4 ms·

There is no hump in the red detectors. The hump exists only in curves we use to translate light into RGB, and is there precisely to help translate that violet i

There is no hump in the red detectors. The hump exists only in curves we use to translate light into RGB, and is there precisely to help translate that violet into purple. https://en.wikipedia.org/wiki/Cone_cell https://en.wikipedia.org/wiki/Cone_cell

tgb 8y ago
Thanks for pointing this out, thought I was going crazy. So the article's explanation of why violet and purple look similar even though they are completely unrelated in spectrum is bogus (the article claims it's due to this secondary peak in the red cone). So what's the real reason?
- pygy_ 8y ago
  Quoting the legend of said picture in the article: """Note that this chart does not show the spectral properties of the cones themselves (but they look similar). It represents the CIE 1931 colour space, which, simply put, corresponds to the signals after they have been processed by the brain."""
  - tgb 8y ago
    Quoting the last section of the article: """The reason why purple and violet look similar to us is because they stimulate our cones in a similar way, but most other animals don’t share the same types of cones and “post-processing”. This means that to other animals, purple and violet may look completely different!""" Emphasis theirs. This is the part I'm calling out. Stimulation of the cones is not relevant! (Of course the cones are relevant but the difference is very much larger and of a different nature than they make it out to be. Most mammals have only two color cones.)
- jacobolus 8y ago
  The article is utterly full of shit. Please disregard it.
blincoln 8y ago
The mean absorbance for the red cones was higher from about 400nm to about 430nm than it was for the green cones in at least one published paper.[1] For some reason that's not clear to me, the graph below the data table doesn't extend the red cone curve far enough to the left to illustrate the hump. [1] _Human visual pigments: microspectrophotometric results from the eyes of seven persons_, page 7 (labeled page 121) in https://royalsocietypublishing.org/doi/pdf/10.1098/rspb.1983.0091 https://royalsocietypublishing.org/doi/pdf/10.1098/rspb.1983...

tgb 8y ago

Thanks for pointing this out, thought I was going crazy. So the article's explanation of why violet and purple look similar even though they are completely unrelated in spectrum is bogus (the article claims it's due to this secondary peak in the red cone). So what's the real reason?

pygy_ 8y ago
Quoting the legend of said picture in the article: """Note that this chart does not show the spectral properties of the cones themselves (but they look similar). It represents the CIE 1931 colour space, which, simply put, corresponds to the signals after they have been processed by the brain."""
- tgb 8y ago
  Quoting the last section of the article: """The reason why purple and violet look similar to us is because they stimulate our cones in a similar way, but most other animals don’t share the same types of cones and “post-processing”. This means that to other animals, purple and violet may look completely different!""" Emphasis theirs. This is the part I'm calling out. Stimulation of the cones is not relevant! (Of course the cones are relevant but the difference is very much larger and of a different nature than they make it out to be. Most mammals have only two color cones.)
jacobolus 8y ago
The article is utterly full of shit. Please disregard it.

blincoln 8y ago

The mean absorbance for the red cones was higher from about 400nm to about 430nm than it was for the green cones in at least one published paper.[1] For some reason that's not clear to me, the graph below the data table doesn't extend the red cone curve far enough to the left to illustrate the hump. [1] _Human visual pigments: microspectrophotometric results from the eyes of seven persons_, page 7 (labeled page 121) in https://royalsocietypublishing.org/doi/pdf/10.1098/rspb.1983.0091 https://royalsocietypublishing.org/doi/pdf/10.1098/rspb.1983...