Tuesday, December 31, 2019

IEDM 2019: Sony 3-Layer Organic+Si Sensor

Sony IEDM 2019 paper "Three-layer Stacked Color Image Sensor With 2.0-μm Pixel Size Using Organic Photoconductive Film" by H. Togashi, T. Watanabe, M. Joei, T. Hayashi, S. Hirata, S. Fukuoka, Y. Ando, Y. Sato, J. Yamamoto, I. Yagi, M. Murata, M. Kuribayashi, F. Koga, T. Yamaguchi, Y. Oike, T. Ezaki, and T. Hirayama combines 3T organic PD pixel with 4T Si-based pixel ideas:

"A three-layer stacked color image sensor was formed using an organic film. The sensor decreases the false color problem as it does not require demosaicing. Furthermore, with the 2.0-μm pixel image sensor, improved spectral characteristics owing to green adsorption by the organic film above the red/blue photodiode, were successfully demonstrated."

6 comments:

  1. Certainly not useable for capturing skin tones.

    ReplyDelete
  2. The red channel would need some help.

    ReplyDelete
  3. The pipe suddenly triples in bandwidth, try running an existing 12MP 120FPS in this mode to your ISP. Due to space constraint, I don't see how equivalent full well capacity can be maintained between the different color channels. You'd have wildly different channel full well capacity which can become a big problem for IQ teams. I don't think each channel would have a separate integration time, thus saturation will be a problem for low full well capacity channels.

    ReplyDelete
  4. Ups. I have a deja vú. Looks a bit like the classic FOVEON sensor idea. Vertical photodiodes. With a lot of crosstalk. Like here. Maybe suitable for traffic light and sign recognition.

    ReplyDelete
    Replies
    1. Albert Theuwissen - Harvest ImagingJanuary 3, 2020 at 8:53 PM

      Hi Dana, I think there are quite some differences between this Sony solution and Foveon's solution. I will explain all details in my workshop at Image Sensors Europe in London coming March. I know you will be there as well, so just join the workshop.

      Delete
    2. Yeah, Foveon had all 3 diodes 3T, whereas here blue and red are 4T and only green is 3T. Still, the absorption characteristics are odd. Green going up at blue again. Blue and Red having terrible QE, red being essentially flat over 600 nm and having another local maxima in green... good luck in color correction and color noise

      Delete

All comments are moderated to avoid spam and personal attacks.