CES 2012 Pico Projector Overview

As part of my marathon training, I ran 18 miles the Sunday before CES and it turned out to also be good practice for attending CES.   I’d estimate I averaged over 4 miles walking the floor and between venues (it was faster to walk the mile to the Venetian than take a bus at busy times of day) plus my morning 3 mile jog.   For this post, I’m going to give some quick highlights of what I saw about pico projectors at CES.   I plan on writing in more detail about some of these items in in the near future.

Over half of the show hours I was in private meetings that I can’t talk about, but I did get a chance to see and hear about a number of pico projector related activities that are public.   I can’t hope to compete with the many people that give you the quick and glossy news of CES that mostly just repeat the company talking points, but as you should come to expect from me, I will be doings some more in-depth analysis with an engineer’s eye of the products.

QP Optoelectronics introduced their “Lightpad” product at CES.   It interfaces to smartphones with an HDMI output and combines a keyboard, DLP WVGA (848×480 pixel) pico projector, rear projection screen, and battery that easily folds up into a thin and light form factor.

While it is not perfect yet, there is a lot to like about the basic concept and they said they got a lot of interest at CES.   It at least starts to address some of the issues with “use model” that I have written about earlier.  I am working on an article that talks about the good and bad points of this concept and where I see this type of product  going in the future.

Syndiant’s biggest news was their formal announcement of the SYL2271 720P 0.31” diagonal LCOS microdisplay and its accompanying SYA1231 ASIC.   Shown at left is an actual picture of the SYL2271 that has been pasted into some cute artwork.  The Syndiant had three SYL2271 720P projectors running in their private suite all showing 720p HD movie content.  All of the optical engines were very much “prototypes” with some optical quality issues and not near production ready.

Syndiant also jointly announced Viewlink’s new Vizcom™ Wi-Fi Cloud-Connected Near-Eye Visual Communication System.  The VizCom system includes a wearable heads-up display with integrated 720p video camera and an AndroidTM smart controller.  VizCom allows content to be streamed directly to the cloud via built-in Wi-Fi or by 3G/4G wireless smartphones, tablets or cellular hotspots. The Syndiant SYL2010 SVGA (800×600 pixel) panel acts as a camera viewfinder and as a display.  There was a working prototype of the display but not the overall product in Syndiant’s suite.   The optical quality of the prototype optics left something to be desired but the mechanical workings of the headset seemed to be very workable compared to other near eye products I have used.

Syndiant had a demo of a 160 lumen 3-D passive glasses pico projector that used two SYL2061’s with a single projection lens in a light engine designed by ASTRI.   The projector would either present 80 lumens to each eye in 3-D mode or 160 lumens to both eyes in 2-D mode.

A number of Syndiant pico projector products were filling about half of 3M’s booth at CES.   There were several more conventional pico projectors like the older MP160 and MP180 plus a new SYL2061 WSVGA (1024×600) based MP220 with 50 lumens.

Additionally 3M was showing a new “Camcorder Projector,” the CP40, which combines a handheld video camcorder with an SVGA pico projector.

Syndiant based products could also be found at AAXA’s and WSOT’s booths at CES and I expect some other places that I may have missed.  AAXA was demonstrating a new projector based on Syndiant SYL2061 panel.   WSOT has a dual panel WSVGA 3-D passive glasses projector similar to the one at Syndiant’s suite.   They also had a demonstration of prototype projector with a 4cc light engine based on Syndiant SYL2030 WVGA (854×480) device.

TI’s DLP certainly had by far the biggest presence of any of the pico projector display makers; although most of the newer products probably should be called “mini” rather than “pico” projectors.   There were a number products based around their WXGA (1280×800) 0.44” panel with products that were from 1.3-inches to over 2 inches thick.  These products were clearly aimed more at business professionals to put in their briefcases and had marketing spec’s of 200, 300, and some with 500 lumens (note these are often their “marketing lumens” which often are inflated by 1.2X to nearly 2X depending on the brand).

All of these WXGA projectors were really designed for wall plug rather than battery operation and have no internal batteries.  But Vivitek did find a way to make their battery powered by adding large external battery packs.   Essentially these battery packs have DC power cord to plug into the DC jack normally used by the AC wall plug power pack.

There could also be found a number of very similar looking WVGA (848×480) DLP pico projectors at the various booths around the show with light outputs ranging from about 30 lumens to as much as 80 lumens.  Most of these projectors include internal batteries.

DLP Diamond Pixel Arrangement

Both the WVGA and WXGA projectors use what is known as “Diamond Pixels” in which the DLP mirrors are rotated 45 degrees in a tile like arrangement show at the left.  This is done to reduce the thickness of the optics (a complex discussion for another day).

The re-sampling/scaling of the image from a normal square pixel grid to the diamond grid  does have a negative impact with high-resolution computer content.  Click on the thumbnail on the right to see the effects of the diamond pixel scaling on a high-resolution test pattern.

A notable exception to the bigger and brighter DLP projectors and much more of a “true” pico projector was used in Sony’s lineup of 4 camcorder models with pico projectors build into backs of the flip-out LCDs monitors.  These projectors used DLP’s 0.22” diagonal nHD (one-ninth 1080p or 640×360 pixels).   It seems to me to be a mismatch to combine a 1080i camcorder with a pico projector that has 1/9th the pixels.

I was told my multiple companies at CES that TI has a major campaign to get all the makers of LCOS pico projectors to carry at least one DLP based projector.  TI provided all kinds of support to get the projector companies to have at least one DLP product and to a large degree they succeeded with companies including 3M and AAXA showing DLP products along with their LCOS projectors.

Microvision "720P" (click on image)

Microvision was showing a new “so called 720p” multimedia projector at CES.  I say “so called 720p” because they would only demonstrate low resolution cartoon like video games on it.  I did ask them to put up a test pattern to show that they really could do 720p (1280×720) resolution but they politely refused.   My engineering instinct is that if someone is claiming HD resolution, they would be showing off HD content.   I also noticed that the 720p projector seems to be off whenever they were not demonstrating it to someone which suggests that there may be some laser lifetime and/or heating issues with the device.

The prototype media player projector was to me surprising large considering they have been claiming the whole PicoP® concept to be aimed at embedded products.  While the light engine optics itself is about 4cc, by the time you add all the electronics and a very large heat sink/heat spreader underneath the projection engine, about 25cc (56mm x 38mm x 12mm) within the media player are consumed (click on the picture above that shows some of the dimension).  Imagine how much bigger still it would be if had to add the cell phone engine and its LCD/OLED display to the package.  Compared to DLP and LCOS projection engines, there seems to be a large amount of electronics associated with LBS.

The same week as CES, Microvision put out flyer with set of partial spec’s on the PicoP engine itself (less any of the media player features).    To a degree, the spec sheet confirms some serious issues with the whole laser beam scanning (LBS) concept that Microvision uses.  The flyer says that at 15 lumens it will be a Class 2 laser product, but in a footnote it admits that the 25 lumen version would be “Class 3R” confirming what I (and others) have said for years about the issues with laser safety standards with LBS.  Note, the cell phone makers have told me that they wouldn’t put anything beyond Class 1 (considered totally eye safe) into a consumer cell phone and LBS type displays would support less than 1 lumen at Class 1; so even the Class 2 rating at 15 lumens I would consider to be a serious problem.

Another interesting indirect admission in the “spec” is that they consume “Approximately 2.0 Watts” at “27% video.”    It seems like a bad job of trying to hide a power problem.  It begs several questions, most obviously, what is the power consumption at some rated (measured) lumens.  If we assume it is for their 15 lumen projector and simply scale up we get over 7 Watts!   To get a realistic power consumption we have to know how “approximately” the power consumption number is and what it covers in the system.   As I wrote previously about the ShowWX power consumption, they seem to be a long way from their power “goals” to fit in an embedded product.

Another little tidbit from the “spec” is that it only has 16-bits per pixel (64K colors which means they have only 6 bits two primary colors and 5 bits of the third primary).  Most products today have at least 24-bits per pixel (8 bits each of red, green, and blue) = 16 Million colors.   This suggests some limitation in the ability to control the colors with their system.

I will have some more comments on the Microvision 720p as well as their 3-D and hand tracking demonstrations in an upcoming article.

Vuzix Holographic Optics

Vuzix was demonstrating an interesting technology for near eye heads up displays.  They have holograms embedded in a thin piece of plastic that can bend the output of a projector 90 degrees, translate and expand it, bend it back 90 degrees, and have it focused at infinity (so your eyes can stay on the real world).

I didn’t get the best picture of it on the above (it is kind of tricky and I didn’t have much time) but it is impressive how they can manipulate the light using hologram light guides.   While the image is in focus and would seem to be acceptable the intended purpose of a near eye HUD/augmented reality display, the image quality is not what you would want for say watching a movie.  Everything seems to have a “glow” to it which I suspect come from the contortions that are done to the light by the holograms.

That’s it for the “overview.”  Certainly my coverage of CES was spotty and if anything I didn’t give a lot of coverage to DLP relative to the number of products that were at the show.  If you have questions or want more details on some subject, please ask.

21 comments

  1. me_wwwing says:

    MVIS specs for the whole projector – Module (W × L × H) 35 × 65 × 6.1mm Max

    for the best image of this – http://www.pbase.com/me_wwwing/image/140957627

    25lm will be used for the auto HUDS. so 20-24 lm for the projector seems to be ok.

    i’m concerned about the lasers life time. we need more info.

    DLP has a lot of chromatic aberration. or what ever the purple highlight around the edges are.

    Karl, did you see the image’s pincushion on the 720p?

    TIA

  2. me_wwwing says:

    Karl, how can a heat spreader be under the IPM, when the IPM is attached with 3 screws to a PCB?

    • admin says:

      Unfortunately, I don’t have a picture of the underside of the projector, but you can get a clue from the side view of the case. Note in the side view I have marked off the 6mm that Microvision “counts” and shown that there is at lease another 6mm plus the top and bottom of the package (so the whole module is about 16mm thick including the case).

  3. Bill K. says:

    Thanks for the Projector Review from CES karl. The indepth nature of your review regarding Microvision Equipment as compared to the other projectors you listed seems to indicate you may have a bit of a bias toward Microvision. Perhaps you should have listed the PicoP at the bottom of your list with the heading “the pico projector I hated most at CES”. [sentence containing explective delete by moderator]

    • admin says:

      The problem with Microvision is that they say so much that is technically misleading or inaccurate that there is a lot of write about and correct. I would have been happy to put up an objective test pattern, take a picture of it and post it, but they only would show low resolution content on their so called 720p device. They claimed to have this really small engine, but when you add in all the electronics and heat spreaders/sinks they need, they are actually bigger than the other technologies. They claim to have the most efficient technology, but when you measure it, they are the least efficient. They have waffled for years about what class of laser safety they would be, but when they announce a product that will be just 25 lumens, they finally admit to being class 3R.

      I get the sense that you are either a Microvision employee or someone that has invested in Microvision. Would you mind stating your interest?

      Is there something you find technically/factually wrong with what I wrote? If so I would be happy to address the issue.

  4. Pete says:

    Karl,

    Did you get a chance to look at samsung’s oled tvs?

    Thanks.

    • admin says:

      Yes, I took a look at both Samsung’s and LG’s OLED TV’s.

      They certainly look good, but the rumor is that the LG is going to cost about $8K. The LG didn’t have an antireflective coating while this gave blacker blacks, it caused a lot of problems with reflections. I didn’t notice the reflection issues with the Samsung. The big advantages over LCDs is the blacker blacks and the virtually unlimited viewing angle and about half the power consumption (the Samsung might be slightly better in terms of power).

      The big technical difference is that LG is using a “white” OLED with color filters where Samsung uses color OLEDs. The LG technology is billed as being more stable since there is a known issue with color OLEDs aging at different rates by color so the color shifts over time. Personally, I didn’t think the LG’s reds looked that saturated but it could have been the content or the viewing environment.

      My overall recommendation to a consumer is would be to wait and see. I think the technology will have to come down in cost fast. I would be a bit more worried about the multi-color OLED approach until it is proven. While OLEDs are working well in cell phones you cell phone display isn’t on 60+ hours a week like some people use their television. The TVs are probably better overall than a good LCD TV, but I think it will be a big premium relative to the improvement in image quality.

      I did upload a couple of videos (click here) I took at the show of both the LG and Samsung OLEDs. It was a bit tricky to take as the camera, a Rebel 3ti, had 18 megapixels and would alias/moire with the OLED’s pixels. I had to very slightly defocus the camera to work as an anti-aliasing filter.

  5. me_wwwing says:

    Karl, there is no reply link in your last post with me. soooo

    >No, Microvision has not said at what point they go from Class 2 to Class 3R.

    Microvision gave us the 25 lm Class 3r rating asterisk. to me if they needed to put an asterisk at 20lm if warranted.
    i think MVIS needs to wait on the FDA before its marketed.

    > patents – i use them to guess at what MVIS can or can’t do. i try to be reasonable by guessing the form and expense of the design.
    since the 2 new Ap Pat came out yesterday. there are now another ways MVIS can correct for some pincushion.
    the date patents are filed give me a hint as to how long they have been working with the ideas. the lens correcting was dated 2010.

    >Microvision made a big deal about working with a Motorola R&D group looking at cell phone applications

    the projector it self is the application to work with a cell phone

    > windshield – i worked 28 yrs in GM’s Engineering Center. i’m well aware of how liabilities are looked are inside an auto company.
    as far as what i know.

    >windshield would act as a mirror

    that’s my point. the loss of light intensity and diffraction of the glass plays into the safety concerns.

    in the next 8-12 months we are going to see what is what in MVIS’s future.

  6. Dave says:

    Is this Microvision part still FPGA-based? This would significantly contribute to both the size and the power requirements.

    • admin says:

      The ShowWX was FPGA based but that would only explain away a small percentage (about 10% to 15%) of the power. Microvision requires a lot of electronics. Pictured below is one of two board that drives their so called “720p” projector. Notice there are two large ASICs (one probably for the mirror control and the other for laser control) plus a small ACTEL (FPGA company) device. There are a number of inductors on and other chips on the board suggestive of DC to DC voltage converters for all the chips.
      Picture of the “720p” electronics:

      LCOS and DLP does everything Microvision does and more with one small ASIC. After a number of years of spending over $30M/year, this is as small as Microvision has gotten the control electronics.

      I believe the second board (not show in the picture below linked to above but was in the ShowWX and is evident as part of a two board sandwich in the pictures 1 to 4 at Engadget of the clear case) has the high power laser driver controller and associated components.

      Most of the power is spent in the active control of the laser drivers. They have to actively analog control the power to the lasers in order to compensate for the time varying speed of the laser scan process and to control the brightness of each pixel. The high speed analog drivers don’t so much turn the current on or off, but rather steer it to the lasers to to the power rail.

      A number of people including mysel have measured the ShowWX and its variants, an the answer keeps coming to about 5W for 10 to 15 lumens. Their latest “720p” spec says they will (I believe this “spec” to be future tense) take “approximately 2.0 Watts” at 27% Video for a 15 lumen projector which works out to about 2.5 lumens per Watt. No matter which way you slice it, their solution is not very efficient in a world were LED based pico projectors are getting about 7 to 12 lumens per Watt (and those technologies will be much more efficient with lasers).

  7. Kyler Thompson says:

    Karl,

    Interesting summary review. A couple miss matches with your assessment is that I noticed DLP boasted their 720p as pico, yet only displayed their panel. When the design requires the optical bench, light sources and fan to function. Thus making the assembly “Mini” as you referenced.

    I also stopped by the MicroVision suite. They were playing HD content and I managed to convince them to play HD content form my Motorola Droid. It functioned without a hitch, very clear and concise content. I can’t imagine any professional allowing a random test patterned for evaluation to be projected when they don’t know its pixel level structure. I also didn’t see any “cartoon” type content being projected, maybe your visit was an early visit to their suite?

    • admin says:

      DLP can call their 1280×800 panel a “pico” but it has an 0.45″ diagonal which is not small enough for any serious embedded applications. The device is finding its way into “mini” projectors with 200 to 500 (claimed) lumens. Anything with these kinds of light outputs is going to need a fan.

      I visited the Microvision suite on the last day of CES. I certainly believe that the Microvision 720p “worked” with HD content, but I’m also sure that does not have 720p resolution. Just because it “plays” 720p content, does not mean that it has 720p resolution. There are a lot of problems caused by their scanning process not matching that of the normal pixel grid. There are additional problems in color depth (please don’t confused color “depth” with “gamut”) caused by the way they have to control the lasers in the scanning process.

      • Kyler says:

        BTW, I’ve noticed your blog discusses an overwhelming amount of information that is not important to the consumer. As an example, everyone knows that the industry leader in consumer sales of LCD TVs is Visio. Many industry professionals not unlike yourself would never purchase an LCD with a contrast ration of 1000, especially when you can obtain significantly higher contrast ratios from Samsung or Sharp. Yet, a Visio LCD TV only has 1000. What is important to technologists is not necessarily important to the consumer. Organizations like SID are commonly referred to as crowing contests. Very similar to an egg factory filled with roosters. Lots of noise, but no eggs. This is also referred to as the technologists paradox, and a well understood reason why researchers rarely make good business decisions.

        When you describe a 720p sign wave format for laser projection and attempt to discredit it using a camera that could never capture the image quality (I’ve seen many technologists like yourself take pictures of displays with 1000000:1 contrast ratios sing their digital camera, very amusing), your data and assumptions could easily be picked apart. When a laser is turned off and on in a pattern of 2080×720, it doesn’t matter if it was done using a saw tooth, sign, or square wave. Pixels are pixels, on and off still remains, on and off.

        As for class level; consumers don’t care, they simply want a brighter image. Next time you pick up a laser pointer, it will have a sticker on it and only technologists will read the class level. At class 3, a consumer will have to look into the laser from 4 inches away and hold it for 4 hours to cause any permanent injury. Batteries on these devices don’t last that long and it’s too easy to place a proximity sensor that would shut down the laser in such an event.

        Many technologists miss the forest because the trees are in the way. Most investors also commonly fall into such traps as well. It’s one of the contributors of why the success rate for investors has dropped from the 1 in 7 in the 80’s to the 1 in 30 of today.

        Thanks for your time…..

        KT

        • admin says:

          I get the “consumers are dumb and will buy anything if it is marketed well” concept. But maybe what you don’t understand, is that there are many fatal flaws in the whole laser beam scanning (LBS) technology. It not just one thing that LBS does very poorly, it is many things that are tangible and measurable. What you have to understand is that the end equipment companies have engineers that are comparing all the technologies against each other.

          Look you can input HD content into devices that don’t have that many pixels and many people will not know the difference. Why even talk about resolution then? I’m not writing for “end consumers” as much as I am writing for people interested in technology. I have shown the various technologies, how I measured them, and how people who are interested in technology can repeat the experiments.

          As for laser class safety, that is one of the FIRST questions I got asked when the subject of laser projection comes up. I have been flat told by major cell phone companies that they won’t want to mess with anything other than class 1 laser safe products. They are dealing with selling products to consumers in a litigious society. They don’t want to risk shipping millions of something only to get forced into a recall of it all. Why is a “real company” going to go with a class 2 or 3R product when there is a class 1 alternative?

          As for “picking apart” my data and assumptions, please be specific. As it stands, you just seem to be saying that measuring anything is a waste of time. You may not be interested but the engineers and managers at companies making decisions to build real products (I don’t consider things like the ShowWX to be real products) have to base their decisions on things that can be measured. They run “shootouts” of technologies to see in which direction they should go.

          As far as “subjective analysis” it should tell you something that so far Microvision has had to pay to make everything themselves and sell at a loss. The best they have been able to do so far is to get some companies such as Uniden Australia (and now maybe Wowee) to rebrand something Microvision made.

          Some of the issue with the laser beam steering claims and then measuring them include:

          1) It has WVGA or and now HD resolution — but by any serious measurement it doesn’t
          2) It is low in power — but if you measure it, it has the worst lumens per Watt of any of the available pico projector technologies
          3) It is smallest — but if you include the electronics and the dissipation of heat due to the higher power it is bigger than the other technologies for a given resolution
          4) It is the only focus free technology — but LCOS and DLP with laser illumination are focus free.
          5) It is cost effective — but it costs much more than DLP and LCOS
          6) DGL are going to be cost effective in 2012 or 2013 — they aren’t

          Then you add to this things like the issues with laser safety and it should be clear why LBS is a technology looking for a solution rather than a viable product technology.

          Finally you make up the reason why startups are having such a hard time today and blame it on your “technologist” straw-man without any evidence. I know many of the reasons why it is so hard in the U.S. today for startups and none of them have to do with your technologist straw-man. One of the biggest is that it cost about 10X as much to go public in the U.S. as it does in much of Asia. It basically cuts off U.S. small startups from being able to go public. If they can’t show a reasonable case for going public, then raising capital is VERY difficult. Venture investors look at “buyouts” by a larger company as a consolation prize with a typical 2X to 3X return, but in the U.S. they are pretty much cut off from the “big win” of a 10X return of going public. Venture investors in the U.S. used to expect a mix of IPO’s with 10X return, buyouts with 2X to 3X return, and some failures with 0 return; but since the U.S. government/regulations has so severely cut out the 10X IPO return since the year 2000, they can get the mix to to work.

Leave a Reply to admin Cancel reply

Your email address will not be published. Required fields are marked *