New Camera Sensor Eliminates Need for Flash

[dilbert]

Some sort of external light is always needed.  I'm hoping that the sensor becomes practical, but, there are
I'm also wondering about the property that he quoters of holding light far longer.  Actually, we want sensors that will reset more quickly so that we get more frames per second.

Well that will definitely help the numbers of cat photos on the Internet swell.

[dilbert]

Sounds fishy to me. The current sensors have about 50% QE. This means that one can only increase it twice, not 1,000 times.

In principle, "sensitivity" is not a well defined term when it comes to digital sensors. They are photon counting machines, missing every other photon, roughly speaking.

Yes, your thinking mirrors mine.  Perhaps its a deeper electron well that holds more photons, but that implies a longer exposure.  The description of "Wang said the key to his new sensor is the use of "light-trapping" nanostructures that use graphene as a base. The nanostructures hold onto light-generated electron particles for much longer than conventional sensors." is a bit vague and does not explain the predicted usefulness for consumer imaging. 

Imagine that the electron well actually has holes in it and that not all of the electrons that you capture stay in there - some get out or some get absorbed by the well.. Giving the well better walls/floor will help prevent that and thus result in a more consistent image being presented.

[brad-man]

Sounds very interesting to me, let's hope it leads to practical sensor innovation. Just because it may eliminate the need for flash, doesn't mean you can't use a flash for fill, highlights, etc. Hopefully a much smaller and less powerful one.

[AprilForever]

Some sort of external light is always needed.  I'm hoping that the sensor becomes practical, but, there are
I'm also wondering about the property that he quoters of holding light far longer.  Actually, we want sensors that will reset more quickly so that we get more frames per second.

Well that will definitely help the numbers of cat photos on the Internet swell.

Dilbert's Law of Photographic Technology: The eventual ramification of every technological advance in photography will be the proliferation of scatological images throughout the internet, predominantly of felines.

[Pi]

Will 51,200 or 102,400 iso soon be the new 800? You'd better believe it.

You'd better not. The shot noise is physics reality. If you keep the Bayer design, there is about 1 stop left for improvement. Then you need a quantum perpetuum mobile for more. Again, the shot noise is not due to some deficiency of the current sensors - it is a part of the image itself, like it or not.

Removing the Bayer design would add another stop or so, roughly speaking. And that's it.  There is a hard limit, which no human ingenuity can overcome. Like the speed of light, or the energy conservation laws. Unless some new Einstein comes and shakes everything we know about physics, there is not much room for improvement.

There is a lot of room for improvement however in the read noise (but below some level not so practical anymore), and long term read noise. But you do not need new inventions for the former, the dark side had it for years. The read noise affects the image mostly into the deep shadows but also creeps into the midtones as well.

[Mt Spokane Photography]

I've started to search for more technical detail as to how a sensor like this might work.  Nokia recently filed for a patent for a graphine based sensor. 
 
http://appft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.html&r=1&f=G&l=50&s1=%2220120205518%22.PGNR.&OS=DN/20120205518&RS=DN/20120205518?ystfuv
 
 
 
There is a little more detail in this article.
 
http://www.slashgear.com/nokia-graphene-camera-tech-tips-true-pureview-without-the-bulk-04245872/

[Mt Spokane Photography]

There was a embarrassing claim which was withdrawn lest year for a similar sounding invention. 
 
http://iopscience.iop.org/0256-307X/28/12/129901/pdf/0256-307X_28_12_129901.pdf

[Mt Spokane Photography]

Yet another research paper detailing how a graphine photo sensor might work.  This from a couple of years ago.  Its a entirely different mechanism for sensing light, and could very well be more efficient.
http://www.mit.edu/newsoffice/2011/graphene-thermoelectric-1007.html

[Canon-F1]

the numbers are nonsense, the whole article misleading.
no word about read noise, A/D performance etc.

ISO/Noise is affected by more then how many photons you can catch.

and the sensor captures photons that don´t belong to visible light too (that is part of how they get such high sensitivity).

no word about how this will work for us photographer and "photography" when you capture IR and UV light.

[dilbert]

the numbers are nonsense, the whole article misleading.
no word about read noise etc.

Why is that necessary?

and the sensor captures photons that don´t belong to visible light too.
so no word about how this will work for photography when you capture IR and UV light.

Filter in front of the sensor, like they do today: IR DSLR conversions are made by REMOVING the filter that blocks out IR light.

Dilbert's Law of Photographic Technology: The eventual ramification of every technological advance in photography will be the proliferation of scatological images throughout the internet, predominantly of felines.

Not quite.

Dilbert's Law of Photographic Technology: The eventual ramification of every technological advance in photography will aid the proliferation of images of felines throughout the Internet.

[alexanderferdinand]

What would Joe McNally say to this statement?
Or David Hobby? Syl Arena? and so on........

[JohanCruyff]

"New camera sensor will eventually eliminate need for Tripod" seems more correct to me.

[awinphoto]

:: Shakes head:: photography is and always be about the capturing of light...  Yes, you can get decent images in dim light with the 5d3, but the usage and shaping of light will always be what separates the pros from the minutia.  knowing the direction of the light, flash or ambient, knowing the lighting patterns, the broad vs short light, how it sculpts the scene... Theres more to photography than shooting in pitch black... 

[pedro]

It's a pleasant inevitability that sensor design and innovation continues at such a cracking pace. Care to remember the gritty sensor on your original Canon 1D?

Lighting will always be important to quality photography to create shape, texture, mood and balance against available light. It's not that long ago that we thought 800 iso was fast, and how incredibly useful it was. Just a few years later we can shoot at 3200 & 6400 with impunity, and 12,800 at a pinch.

These speed gains now have less to do with being able to get an image in lower and lower light, but having the ability to stop action in environments where it previously the stuff of science fiction. Will 51,200 or 102,400 iso soon be the new 800? You'd better believe it.

-PW

I hope to see that within my life time. Turning 50 next year. Maybe not the new ISO 800. But an ISO 51.200 on my current 5DIII with the IQ of my ISO 3200 in RAW would be intresting. If there is something in photography I really strive for, then it ultra low light photography. 10-12 years down the road from here, where will we be at sensorwise?

[TAF]

The problem of durability could be solved by exchangeable sensors. Would be good idea just with CMOS sensors - I would like to convert my 20D to a B/W camera by removing the RGB Bayer pattern (and increase the sensitivity by a factor of two or three) but would like to have the chance to try the removal procedure on two or three sensors (not cameras).

I've wondered the same thing, so I have to ask - is that even possible?  It would certainly be a fascinating thing to experiment with.

Since the Bayer filter is right on the chip, I'm guessing one would need a clean room and some very exotic micromanipulators.  But even then, how does the software in the camera respond?  Would the nice people at Magic Lantern need to provide some new firmware to make things work?

If possible, it would be cheaper than a Leica, even if you started with a new Canon...