% | $
Quotes you view appear here for quick access.

Tessera Technologies Inc. Message Board

  • jacosa jacosa Nov 3, 2010 12:31 PM Flag

    Optics thoughts

    I think I can guess what made the WLO business unattractive. It comes down to image quality. The quality of a captured image is a combination of number of pixels, number of brightness values resolved per pixel and the accuracy with which the recorded values track the actual brightness. For a constant wafer technology, quality is proportional to area. The kinds of technological improvement that allow increased image quality from a given area of silicon are a little different from the kinds of improvements that allow increased device density. Tessera put a lot of effort into getting out a very small VGA-grade chip-level camera (2-3 megapixels; 256 levels per color, 1% accuracy). The implicit hope was that there would be a market for such a thing, and that IC technology would develop in a way that permitted easy improvement to consumer photo grade (8-12 megapixels, 256 levels, 0.5% accuracy)(some cell phones are now in this range). What has happened, apparently, is that the VGA-level cameras are a cheap commodity hard to sell at a profit, and the desirable resolution level still requires so much sensor area that Tessera's clever miniaturization techniques are unnecessary.

    One thing that disappoints me is that I was really hoping to see compound-eye consumer cameras (say six 1/4 square inch sensors having their signals combined by the processor). Could have sold a lot of units, but I guess the technology ain't ready.

    About the value of the Shellcase IP: the way I read the white papers, some of the technology used to put all the electrical connections to the camera module on the non-photosensitive side could be seen as alternatives to parts of the microPILR technology package. That is the company's crown jewels, so owning related stuff is worth something, even if it is mothballed.

    SortNewest  |  Oldest  |  Most Replied Expand all replies
    • you finally got a reply to all of your concerns. They have wiped out the whole optic field starting from the director who took care of it...
      To be honest, I have sold all my shares at pair value more less, the risk for a negative outcome in the various courts its too much of a risk for me. The optic biz which should have been the propellent for the new generation revenues after several patent expire has failed. I will consider reinvesting in TSRA only on the low teens

      • 1 Reply to giomesc
      • Ok, I just want a bit of landscape. I'm pretty sure "x" is the same as tesseriana. Are you? I ask only because of the ESL writing style.

        The new optical business isn't gone; EDOF and zoom remain at a hardware level and the whole software package is still there.

        There was never a prospect that optics would be more important to Tessera's future than microPILR. The drive to higher connect densities seems to have resumed after a recession pause. We'll get answers on that over the next 18 months.

    • yi misunderstood how it work. my 3MP phone camera tke very clear and good pitures, as good as my big camera. i heard that thy just couldnot make it.

      • 1 Reply to xtessera
      • "yi misunderstood how it work. my 3MP phone camera tke very clear and good pitures, as good as my big camera."

        You can't be serious about pictures. A lot of us are. Comparing the pictures from any last-generation phone camera to those from the mid-consumer-range Canon Powershot [for instance] is a bad joke. And in terms of information content it should be: 3 million pixels with 21 bits of information per pixel (24 bits are saved, but the least significant one for each color is pretty random) gives 63 million bits of information; 8 million pixels times 24 bits per pixel gives more than 3x as much information.

        The observation I started from is that despite enormous demand for better-than-consumer cameras with the consumer-size image sensor (1/2.5 inch), those cameras remain scarce and expensive. While sensor size for a given quality has shrunk maybe 20%, regular chip size for constant processing power has shrunk about 90% (and tolerable surface area for a phone design has increased). Furthermore, one of the key advantages of Shellcase is avoiding "dust" on the sensor. Small chips can be ruined by a bad spot; larger chips aren't ruined, and the errors caused by dust can be ameliorated by...added power in the processor.

        So physics / engineering / consumer preferences have gone the wrong way for the chip-scale camera. The physics may have been predictable, but the other two weren't.

    • Thanks jacosa, as always your input is knowledgeable, helpful and appreciated.

      - Rummy

32.74+0.37(+1.14%)Nov 27 1:00 PMEST