD parents later spent 3k in repairs, with it's eventual demise in 2011.
Today the brightest projector for cinema comes from Barco and fires up to 75.000 Lumens onto the wall. Considering the 3000-ish Lumens of a standard tabletop consumer projector, this is truely incredible. To be fair this one only usable in stationary applications, since the projector and the two separate AC units is weighing over 500kg in total. For rock-and-roll purposes and when moving often you better get your hands on a Panasonic RQ50K delivering 50.000 lumens at only 160kg including mounting hardware with no extra AC required. That still gives you, when used with two projectors, more light than from one Barco projector. And - yes they can produce 4K.
The whole concept of hinged mirrors makes me wonder if we could use this with a light bending substance like pearl to make a display similar to e-ink but not limited to black and white
I love my Samsung DLP TV from 2008 and still going strong. Never had to buy a bulb thanks to the LED lightsource. Too bad that was the end of DLP tv's. Their durability is amazing
Thanks a lot! I'm very new on this topic and interested. I know MIT made huge investments on MEMS but don't know what their outputs are despite we have close relationships with them.
What do you think about the idea, that computer devices don’t compute, but rather receive Bitmap data (per wireless connection), for every frame that the screen uses. And the “computer“ device only sends the input from the touch screen or keyboard. Combined with rf mems switches for energy efficiency. I thing this is the next big game changer.
Recipe to make a projector: 1- make enough light for the brightest pixel in the brightest scene of the film, for every pixel, all the time 2 - throw all of that away, except for the little bit you need
Never knew they won an Oscar for that Technology. As soon as I learned about DLP projectors, I had to have one. What an incredible and ingenious technology. My home theater projector technology of choice is DLP. I have 3 single chip, lamp-based ones that use high-speed color wheels that produce nearly zero rainbow, and one, single chip, LED based one with no color wheel, and no discernible rainbow at all. Early single chip DLP projectors were pretty bad with rainbow, but the latest ones (home theater level, anyway) have eliminated that issue. Plus the addition of frame interpolation and 3d make the experience even better. I prefer single chip projectors over 3 chip. Single chip means that there are fewer optical elements in the light path, making it much simpler with no need to converge anything. Plus if it used an RGB, LED light source, it runs much cooler, and therefore quieter, and eliminates the need for a color wheel. Currently all my projectors are 2K. They are bright, vivid, and sharp. Even siting 8 feet from a 150" screen, I can not discern the pixels. Maybe there is a 4K one in my future. No rush. I'm quite satisfied.
![Lp. Сердце Вселенной #60 РОЖДЕНИЕ ЛОЛОЛОШКИ [Финал] • Майнкрафт](http://i.ytimg.com/vi/YoR0pAV9FVQ/mqdefault.jpg)
Born too soon to explore the earth
Born too early to explore the galaxy
Born just in time to experience dank MEMS
Ok. I think I laughed too hard at this. That was a good one.
Not too early to explore the Earth.
Still plenty buried from times past both fossilised and from ancient human history.
I'm intimately familiar with this topic because I was an early adapter of DLP projectors for home theater use.
1-chip DLP has a serious flaw, in that, at any given moment, there is only one of the primary color being projected onto the entire screen. Some people are much more sensitive to the so called "rainbow effect." Moving your head quickly can exacerbate the issue. They reduced the problem by either spinning the RGB wheel faster, or doubling up the colors into RGBRGB wheel, so that one color stays on the screen for a shorter amount of time, each.
Technically, even 3-chip DLP will have many pixels displaying just one color at some moments, but these are per-pixel effect, not whole screen effects.
Fundamentally, what is projected onto the screen is a pulse-code-modulation pixels. Each pixel is either ON or OFF at any given time, and in case of 1-chip DLP, in one of the Red/Green/Blue.
There is no grayscale. It's just ON or OFF. Each pixels blinks at incredibly fast rate, you can't normally tell, and the brightness is controlled by the rate of ON time, vs. OFF time (hence PCM). For example, 50% brightness would be represented as ON/OFF/ON/OFF/ON/OFF and so on. 1/3 brightness would be ON/OFF/OFF/ON/OFF/OFF, although they could throw in some randomness to making it look more natural to human brains.
a fun detail about the rainbow effect is that it gets significantly worse if you're not directly looking at it. The human eye is a lot more sensitive to motion at the fringes of vision (higher density of cones) than at its center and so if you look askance at it you see the flicker much more easily. Once you notice this it's really interesting how many other things you can see flicker--fluorescent lights, LCDs, and more!
@@SalmanHusainGplus Helps when you squint a bit and move your head quickly from side to side! I've always had some fun when going to the theater looking for the effect but never knew what caused it. Amazing!
OK but why would you even need a projector now that you can get a 77" OLED for $2500 and the tech has been making meaningful improvements past few years? And if you can afford a movie theatre in your house, wouldn't you go for a micro LED or something, projectors just look washed out in comparison and probably can't satisfy a modern nit whore anyway.
Yeah, I really hated 1-chip DLP projectors we had in school because of that rainbow effect. But it seems most people don't see or notice it, lucky for them.
@OlafWillocx I imagine they were able to control brightness with the original DEFORMING-Mirror-Device, but they just couldn't manufacture a device with all those hundreds of thousands (to millions) of mirrors that deformed at the same rate. I doubt it's practical even with current technology to map each pixel to different analog algorithm to get a consistent image across the whole device. DIGITAL gets rid of this problem. Slight variance among individual mirrors won't matter at all.
I also remember reading about how the molecular fatigue behavior is entirely different at nano scale. If macro scale physics numbers were applied, there's no way the hinges would last long enough to make the DMD chips viable, but they apparently don't behave that way, and are incredibly long lasting at nano scale.
The entire field of mems semiconductors is just truly incredible.
Is photonics and semiconductor same i think u are wrong
@@ggboss8502 MEMS and semiconductors are related since many MEMS devices include control circuits and they share similar manufacturing processes. Photonics is a very specific branch of light manipulation using technologies similar to semiconductors (you can create voltage controlled light switches, for example), but in general these do not employ MEMS techniques. Having said that, MEMS devices such as the DMD described in the video can be used within photonic systems but only for relatively low speed switching compared to true photonic switches.
@@ggboss8502 I think the main difference is one is mechanical, and the other is electronic? But mems also depend on semiconductors for their functioning. So I consider mems to be a subset of semiconductors, perhaps at the extreme going into nanotech.
@@ggboss8502 sorry I just realized you mentioned photonics and not necessarily MEMS or even optical MEMS tech. But consider this: the semiconductor fabrication process relies to a large extent on optical technology. It's essentially a reverse microscope. So I consider optics and semiconductor tech to be 2 sides of the same coin. It's basically at the same size for one to be manipulated by the other
Soo im wondering if projector mems technology has been considered for use with vr/ar technology? The way how the eyes works is that light is bouncing around in a room so maybe u could shoot laser inside of eyeball to make hologram? They should use this and try to make a technology that totally shoots lasers directly into your eyes and you hallucinate holograms and 4dimensional VR/AR games and slowly burn your eyes out of your skull. That sounds to me like great idea to further drive the technocratic societies attention spans down the shitter. Then no more ‘mental illness’ because everyone will be surviving in a scarce environment.
That micro hinge mechanism is fantastic and to think it's possible to mass produce something that small is bonkers.
When you make normal semiconductors mems is huge dimensionally.
Getting the hinging to work really is as difficult as you think. It was the final yield limiter to really get ironed out.
@@rydplrs71 I'm excited to see mems speakers finally break onto the scene and see a similar revolution in the audio world to match the growing mems microphone market, it's been a long time coming.
Some aspiring vendors like Australian Audiopixels have been laying it on thick claiming to be able to fully replace current loudspeakers using flat mems chip arrays - I have to say I'm skeptical but more than willing to be surprised if I could get these into my room which doesn't have anywhere near the space for a decent 5.1 speaker set up.
@@mnomadvfx there are mems miccrophones?!?!?!
yeah and then do it a billion times on one chip, then make a billion chips.
tiss magic really.
DLP stuff is always fascinating to me. I had the privilege to work at TI in the UK from the late 90s to mid 00s, although not as part of the team supporting the DLP business. The DLP guys were generous with their time in explaining stuff and TI also produced quite a few internal papers on how the technology worked and the development history. I still have my copy of the TI Technical Journal which was dedicated to this topic.
Some theaters use DLP projectors, but they don't have the contrast range that LCOS (liquid crystal on silicon) projectors do. You didn't mention the pulse width modulation used on the mirrors to generate shades of gray in each color, but it was a good overview. Thanks for all you do.
My understanding was that the contrast ratio of DLP was effectively infinite, because to generate black, the DMD just needs to dump 100% of the incoming signal into the void rather than onscreen. Are there other factors I’m missing?
@@PaulFisherere are several sources of flare in DMD. 1) the material below the mirror itself does show through a bit since it has to be exposed for a mirror to be free of its adjacent companions. While this is typically coated in a material selected to be as black as possible over the desired wavelength range, it’s not perfect. So even if 100% of light directed at the mirror is diverted, there will be some reflected light from the substrate. 2) the mirror can only tilt so far. Ideally you’d direct all the light onto a light dump orthogonal to the exit pupil, but the mirrors can’t flex 90 degrees, only about 10-12. So some light will be scattered from the mirror edges and the anchor via structure, letting a little light through even in the static full off position. 3) ideally, your light source would have total spatial coherence (or equivalently, be a point source at infinity). Real light sources are only partially coherent and have finite size, so some photons impinge on the mirror at a small angle. When these are reflected in the off state, some small amount of light will still go through the exit pupil. 4) the light dump material is chosen to be as black as possible and actively cooled from behind. However in practice, a true black body does not exist, so there will be a tiny amount of reflection which will bounce around the DMD cavity. Some of this will escape through the pupil.
@@Grak70 We had a Christie 10,000 lumen 3-chip DLP projector which was replaced by a Sony LCOS-type projector. The DLP unit had noticeably higher black level than the Sony. I don't recall the light output rating of the Sony, but it used a xenon lamp where the DLP projector used two mercury lamps - so it wasn't exactly a fair comparison. However, in addition to higher light output, the Sony had a much lower black level than the DLP unit from Christie.
@@curtlundgren6867 sounds about right. Off state black levels are a major reason DLP became obsolete as LC black levels improved and then LED off-current improved. Then came active LED, OLED, and then quantum dots with ever better black levels. There was a lot of room at the bottom for solid state light solutions, but DLP could never quite keep up. MEMS in general are much more finicky and lower yielding to fab, not to mention they face very different challenges to CMOS. TI repurposed a lot of silicon equipment to make DMDs and I think the ROI of using apples to make orange juice was difficult to justify in the long term, especially with such stiff competition from Japan, Korea and later China. Once China started pumping out decent quality LED displays that kicked even the best DLP’s ass, the margins in display evaporated.
Thank you both for the explanation!
Cool that you mention Mike! I love his channel. Highly recommend it for the electronic nerds like me.
Awesome! I asked for this topic last year! I’m so glad to see you explored this cool tech.
Always love when another underrated UA-camr I've been following for a long time gets a mention somewhere. Mikeselectricstuff always brings the goods, just like Asianometry.
The amount of time you invested to research this must be astounding... great job. I find it very odd how some companies with just one major product line that costs tens or hundreds of thousands of dollars simply fail horribly once a new product arrives, rather than trying to adapt.
If the competitor patented their tech and the out dated tech is unable to complete there's not a company can do. Even if it's not patented if the new technology is fundamentally different the old company won't have any experience with the new technology and would struggle to compete. That's why R&D is so important.
15:00 Whoever came up with this mirror configuration must have been a genius. Three beams of almost identical frequency light splitting in entirely separate directions, then coming back perfectly in line with each other. On the diagram it looks like chaos, yet every angle of that configuration is a careful design choice. How do you even approach that? Looks like magic to me.
Wow... did not know that MEMS have been around so long. Very cool. Always learning something new from watching your videos. :-)
Are you the creator of nexus wifi modem firmware "hack"?
My home theater system uses this along with a laser light source, and the video quality, the color reproduction, the resolution and the contrasts are outstanding!
You've taught me so much stuff that I wouldn't even know to ask about and you make things that seem "boring" absolutely fascinating bro.
I get so excited whenever you post a video that when I see a new thumbnail I make noises lol.
Thanks, fascinating story.
I worked with a GE Talaria projector in the early ‘80s, it gave a sharp and extremely bright image (which was essential because it was being used as a back projector in a well-lit TV studio). The thing did have drawbacks, slow to warm up, lousy colour purity and an interruption to the video feed would generate a line of bubbles in the oil which would take a couple of minutes to clear as the disc slowly rotated.
As I recall, the potential for the oil to contaminate the vacuum in the light valve was circumvented by a catalytic device that ‘cracked’ large molecules to smaller ones that could be absorbed by a getter.
I remember reading about MEMs as far back as the 80s, in EDN. Saw my first TI DLP projector, a consumer grade one, at the CES trade show, either in the late 90s, or early 2000s. I, too, once worked for a company that won an Oscar (Moviola). Though they won that Oscar decades before I was born.
I love projectors! Especially projection mapping. I learned so much and this one I had never imagine before. Super cool.
Everyone raves about those Panasonic three chip DLP laser projectors, but I’m not sure how many industry folks know the nitty-gritty like you illustrated there.
You can't imagine how much i love what you do on here. I just hope that you won't burn out yourself! Thanks a lot and take care please
Nearly all laser projectors in cinemas still use DMDs/DLP for their imaging, the lasers simply replace the xenon arc lamps as the light source and can improve contrast due to the narrow light band of the lasers
And also the ability to blank the lasers in between frames.
@@moki123g generally they dont, shutting the lasers off between frames keeps them from producing stable colours and in cinema units colour accuracy is paramount.
blanking is more something done by the DMDs themselves by simply putting all pixels to the off position and holding them there. It is used fairly extensively even with arc-lamp based units, principally its used to keep you from seeing the image go hilariously terrible while the rest of the imaging system re-adjusts for a new raster timing/resolution.
the main benefit to the lasers is their narrow colour bands means less light that doesn't go to the DMDs, which increases contrast because there is less 'unwanted' light coming out of the projector
I would have never new such a thing without your video, big tanks.
I absolutely LOVE your channel. I love how you deliver information on topics I personally know very little about in such a concise and to-the-point way. You make me want to learn as much as I can, thanks for that
Excellent presentation. I’ve been working on “light valve” computing for well over two decades. It’s particularly useful in new areas including quantum computing.
I had a single chip DLP retroprojector around year 2000. Although it was only SD, predating HD by a couple of years the image was great. If you examined it close up the pixels didn’t have black around them, unlike the plasma screens that were popular at the time.
Amazing doc, I had a series of Mitsubishi CRT based TV's in the 1980's, and the last one I had was a 65-inch and had three 9-inch glass focusing lenses. The innovation of the home versions of projection TV ran parallel in R and D and must have helped fund design of the new, huge screens we take for granted. In the 1980's, people looked at 65 inches of TV and were like, bro, you like TV, or what? Smirks sometimes followed. We took the snide comments in stride, and continued to support the early expensive tech. We were pioneers!
Wow DARPA saw this coming a mile away, now MEMs are the future for time keeping accuracy. Amazing.
Yeah. The US military spends fortunes to be the most technologically advanced organization on Earth. DARPA tends to be the first adopter for many new technologies, because the DOD cares a lot about gaining any possible advantage and has the money to singlehandedly pay for the massive development costs of new technologies that don’t have other uses yet.
15 years ago, a company called Glimmerglass made 3D MEMS capable of directing 512-1024 frequencies of light for switching in optical networks. Cost of each unit to manufacturers incorporating it into their products was over $200,000, as I recall!
I made MEMS engines for DARPA.
They didn’t work but they did lots of cool things.
I was working with Synelec and Eyevis for 25 years with DLP projectors for large format video walls for conference rooms, security centers, power plants, and so on.
DLPs are in automotive headlights now. As ridiculously expensive as they are, it's one of the most exciting modern semiconductor features in cars to me.
I just use aftermarket LED light bulbs to blind the people in front of me. The key is to use purplish and blue color which really distracts them and makes them turn their heads away from the blinding light.
@@ahndeux makes it a strobe light and the pedestrians would crawl before you
@@megalonoobiacinc4863 Anything to slow them down would work. Thanks for the strobe idea. A nice multi colored strobe led would make them crawl on their knees.
Another phenomenal video, thank you!
Been aware of this tech for a long time now, but it still baffles my brain that it actually works
Little mirrors be reflecting :D
They do be shiny tho
15:00 I would argue the more significant advantage to braking out the RGB to dedicated chips, is that you get rid of the color disintegration effect (as the colors are no longer using time based sequencing, and the time base is instead confided merely to the luminance component of the image) one experiences when you pan your eyes over a single DLP projection, where you see individual color trails (instead of the final intended composited color) streak through your vision, a very noticeable and irritating effect of single chip DLP.
Also, sometimes, commercial projectors use dedicated RGB light sources, instead of a prism, to separate out the color channels.
This was a seriously good episode! Thank you.
I rarely comment on any video. But after many videos and months following you, and after discovering some other scientific divulgation channel I must say, I appreciate your research, very well done; but I also have to say I really like your voice for narration.
Other channels, despite having a well researched topic, don't have the skill/knowleadge to make ther voice a pleasent tone to be listened.
Keep the good work, love your videos, would love to see a future video about China's nuclear power and the state of nuclear power generation in Asia overall after fukushima and russia-ukarine war
I was not aware of this tech application. Nice research!
This is my first time coming across this technology. What a nice place to learn about it! Subbed.
Love my home single chip DLP. It uses 4 color RGBY variable speed wheel, can do true 24hz and 48hz, its amazing. Only downside some rainbow effects common to wheel projectors and chip resolution is 1024x576p down-converted from source HD, looks way better than that implies though, basically cinema quality on a 2 meter (80 inch) wide screen
Cinema laser projectors are still DLPs. They just use three laser arrays for the light source rather then a xenon lamp and a prism.
There are several sources of flare in DMD. 1) the material below the mirror itself does show through a bit since it has to be exposed for a mirror to be free of its adjacent companions. While this is typically coated in a material selected to be as black as possible over the desired wavelength range, it’s not perfect. So even if 100% of light directed at the mirror is diverted, there will be some reflected light from the substrate. 2) the mirror can only tilt so far. Ideally you’d direct all the light onto a light dump orthogonal to the exit pupil, but the mirrors can’t flex 90 degrees, only about 10-12. So some light will be scattered from the mirror edges and the anchor via structure, letting a little light through even in the static full off position. 3) ideally, your light source would have total spatial coherence (or equivalently, be a point source at infinity). Real light sources are only partially coherent and have finite size, so some photons impinge on the mirror at a small angle. When these are reflected in the off state, some small amount of light will still go through the exit pupil. 4) the light dump material is chosen to be as black as possible and actively cooled from behind. However in practice, a true black body does not exist, so there will be a tiny amount of reflection which will bounce around the DMD cavity. Some of this will escape through the pupil.
When I was in Chemistry grad school, I proposed using digital micro mirrors to make a Hadamard transform near infrared spectrometer. It would have some advantages over a Fourier transform device in terms of noise elimination but was wavelength limited by the sizes of the mirrors.
You weren’t able to filter it and polarize it? I-Line︅ steppers used this principle before the light source gets to the lens.
I didn’t actually make one. I was just required to make a research proposal as part of getting a PhD.
The problem was that if you went into the infrared, the wavelengths would start getting longer than the width of the mirror, and if you went into the visible there would no longer be a noise advantage from multiplexing the signal due to the sources of noise found in visible vs infrared detectors. (It’s been over 25 years so I’m shaky on the details).
Think you mean 1/180th of a second instead of 1/60th of a second. Each color being 1/3rd of the 60Hz frame time. @14:12
The last time I used a paper airline ticket was about 2015 when an intercontinental flight was cancelled and the service desk put me on a different flight. When they handed the paper ticket I was actually worried that it would not be honored for my connecting flight but it worked.
Most IMAX projectors are still using analogue film projection.
In 1999 I was living in New York, I drove over to New Jersey with a friend of mine that owned a movie theater, to see the Phantom Menace digitally displayed, having seen it already in a regular theater. I did have some issues with small swirling artifacts, but it seemed more like a software compression issue (additive/subtractive images) than a projector issue. My friend got us into the projection booth after the movie for a tour. We had to go through security and sign some papers, but it was worth it. What I found more interesting than the projector, was the computer system it was using, if I remember correctly, it was called a 'Pluto Box'. It had redundant drives that it could switch to incase of a problem, but you could program it a month in advance for showtimes, where it would also work the lights, sound, switch languages, and even pull back the curtain (if it had one). Seems like the norm now but was really impressive 24 years ago! =)
5:00 You missed mentioning "adult films" in your list of what the Eidophor was used for (with reference to the image at 4:44)
Came here looking for MEMES but I guess MEMS is good enough.
I’m surprised you only compared DLP to other video technologies with no mention of the 35mm (and 70mm) film that was THE standard (and only option) for cinema before DLP. 35mm was largely in use up until 2008/9 and slightly beyond.
This episode had more to do with the projector rather than the storage medium. It is a big part of the story, but maybe it is an aspect told later/elsewhere.
This leaves me to wonder if we could utilize DMD tech in an analog/digital highbred processor unit design.
I think one of the prayers for optical processors is the fact it can minimize the capacitance, resistance, and magnetic issues of traditional electrical pulse based processors.
I can imagine an array of these being able to do basic arithmetic on a near continuous basis. Although larger in footprint than tradition semiconductors, they may be able to reclaim some of that space with cross beam refraction within a near vacuumed package and the use of being able to cross beam paths… something traditional traces can’t do.
Just something I often think about.
I've been curious for how circuits are validated in general... do you think a video about validation techniques would be good? It's one of the unique areas in the industry where most the methods are standardized in some way or another
Car manufacturers are experimenting with DMDs for beamshaping in headlights. For example, check out Mercedes-Benz's "DIGITAL LIGHT" demonstration at the Geneva motor show in 2018.
Just to be clear, when you only have digital switching of the brightness, you have to have much finer control for when to turn each pixel on an off so that the right amount of light gets through. The control circuits essentially have make adjustments thousands of times a second, so the "refresh" rate for modern LCD and DMD are in the thousands of Hz.
it's because of your charm that I don't mind you calling my human visual system slow and incompetent. I love your voice.
Around 1993 my employer, Alphatronics, made probe cards for testing IC dies while still on the wafer, and one customer was TI. A common error for the probe tester was for the lift device to fail before the probes were translated to the next die. This was generally catastrophic for the probe card and the wafer.
We got one probe card back from TI with bent tips and a gray smudge on the underside. Under a microscope the smudge resolved into itty bitty glitter- someone at TI had about seven million years of bad luck coming.
I'm pleased to see that the team at TI working on the project near the time I left in 1993 were, it seems, very successful. That line about TI building a fab to make the DLP is not what happened. The DMOS IV fab had been built to produce the .5um products. First was the 16MB DRAM. The DLP development team under Hornbeck were officed in that building and their lots ran on that line, but the driving force for the fab was the DRAM.
Would be nice to see a video on all the different projection systems. Especially covering Epson's 3LCD and JVC/Sony LCoS / SXRD
One effect Optical MEMs and DMDs had on my life experience was I mostly stopped going to the movies - while good they don’t have the brightness I remember from film projectors. A shot of the sky used to almost hurt your eyes and now is a bright grey.
Another excellent presentation!
Analog DMD technology was used to control the hexagon mirrors on the JWST. Crazy to think a projector was used to build the most advanced telescope
I worked for Hughes in the 90's. They had a very weird mix of businesses including DirecTV.
Another fascinating presentation, great work👍🏻👍🏻
1/60 of a second is enough to perceive the flickering of the colorwheel. It was the major drawback of DLP projectors. It was noticeable especially when our eyes make saccadic movements.
wow, that eidophor is just crazy complicated. I never even imagined that theater projection was so high tech that long ago. I have known about the DLP chips for a while though, very cool tech.
Thank you for the historic research and highly interesting presentation!
@13:26 I first thought this was a very bad AI image before realizing the faces were blurred to protect the identities
Nice shoutout for mikeselectricstuff, he is one of my favorites!
Dlp is also being used for 3d printing. It's interesting to see it circling back to printing.
It's incredible how many talented engineers there are and have been.
This tech is super impressive. Sounds like alien technology.
The ability to project movies from digital stream rather than physical media opened up opportunities for the studios to have better control over leaking new releases before an official premiere presentation. The high profile productions are streamed into the projector’s buffer remotely from a trusted data center. Local operators cannot copy or access the movie stored in the projector, they can only play it at agreed times.
The only bootlegging option is to record the new release with a camcorder. But thanks to the brightness and resolution of the DLP projector the quality of the bootleg is often good enough.
"[Reminds me] of cheez-itz, which I love"
So that's why I got that cheez-itz puff Ad, thought I was safe here! LOL
15:45 Huh??? You seem to have a very small world.
Everyone i know in Asia has seen The Phantom Menace.
I've watched it myself in the theaters back when it came out here in Asia.
Fascinating!!! I love how your videos bring light (pun intended) to things the average person would never think about! 😊
The Pilgrim Theater you show in the photograph on the left is in Boston, not NYC.
Asianometry: Please look into doing a video on Dahon and Tern bicycles. Its may be a bit off the beaten path for you but I suspect will be a topic you might be interested in.
3:35
Glad to hear we still partake in the same traditions near 100 years later.
I misread the title as Optical MEMES. Truly disappointing, but understandable.
Folks, there's a bot in the comments. Go to their channel, select «About», hit the flag and report the user for impersonation.
I just did exactly that... "using Asianometry identity for Whatsapp scamming" in additional comments before submitting... I hope that works.
Oh man , mems+LiDAR would be epic. Also what about mems+laser projection?
My half brother still has what's probably an 80" DLP TV from the late 90s. Still works perfectly but modern displays make it seem like a relic now.
Since Asionimetry did such a nice series on the semiconductor world, here is another application for MEMS in the world of lithography: for some systems this method was used to shape the illumination beam. See: ASML Flexray...
the picture at 11:12 is a CRT projector, not an LCD
I think my favourite use for optical MEMS would have to be Sam Zeloof's modification of a DLP projector for doing maskless photolithography for DIY silicon chip manufacture.
Great video, subscribed!
I have a Panasonic laser DLP projector. It is quite large but has great sharpness and brightness. You can get these used with few hours on the laser for $300 for a 1080p model.
At 11:19, when talking about LCD projectors, isn't that a CRT projector shown there?
I was looking for a comment like this as yes I believe that is a CRT projector.
Ah you just reminded me of the old DLP projection TV my parents bought way back in 2005 from Samsung... It looked good while it worked, but it was a total money pit (it's primary push was that it used the then novel mems to achieve brightness and contrast ratios that put plasma and later many early OLED to shame). It's MSRP was $5k USD parents later spent 3k in repairs, with it's eventual demise in 2011.
15 "foot" movie screen? Square "inch"? It is not the dark ages anymore, use standard units of measurement.
Really great video ! but man.. Have you read what's Pilqrim cinema projection for tonight ? xD at 4:41
Thank you for your broadcasting 👍
Help for me ☺️
Today the brightest projector for cinema comes from Barco and fires up to 75.000 Lumens onto the wall. Considering the 3000-ish Lumens of a standard tabletop consumer projector, this is truely incredible. To be fair this one only usable in stationary applications, since the projector and the two separate AC units is weighing over 500kg in total. For rock-and-roll purposes and when moving often you better get your hands on a Panasonic RQ50K delivering 50.000 lumens at only 160kg including mounting hardware with no extra AC required. That still gives you, when used with two projectors, more light than from one Barco projector. And - yes they can produce 4K.
The whole concept of hinged mirrors makes me wonder if we could use this with a light bending substance like pearl to make a display similar to e-ink but not limited to black and white
>tiny mechanical stuff on a chip
when are they going to try making the worlds smallest timekeeping device?
I love my Samsung DLP TV from 2008 and still going strong. Never had to buy a bulb thanks to the LED lightsource. Too bad that was the end of DLP tv's. Their durability is amazing
i'm studying in software engineering and your videos are increasing my interest in the computer engineering field
same, but I come from the opposite angle. I'm a physics student, started listening to Asianometry in the pandemic, and now I'm branching out to ECE
@@VedJoshi.. cool man!
you do a dam good job brother.
Thanks a lot! I'm very new on this topic and interested. I know MIT made huge investments on MEMS but don't know what their outputs are despite we have close relationships with them.
I had a 1 chip vga dlp projector in 1999. The problem in the home is always that the cost of the bulb is more than a decent LCD TV.
Not anymore with laser, though the projector is easily 10x the cost. :D
What do you think about the idea, that computer devices don’t compute, but rather receive Bitmap data (per wireless connection), for every frame that the screen uses. And the “computer“ device only sends the input from the touch screen or keyboard. Combined with rf mems switches for energy efficiency. I thing this is the next big game changer.
Recipe to make a projector:
1- make enough light for the brightest pixel in the brightest scene of the film, for every pixel, all the time
2 - throw all of that away, except for the little bit you need
Never knew they won an Oscar for that Technology. As soon as I learned about DLP projectors, I had to have one. What an incredible and ingenious technology. My home theater projector technology of choice is DLP. I have 3 single chip, lamp-based ones that use high-speed color wheels that produce nearly zero rainbow, and one, single chip, LED based one with no color wheel, and no discernible rainbow at all. Early single chip DLP projectors were pretty bad with rainbow, but the latest ones (home theater level, anyway) have eliminated that issue. Plus the addition of frame interpolation and 3d make the experience even better. I prefer single chip projectors over 3 chip. Single chip means that there are fewer optical elements in the light path, making it much simpler with no need to converge anything. Plus if it used an RGB, LED light source, it runs much cooler, and therefore quieter, and eliminates the need for a color wheel. Currently all my projectors are 2K. They are bright, vivid, and sharp. Even siting 8 feet from a 150" screen, I can not discern the pixels. Maybe there is a 4K one in my future. No rush. I'm quite satisfied.
I loved our DLP rear projectionhd tv. Sad that went away as a product