Naked Science Forum
General Science => General Science => Topic started by: jim on 06/08/2008 23:22:58
-
jim asked the Naked Scientists:
Is there a way to put a standard to colour and sound? ie a green colour has a sound frequency of xxx , know what i mean?
Thanks ! love your show (http://www.thenakedscientists.com/HTML/podcasts/) -
Jim Bremerton Washington -us
What do you think?
-
Monochromatic light can be characterised by a frequency (around 600THz - that's some 1000000000000 times higher than "middle C"), but more commonly by wavelength; eg 750nm (very deep red) through to 350nm (very deep violet) - green being around 550nm.
But there is no fundamentally meaningful connection between light colours and sound frequencies.
-
colours of noise ? (http://en.wikipedia.org/wiki/Colors_of_noise)
-
Colour is something that you experience in your brain; it is a subjective thing. No two people have exactly the same appreciation of colour but a lot of testing revealed a consensus amongst observers and a number of 'measures' of colour emerged.
Human colour vision involves three sets of sensors on the retina. They, tell your brain the relative proportions of red(ish), green(ish) and blue(ish) light. These are, efectively, the same as the Red, Green and Blue signals used in TV. BUT, as sensors, they are not what an engineer would gave gone for (nature's methods seldom are) and their 'analysis curves' overlap. However, the brain uses the three signals to come to an opinion about what colour to call an object.
A colour does not necessarily correspond to a particular frequency of light. The colours that you get when you split white light with a are specific wavelengths but you can produce the same sensation as you get for ,for example, the spectral yellow colour from a sodium lamp by adding appropriate amounts of red light and green light. The three signals to the brain may be different but the brain gets the same sensation. This, luckily, is how colour TV manages to work with just three phosphor colours.
There are three 'tristimulus' values which describe any colour of any brightness that you can see. You can actually separate the Brightness (luminance) signal from the colour (chromaticity) signal, which is the relative levels of the sensor outputs.
See the following link for a picture of how colour can be described. Notice how the 'spectral colours' are along the curved top bit and the 'non spectral' colours lie underneath the white, central bit.
The way that a particular sensation can be produced by adding proportions of two or more other colours is well shown on the diagram.
http://hyperphysics.phy-astr.gsu.edu/hbase/vision/colper.html (http://hyperphysics.phy-astr.gsu.edu/hbase/vision/colper.html)
The next link shows how you can produce some (but not all) of the possible colours that a human can see.
http://hyperphysics.phy-astr.gsu.edu/hbase/vision/colrep.html (http://hyperphysics.phy-astr.gsu.edu/hbase/vision/colrep.html)
TV systems use three phosphor colours which form a triangle on the CIE diagram. They can produce a good match for any colour falling within the triangle. Colours which lie outside the triangle (very vivid colours) get 'pulled in' onto the edge of the triangle. If you look at a crowd of people with bright clothes on TV, you may notice that many of them appear to be wearing the same shade of coat - that may be because they are different but end up in the same place on the CIE diagram and are displayed as the same.
-
There's the Pantone system for colour matching:
http://en.wikipedia.org/wiki/Pantone
-
Yes, that's true, but the Pantone Colours are still measured (or can be measured) using CIE colour coordinates. The 'Swatches' of Pantone colours are used for convenience and cover a limited set of colours.
-
Is there a standardised system for describing colours?
Yes, there's the various CIE (French, but roughly translated as Commission on Illumination) standards, eg. (x,y). Chromaticity (colour, neglecting brightness) can be represented on a two-axis diagram.
Sorry to plug my website again, but there's a nice CIE chromaticity diagram to be found at http://www.techmind.org/colour/ at about one-third of the way down the page.
There are also device-specific ways of putting a numerical value on colour, such as (RGB) for a known/standardised display, or (U,V) or (Cr,Cb) in the world of television...
-
There's more to colour than the cie chart; where's brown?
-
There's more to colour than the cie chart; where's brown?
That needs to be mentioned - thanks.
Brown has the same chrominance values as yellow. It just has a lower luminance value.
The cie chart normalises luminance in order to get things on one two-dimensional scale. The full graph would be a vertical column to include luminance.
'Sludgy' colours only look sludgy when held against brighter objects. I was fascinated to learn that, apart from the luminance values, skin tones are more or less identical for all races. I guess that means that the pigment is more or less neutral.
The eye / brain are too smart for their own good, sometimes. Between them they do their very best to eliminate the effects of illumination. If you cut a hole in a yellow card and put an identical card behind the hole, if you can manage to get a really deep shadow on the yellow bit behind (and you can suspend disbelief a bit). The rear bit will look brown (ish).
Failing that, use Photoshop (o.n.o.) and look at the RGB components of a yellow and a brown area of a picture. You'll find their ratio is much the same although their absolute values differ.
-
There's more to colour than the cie chart; where's brown?
That needs to be mentioned - thanks.
Brown has the same chrominance values as yellow. It just has a lower luminance value.
The cie chart normalises luminance in order to get things on one two-dimensional scale. The full graph would be a vertical column to include luminance.
'Sludgy' colours only look sludgy when held against brighter objects. I was fascinated to learn that, apart from the luminance values, skin tones are more or less identical for all races. I guess that means that the pigment is more or less neutral.
The eye / brain are too smart for their own good, sometimes. Between them they do their very best to eliminate the effects of illumination. If you cut a hole in a yellow card and put an identical card behind the hole, if you can manage to get a really deep shadow on the yellow bit behind (and you can suspend disbelief a bit). The rear bit will look brown (ish).
Failing that, use Photoshop (o.n.o.) and look at the RGB components of a yellow and a brown area of a picture. You'll find their ratio is much the same although their absolute values differ.
Agreed, brown is just dark yellow or dark orange - the eye only perceives it as the darker colour when there's significantly brighter colours elsewhere in the field of view and/or recent visual memory.
Similarly (but simpler) is grey. A colour may look white until you place a whiter box around it, then the previous looks grey. Then put an even brighter box around that... (a nice PowerPoint demo in a darkened lecture theatre).
I guess the human visual processing is mainly designed for observation of reflected colours, where you have some notion of the overall ambient level of illumination, ie some feel for the total reflectivity of any given object. It all depends on context.
Also, I suppose, a very dark object will only really look "black" in the presence of much brighter surround. (consider a traditional TV screen - looks murky green when switched off, but you'll still believe you're seeing black within images when it's powered up).