Camera vs Eye Comparison

[erickulcyk]

Hello,

It is my understanding that the hum eye has 3 types of cones to see color, and rods to see brightness. And it is the rods which allow us to see well as night. I also gather that the reason cameras perform so terribly in dark conditions, is that the digital camera senses rgb color, which seem similar to cones. Do some cameras also have of rods, which would allow you to take great pictures in the dark (just not much color)?

Thanks,
Eric

 

[Ernicus]

I know this isn't any help, but maybe someone will remember and find the thread, but there was an example...in graphic form, of exactly such thing as your asking about here in the forum somewhere. Maybe someone can find it and link it.

 

[rexbobcat]

Cameras see in color at all times unless you use specific software or film. Rods are black and white, which is why humans see in b&w at night.

So no the camera does not have rods. It does not function as organically as our eyes do.

The closest comparison I can think of is ISO.

 

[Trever1t]

you are way over thinking ... unless you are that kinda engineer. The camera sensor is barely comparable to the complexity of anything human.

 

[Sw1tchFX]

And poor old digital cameras can only see 255 different variations of red, green, or blue each

 

[KmH]

Digital cameras don't 'see' any color at all. We'll get back to that in a moment.
For more about the RGB color model see - RGB color model - Wikipedia, the free encyclopedia

Human eyes work in a non-linear way. The analog image sensors in a digital camera can only 'see' in a linear way, and the image sensor only see's luminosity (no color).
http://www.adobe.com/digitalimag/pdfs/linear_gamma.pdf

A pixel is a device that developes an analog voltage based on how many particles of light strike it during an exposure. A single particle of light is called a photon. Photon - Wikipedia, the free encyclopedia

The voltage a pixel developes is very small, even if a lot of pixels fall on it, and the voltage has to be amplified. How much the voltage gets amplified is what the ISO setting on the camera is about. At low ISO the voltage is amplified less than when the ISO is set to a higher value.

Once the voltage of all the pixels has been amplified, all the pixel analog voltages are changed to a digital value in a analog-to-digital converter. Those digital values are either 12-bit (4096) or 14-bit (16,384) discrete values of luminosity (no color). http://wwwimages.adobe.com/www.adob...ly/prophotographer/pdfs/pscs3_renderprint.pdf

Most digital camera have a Bayer Array in front of the image sensor. The few DSLR's that have a Foveon image sensor (Sigma cameras) are different.

This is what is in front of a 4 pixel square of pixels. One pixel only developes its voltage based on the intensity of the red part of the light hitting it, one pixel only developes its voltage based on the intensity of blue part of the light hitting it. Two pixels only develope their voltage based on the intensity of green part of the light hitting it, because human eyes are more sensitive to green light. Bayer filter - Wikipedia, the free encyclopedia

Before the digital values that come out of the A/D converter can be made into an image several steps have to be taken. Based on the layout of the Bayer Array, a software algorithm interpolates what color of light fell on each pixel. that process is called 'demosacing'. Even though a single pixel was covered by a red Bayer Array element, the digital signal levels of the adjacent pixels is taken into account by the demosaicing algorithm to infer how much blue and green light actually fell on the pixel covered by the red filter. The same is done for the other adjacent pixels. Active pixel sensor - Wikipedia, the free encyclopedia
Next a non-linear gamma curve has to be applied so it looks like something humans eyes actually see, because what the image sensor produces doesn't come anywhere close.
http://en.wikipedia.org/wiki/Algorithm

http://www.bhphotovideo.com/bnh/con...getpage&Q=newsLetter/The-Photographic-Eye.jsp

http://en.wikipedia.org/wiki/Foveon_X3_sensor

Below is an approximation of what the image sensor actually produces, having a linear gamma curve and being grayscale, followed by an unedited Raw conversion (using ACR) that has a non-linear gamma curve applied and interpolated color.

 

[Ernicus]

I thought it was a thread you posted in Keith, but I didn't want to mis quote ya. Great information. Always a good read when you post techno stuff.

 

[pgriz]

Actually human perception is much more complicated than what has been discussed. Beyond the ability of the eye to capture the photons, there is an enormous amount of processing that occurs at various levels (retinal, edge and motion motion detection at the eve level, combining/splitting the stereo images at the optic chiasm, and the direction of the different signals (movement, edge detection, colour, contours) to different parts of the visual cortex, and the processing of this information to form the awareness of what we see. Our brains construct a mental model of what we see around us, and use frequent updates from the visual subsystem (but also auditory, and tactile) to update this model on a continuous basis. Evolution has resulted in our physiological system giving priority to certain signals (a sudden change in the visual field is interpreted as movement and results in extremely fast reaction in the brain), whereas other signals are used to supplement and cross-verify sensory information (which is why alcohol and other drugs that affect sensory processing can create such a havoc with perception). \

For those interested in this very fascinating area, two links that may be useful are: Visual system - Wikipedia, the free encyclopedia and Visual perception - Wikipedia, the free encyclopedia.

 

[KmH]

Where did the OP go? :er:

 

[unpopular]

Prob 75% of what we see is a result of post processing.

 

[Compaq]

You can argue, what is reality? Is it what a camera captures? Or is it what we, as humans, perceive? We disregard very much information that reach our brain, because it's not essential. Take a painter, for example. He or she might paint a romantic landscape that's based off from a real place. But painters work by inclusion, they decide what they want to include. If there as a motor road running through, the painter doesn't need to include that, and so that's not part of the painter's reality, or rather, perceived reality. The painter saw a wonderful landscape, and disregarded the road.

Photographers work by exclusion, and our brain is the same. We exclude (disregard) certain information we don't need. We never notice our nose, yet it's clearly visible all the time. Whose to say we disregard more? Reality is hard to define, because it's a floating term, almost an abstract term.

This might be a little on the side of the debate, though :/

 

[KmH]

100% of what humans see is post processed. Visual perception - Wikipedia, the free encyclopedia

As it is we see only a vary narrow slice of the total electromagnetic spectrum. We see wavelengths of light that are between 380 nm and 760 nm (790–400 terahertz in frequency).

If the sun was blue instead of yellow, grass and other things that are green, would be purple.

 

[rexbobcat]

HOW MANY "FRAMES PER SECOND" DOES THE HUMAN EYE SEE AT APPROX.?

I typed it in caps because I wanted to semi-derail the topic from all that other stuff I have no idea about.

I seriously would like to know, though.

 

[unpopular]

between 15 and under 60, depending on conditions and the individual.

 

[unpopular]

You can argue, what is reality? Is it what a camera captures?

We cannot directly sense reality, everything we see is photo energy that is converted into electrical energy which is interpreted through memory. So what we argue as "reality" is actually photons reflecting off the subject, focused through the eye and converted into electrical energy and interpreted by memory. I repeated that on purpose.

The same can be said of kinetic energy being converted into the sense of touch, chemical energy to taste and smell.

So there is this sort separation between my ego and the world around me which we cannot get around, and furthermore, there is a momentary gap between the light reflecting off the subject and my mind interpreting the electrical signals which the light is converted to. Even if we could interpret this data instantaneously (we can't) the speed of light would cause it. Likewise, we can't perceive this gap either, though some neurologists have implied that Deja Vue may be a side effect once things get out of sync.

So really, a sensor without a lens collecting unorganized photons is probably more accurate to the nature of reality than our eyes - after all a lens is a device created to organize light in a way that we can recognize, but it is still a representation, and moreover our viewing of such an image is a representation of that representation.

The only conclusion I can come to is that the chair I am sitting on is the only true observer of reality; unfortunately it's very inability to observe is what makes it so.