12

I've noticed when I am in very low light (light that isn't in the same room I'm sitting in), when my eyes adapt to the dark that I see colored particles. Do these colored particles resemble noise in the photography world?

Something like high ISO noise, but the particles are less colored. Also I noticed that this happens when you apply pressure on your eyes, or when you stare at something at dark I can feel the grain; it's not smooth as it looks under light.

7
  • Can you clarify what you mean by "when I set in very low light"? Also, would it be possible to post an example as I am not sure what you are referring to.
    – dpollitt
    May 24, 2012 at 2:26
  • 5
    @Akram: A sample photo, preferably with EXIF data or at least an explanation of the camera settings used, would be extremely helpful here. :)
    – jrista
    May 24, 2012 at 2:37
  • 8
    @jrista: unless Akram has very special eyes, they don't write EXIF data. :)
    – mattdm
    May 24, 2012 at 2:47
  • @dpollitt: are you certain that the why isn't actually similar in both cases?
    – mattdm
    May 24, 2012 at 2:52
  • 1
    I think you may have Visual Snow visualsnow.eu
    – GypsyKing
    Oct 25, 2015 at 5:49

4 Answers 4

17

The sensor systems are different enough that direct comparison is hard. There are some similarities, but the sensor post processing is exceptionally well tailored to remove undesired artefacts and the maker has not provided a means to turn noise reduction off.

Also, the image is developed by a custom algorithm and the system does not allow access to the RAW data.

Pressing the sensor is cheating and does induce artefacts as signal can be generated which is sensibly indistinguishable from photon stimulation within the criteria established by the wetware. Pressing the sensor, either through the flexible housing or directly may cause degradation or destruction and is not within the standard operating conditions or guaranteed worst case specs and so is not covered by warranty.

There are two sensor systems whose outputs are combined (something like Fuji's dual site size sensor but totally different).

You'll read things like:

  • The eye has about 100,000,000 "rods" which are monochrome only sensors. There are about 5 - 10,000,000 "cones" which are colour receptors but less sensitive than the rods. Most of these are in the centre of the eye in an area about 0.5mm across (Work that out for sensor cell area !)

To make rubbish of that statement, you'll also read that

  • there are RGB cones but far fewer blue than R&G and the blue are outside the centre but far more esensitive than the R&G so overall the RGB sensitovity is about the same.

Whatever ...

enter image description here

As light levels drop the cones start to stop working. For my eyes - which seem reasonably standard in thsi respect (and not others) at 20 lux colour is not too bad. At about 10 lux you can still see colour but notice it starting to suffer. From there it fades away and by 1 lux it's essentially monochrome. Bright moonlight is a few tenths of a lux. Stumbling around a room that is so dark that you can sort of see doorways so as to get through then level is somewhere under 0.1 lux so by 0.01 vision per se is largely gone.

BUT and the reason why the above is worth saying at all (maybe) is that the dark adapted eye can detect a single photon. If you are in total darkness you will not see every single photon as there is substantial dead area between the sensors, but if a photon strikes a sensor it will fire and you will see a spot of light. What that spot of light registers as is uncertain. If it fires a rod you'd expect monochrome. Whether it's able to fire a cone may depend on energy level - so if so you'd expect blue flashes to be more common.

Finally, long shot: and this is a maybe, you MAY be able to see secondary emissions from Gamma rays !. Gamma ray "telescopes" work by looking for secondary emissions caused by high energy gamma rays striking atoms in the atmosphere and causing a visible photon emission at lower energy. Vanishingly few of the high energy gamma rays make it to earth surface (to contribute to the background count you hear on a Geiger counter) but perhaps a dark adapted eye gets the benefit of a few of these knocking secondary particles off other parts of your eyes! Maybe.


Added.

Relevant (maybe :-) )

http://hyperphysics.phy-astr.gsu.edu/hbase/vision/rodcone.html

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

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

Video

Good: http://www.cis.rit.edu/people/faculty/montag/vandplite/pages/chap_9/ch9p1.html

Goodish: http://www.vetmed.vt.edu/education/Curriculum/vm8054/eye/RODCONE.HTM

Eye: http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/V/Vision.html

9
  • When your source says 'RGB sensitivity is about the same', they mean that R&G sensitivity is about the same as B sensitivity. Cones are less sensitive to light in general than rods. This is why you see things better in the dark out of the corner of your eye, because the light is coming in at an angle and hitting the rods around the edges of the retina. Many armed forces train their members to scan the terrain this way at night. May 24, 2012 at 5:58
  • @ElendilTheTall - Yes, I meant the RGB sensitivity the way you explained it. I'm not sure what it sounded like I was saying but evidently it's not clear enough. What I meant re "rubbish" is that the "all in the centre" idea is not true for blue. Must make demosaicing interesting - think about it - when eye is turned to get high res in centre the blue cones don't benefit. So blue colour channel should be lower res than RG even though same sensitivity overall for big picture. May 24, 2012 at 7:02
  • Maybe it is - human colour vision is relatively easy to fool, because the brain 'fills in the blanks'. May 24, 2012 at 9:08
  • This is a great response! :D May 24, 2012 at 10:15
  • +1 for the dark adapted eye can detect a single photon. this is actually what happens and I updated my question with this note
    – K''
    May 25, 2012 at 1:45
2

First remember that color is only an illusion born into your brain: most of mamals will have a colorspace reduced to red and blue, birds have extended colorspace seeing also in UV bees see yellow blue and UV. Show a picture of a flower to a bird or a bee it will not recognize the color (as our device do not record the UV). The color is build using a combination of the light intensity (from the rod) and the color signal (from the cone)

For human eyes perception details (and a nice picture of the cone on the retina) have a look to http://www.beercolor.com/color_basics1.htm

A very important point is to understand that the perception of the outer world through the eyes is NOT the processing of a simple image: The eye has a good resolution only at the center (where it also see color) therefore when you look at something your eyes will scan the scene geting bits of information and your brain is caching the data, extrapolating some part of your field of view and reconstructing an image. Furthermore ther is a remanance of the image on your retina (used to make you belive that there is some movement in a movie)

You can be aware of parts this process by thinking that when you look at a scene everything is in focus, you eyes are not so good: this is a composit. Think also that your eye has a blind spot that you never notice (the image is extrapolated) there sre some experiment which allow you to evidence it (see the test on http://en.wikipedia.org/wiki/Blind_spot_%28vision%29) RImage reconstruction can also be tricked: this is optical illusion

The color that you see when you press on your eyes are due to mecanical constraint on the retina (the normal behavior of cone is that the pigment they contain will elongate when reacting to light causing a pressur which originate the nervous signal, one may also experience such color patch in some headheach or when hurt in the back of the head. In this case the signal originate directly into the visual cortex.

In low light it is not clear if the noise you see originate in the device (your eyes) or in the processing (your brain)...

2

You are referring to Phosphene. This doesn't have anything to do with photographic noise.

2

Your description sounds a lot like the condition known as visual snow, which some people liken to snowy interference on an old analog television. The linked Wikipedia page says:

The cause is unclear. The underlying mechanism is believed to involve excessive excitability of neurons within the cortex of the brain.

If true, then perhaps what you're talking about is related to the high-ISO noise recorded by a digital camera in the sense that it's more or less random data added to the image, but the place where the noise is added is different. In a digital camera, the noise comes from the sensor; with visual snow, it's apparently added in the brain, so later in the imaging process.

It seems that not much is known about visual snow, and there's not even clear agreement in the medical community about whether it's a real problem. The Guardian had an interesting article about the condition, which you can read here: The mysterious eye condition of 'visual snow'​.

1
  • 1
    that's an interesting article, thanks for the link
    – K''
    Jan 22, 2018 at 16:59

Not the answer you're looking for? Browse other questions tagged or ask your own question.