New Sensor a Thousand Times More Sensitive Than Current Camera Sensors

Graphene seems to be the new miracle material, too bad no one ever actually releases a product made from it. In due time I guess....

http://www.sciencedail​y.com …/2013/05/130530​094624.htm

It is also still rather new, and production methods of graphene are still not ready for industrial deployment. That stuff is still in the realm of the lab. Give it time.

So will this thing will be able to capture enough dynamic range to match the human eye?

If so, then its a great step in the right direction.

http://spectrum.ieee.o​rg …s-more-sensitive-to-light and check the comments.

The 1000x is compared to previous graphene sensors and not to sensors on a whole. Whoever originally wrote the press release messed up.

Sounds like ISO worries are gonna be a thing of the past and camera prices are gonna drop through the floor. Anyone up for noise free ISO 10,000,000?

Ginga wrote in post #15986593
So will this thing will be able to capture enough dynamic range to match the human eye?

If so, then its a great step in the right direction.

The human eye actually doesn't have a huge dynamic range, and has terrible resolution and sharpness. It's just that we can look quickly at the bright part of a scene (e.g. the sunset) and look quickly at the dark part (the foreground) and our brain puts everything that we see together into a coherent scene that appears to have huge dynamic range, just like HDR. Or we can look at something with only limited detail, and our brain fills in the gaps automatically (like the Heal brush) - this can sometimes backfire, which is why we see optical illusions the way they are.

Essentially, our eyes are third-rate cameras, but are attached to a super-high-powered, automatic version of Photoshop.

boerewors wrote in post #15987008
Sounds like ISO worries are gonna be a thing of the past and camera prices are gonna drop through the floor. Anyone up for noise free ISO 10,000,000?

Not going to happen, because light is quantised - you can't have less light than a single photon. All it means is that the vast majority of noise will be shot noise from the incoming light, rather than noise from an electronic source.

Ginga wrote in post #15986593
So will this thing will be able to capture enough dynamic range to match the human eye?

If so, then its a great step in the right direction.

Sensitivity <> Dynamic range

Shadowblade wrote in post #15987148
The human eye actually doesn't have a huge dynamic range, and has terrible resolution and sharpness. It's just that we can look quickly at the bright part of a scene (e.g. the sunset) and look quickly at the dark part (the foreground) and our brain puts everything that we see together into a coherent scene that appears to have huge dynamic range, just like HDR. Or we can look at something with only limited detail, and our brain fills in the gaps automatically (like the Heal brush) - this can sometimes backfire, which is why we see optical illusions the way they are.

Essentially, our eyes are third-rate cameras, but are attached to a super-high-powered, automatic version of Photoshop.

Hey, I really liked my eyes up until now!

Interesting read.

boerewors wrote in post #15987008
Sounds like ISO worries are gonna be a thing of the past and camera prices are gonna drop through the floor.

can't wait until it's in cell phones

This is another example of lazy journalism.

First, the paper states that there is a 1000x increase in sensitivity compared to previous graphene sensors, not current camera sensors. In fact, they have a sensitivity in the nW range, which is far worse than current silicon sensors.

Second, the paper states that the gain has been improved compared to traditional silicon photodiodes. However, it does not look at the signal to noise ratio of the graphene detector. So while it might do a better job at amplifying the signal, it will also amplify the noise equally.

Third, photons are quantized, and therefore inherently noisy. You need lots of photons to hit each pixel to average that out.

Fourth, assume you capture an image with less than 10 photons per pixel. How useful will that be for your image if you only get 10 values in each color channel?

Current silicon sensors are already almost at the theoretical physical limits in terms of quantum efficiency and signal to noise. So you're unlikely to see any more huge advances in that regard. Where there is still room for improvement on the sensor part is reading out the captured signal. The amplification of the signal is where most of the noise comes from.

So in summary, while graphene may be the way to go in the future, it certainly won't be improving our sensors anytime soon.

Stone, you are right! New graphene camera sensor is much better. It takes better dim-light photos and cameras fitted with a new revolutionary sensor will soon be able to take clear and sharp photos in dim conditions. The new graphene-based sensor is the first capable of detecting broad spectrum light, from the visible to mid-infrared, with high photoresponse or sensitivity.

hwan wrote in post #15991763
This is another example of lazy journalism.

First, the paper states that there is a 1000x increase in sensitivity compared to previous graphene sensors, not current camera sensors. In fact, they have a sensitivity in the nW range, which is far worse than current silicon sensors.

Second, the paper states that the gain has been improved compared to traditional silicon photodiodes. However, it does not look at the signal to noise ratio of the graphene detector. So while it might do a better job at amplifying the signal, it will also amplify the noise equally.

Third, photons are quantized, and therefore inherently noisy. You need lots of photons to hit each pixel to average that out.

Fourth, assume you capture an image with less than 10 photons per pixel. How useful will that be for your image if you only get 10 values in each color channel?

Current silicon sensors are already almost at the theoretical physical limits in terms of quantum efficiency and signal to noise. So you're unlikely to see any more huge advances in that regard. Where there is still room for improvement on the sensor part is reading out the captured signal. The amplification of the signal is where most of the noise comes from.

So in summary, while graphene may be the way to go in the future, it certainly won't be improving our sensors anytime soon.

so it's all a bunch of smoke and mirrors

It's a step in the right direction, but it'll be a while before we see it used as camera sensors.

Shadowblade wrote in post #15987148
Essentially, our eyes are third-rate cameras, but are attached to a super-high-powered, automatic version of Photoshop.

Great way of putting it.
I'm stealing this!