Eric Fossum

Sorry, RGBW already means something - R,G,B and "white" aka "clear"

This looks like a sensor issue - some low level row noise, probably from coupling somewhere in the circuit design.  Unusual for Sony, but maybe you are pushing the performance and it is showing ...

I attended the ULITIMA meeting at Argonne National Laboratory last week which was focused on ultra high speed imaging. The slides from the talks are available. I am linking to Vlad's post on his ...

FYI from the 1990's: Design and operation of self-biased high-gain amplifier arrays for photon-counting sensors

I heard this idea from a friend in about 1992 and talked him out of it as a bad idea for visible light sensors.  It is still a bad idea in many ways and I think MIT/LL can only use this idea with ...

Well, you are forgetting several important things.  1.  The voltage from a single electron has to be large enough to trigger a continuous-time comparator, with low BER (meaning very low noise).  ...

Thanks Dan. I know Shoji well but I did not see this paper. It is a pretty good review paper although he misses some early work. I appreciate your bringing it to my attention. I don't agree with ...

Richard, I think you are correct. BTW, here is one of my favorite papers which dates from 2001. We used two different signal chains with different gains, and two different sensitivity ...

BTW, in the QIS patent specification (circa 2005) it is mentioned that color filter can cover a single jot or a group of jots. This is because it can reduce color cross talk and also make it easier ...

meh, Honestly, nothing in here of interest.  And there seems to be a disconnect between figures and captions - if not intentional then really strange. Worst case was the TCAD charge transfer ...

The QIS concept (aka DFS) started when my friend Junichi Nakamura was planning his book on Image Sensors and Signal Processing for Digital Still Cameras . He asked me to write the last chapter on ...

The improved link is now edited as above. Thanks Dan for posting this post. The IEEE Spectrum Chip Hall of Fame is something all of us former Photobit-ers feel quite good about.  I feel bad about ...

I am not sure I would put it that way, unless there is a baby-development joke in there. She worked at JPL pretty much up until our daughter was born. Then a couple of months after that, while on ...

I have written to LDV Capital to ask why they deleted or privatized the video. Stay tuned.

Me too. Seem like photon shot noise impact on IQ is quite different if the photon signal is all in one pixel (and the rest of the image is black) vs. spread out at low level across the image. PS. ...

If the dark current were the same in every pixel every time then it would NOT be considered noise, just sort of an offset. In an image it would appear as a sort of fog if the offset were not ...

The so-called doubling temperature depends on the device, and either goes as the bandgap or half the bandgap.  Half the bandgap is fairly common which gives a doubling temperature of 10C.  Full ...

Yes, this is due to white noise in the CCD source-follower transistor, whose total impact can be reduced when operating at lower frequencies.

This is not really a sensor temperature issue (glow is not due to a hot amplifier circuit). The glow is due to collision of transistor electrons with the silicon lattice, leading to near-band gap ...

This is to reduce dark current to allow very long exposure periods. This is different from  readout noise, which is mostly unsupressed 1/f noise from the in-pixel source follower. There is a ...