The huge problem is that we lack vocabulary to talk about unique qualia. Our words come from talking to other people and if nobody around us has the same qualia as we are, nobody gave us a word.
At the moment I’m learning to distinguish colors better via an Anki deck. I use the CSS color name definition.
Seeing the difference between navy and midnightblue is still hard for me but I’m confident that I can learn it with practice. Some day I will hopefully even be able to tell apart snow from floralwhite.
I like the particular deck and if someone wants to train their color perception I’m happy to share it. It’s build in a way that you get progressively more difficult decisions and provides years of fun at 5 new cards per day.
I would also like to create a deck to train sound perception. Does anyone know of a good tool that can automatically produce sound files with a specific pitch for pitch training? At best a tool that can be used via the command line.
Even better than that is this series of blog posts, which talks about color identification across languages, the way that color-space is in a sense “optimally” divided by basic color words, and how children develop a sense for naming colors:
How well calibrated is your monitor? My wife recently had a “That’s not what I wanted!” experience after discovering that the brightness levels which are optimal for watching movies are not optimal for designing artwork which looks the same after printing. Unless you’ve done some careful tweaking you might not be learning the colors you think you’re learning.
For sound generation, what operating system are you using? Alsa (at least on Ubuntu 13.10) comes with “speaker-test”, a command line program which can be used to play specified frequency sine waves, and alsa itself can be configured to allow you to save sound output to files.
..And that’s pretty much the story of how at work we ended up with a hideous orange conference table instead of the nice warm brown our department chair envisioned
Unless you’ve done some careful tweaking you might not be learning the colors you think you’re learning.
Calibration is something I don’t learn, at the moment I can more about fine color distinction.
For sound generation, what operating system are you using? Alsa (at least on Ubuntu 13.10) comes with “speaker-test”, a command line program which can be used to play specified frequency sine waves, and alsa itself can be configured to allow you to save sound output to files.
I use windows, but I might write a script and run it on a linux computer to get files, I will think about it.
Calibration is something I don’t learn, at the moment I can more about fine color distinction.
You’re teaching yourself to associate labels (e.g. “midnight blue”) with particular colors.
You seem to be doing the colors on your monitor and I doubt your monitor is color-calibrated. This implies that you’re looking at a somewhat-randomized version of a true color. Your monitor also has a limited gamut (the set of colors that it can display) and for some monitors the gamut is very limited to the extent that they have to simulate some colors by dithering. Most uncalibrated monitors are too bright and too blue.
Basically, display of proper colors on a screen is complicated and you must take active steps (e.g. color calibration) to make it happen.
You’re teaching yourself to associate labels (e.g. “midnight blue”) with particular colors.
Actually the css label is “midnightblue” without a space. For my purposes I don’t think it’s a problem to treat the word true as meaning whatever the hardware interprets it too mean.
I want to get better at precision of color distinction. Having accurate labels is secondary.
I don’t think it’s a probably to treat the word true as meaning whatever the hardware interprets it too mean
Your sentence doesn’t make sense, but keep in mind that we’re talking about what your local, particular, individual hardware interprets it to mean. My hardware (my screen) will likely interpret it differently. Screens of random people will interpret it differently again.
Also a lot of LCD screens use what’s known at TN (twisted nematic) panels and the great majority of them are 6-bit panels, that is, they drive pixels at only 2^6 = 64 levels. So, for example, they can only show you 63 levels of blue (0 is off/black) even though the standard software treatment of color is 8 bits per primary for 256 levels. The TN screens mitigate this problem through dithering.
Human vision is highly adaptive to ambient light, in particular its color temperature. We perceive colors differently in the sunlight, in the shade, or under a tungsten lamp. And these are full-spectrum light sources. Fluorescent lights are not, they have spikes and voids in their spectrum and that affects certain colors under them as well.
Yes, there should have been the word “problem” when I wrote the word “probably”.
As I said, it’s complicated.
I know that human vision is complicated.
I however also know my bit about measurement theory. The important thing isn’t that a measurement is “true” but that that it has features like sensitivity and specificity. Accuracy and precision are other words to speak about measurements.
When dealing with a complex subject the important thing is whether your map of reality is good enough for the purpose for which you want to use it. I think the way my computer models colors is good enough to produce a valid training effect that helps me to get better at distinguishing colors.
I could create a randomly generated filtered deck out of 100 mature Anki color cards and test it at the computer of someone else and see whether the particularities of my specific computer monitor produce a problem. It’s a question that has an empiric answer and in a year when I have put more training into the colors I probably will do something like that. At the moment I have one day 70% correct cards and the next day 85% right cards, so it’s not stable enough for good test.
My notebook has a slightly different color profile than my 24″ monitor. Switching between the two doesn’t produce issues I can perceive. It doesn’t feel like everything is suddenly wrong. My brain seems to be able to make the necessary adjustments that I can still answer the cards correctly.
Let’s use meatspace examples. Do you want to be able to look at a wall and say “this is color X”? Do you want to to be able to look at two walls side by side and say “These are different colors”? Do you want to be able look at a wall, look at another wall the next day and say “This is the same color as I saw yesterday”?
Being able to answer the question: What color is this isn’t very useful. It’s not what separates the person who rather wants to be deaf from the person who rather wants to be deaf. I want is a cure for my partial blindness. I want to perceive more bits of information through the visual channel.
I don’t care for ‘is’ or ‘true’. I’m post- aristotelean. To quote the constructivist Heinz von Foerster: Truth is the invention of a liar. Just in case you think, I’m off-topic, I’m not. Those Anki cards are a result of among other things reading Korzybski’s Science and Sanity.
To go back to questions I want to be able to notice if a website I visit changes their color scheme in a way that exchanges navy with midnightblue just as I’m now able to notice a change from red to green.
Given that your brain’s processing bandwidth is severely limited
I don’t think there a good reason to believe that. I think quite often the limiting factor is time spent in deliberate practice and not lack of neurons or similar hardware problems.
Does anyone know of a good tool that can automatically produce sound files with a specific pitch for pitch training? At best a tool that can be used via the command line.
Sounds like you want sox. To make an mp3 that plays an A 440 for 1 second you would do:
sox -n a440.mp3 synth 1 sin 440
But note that most real world sounds are a combination of many frequencies, so training on sine waves may not be what you want.
But note that most real world sounds are a combination of many frequencies, so training on sine waves may not be what you want.
I would think that training on them provides useful skills that generalize more broadly. It’s probably not perfect but it’s easy to create cards with binary choices that can get progressively more difficult.
The goal is getting to a point where I engage into deliberate practice of distinguishing sounds and using Spaced Repetition to do it.
If anyone who’s good at sounds has better ideas about creating a Anki deck to train distinguishing sounds, I would be happy to hear ideas.
I also try to train phonemes, but creating good cards for it proved to be hard. The first cards I created where simply to hard for myself as I’m not good at audio perception.
I can hear a lot more in a Salsa song than I could hear 5 years ago.
I think that it’s worthwhile to invest significant time in getting to perceive more bits of reality in daily life. I’m still at the phase of experimenting about how to train myself and others to have richer qualia, but I think it’s an area worthy of further investigation.
Pitch seems to me like a very straightforward concept, but I’m also willing to learn other ways of distinguishing sounds.
I also try to train phonemes, but creating good cards for it proved to be hard.
Instead of phonemes in isolation, it should work to train on them in words as minimal pairs. For example, to train the difference between /b/ and /d/ you would test discrimination between /bog/ and /dog/, /cab/ and /cad/, /cabby/ and /caddy/, etc.
Under different light contexts an reflective object might be closer to either midnight blue or navy. Have you attempted using paint chips or something to test yourself under sunlight versus florescent light or anything?
Under different light contexts an reflective object might be closer to either midnight blue or navy. Have you attempted using paint chips or something to test yourself under sunlight versus florescent light or anything?
I know at the moment it seems to me that the colors are far enough apart that light conditions at my PC are not the main problem.
My notebook is slightly differently configured and I didn’t judge many Anki cards wrongly when answering on the notebook instead of my main monitor.
I know at the moment it seems to me that the colors are far enough apart that light conditions at my PC are not the main problem.
Your monitor emits light, so the light conditions matter less, mostly needing to overcome the ambient light (laptop in sun).
Most things don’t produce their own color though, they reflect varying amounts of the incoming spectrum. If that incoming spectrum is different, the outgoing spectrum is different. You can take advantage of that in various ways, but it might also confound the question of what color “is” this object.
Or maybe you automatically take that into account by using the ambient light as a reference, I was wondering whether you had tested for that or not?
Or maybe you automatically take that into account by using the ambient light as a reference, I was wondering whether you had tested for that or not?
At the moment I haven’t tested. I spent a total of 7 hours with the average of 5 minutes per day on the Anki cards and I seem to be getter better at color distinction.
Every card provides a binary choice. I have cards that present me with two color words and a large circle that’s filled with the corresponding color. The difference of the colors is at the beginning stage where I’m still at least a total 32 different hex values.
I also have cards that ask for the hex value of the colors.
There were some days were I traveled and used my laptop in other light conditions. They weren’t a problem.
For the time in 7 years I created cards to distinguish 4dc636 from 4dc637. That might run into issues with light conditions. I don’t know whether it does and whether the human mind is trainable to distinguish colors as finely, but I will find out if I continue spending my 5 minutes every day.
but it might also confound the question of what color “is” this object.
I don’t like “is” anyway for the reasons Korzybski layd out.
I want to increase the amount of bits I perceive through the visual channel. It’s an open experiment.
The outcome might be that I have color distinction in a few years that allows me to impress people with stunts. I might learn something valuable about colors that can be made into scientific paper or blog post. I also expect that I will get better at usability design even if I don’t get superhuman color perception abilities out of the project.
But you are right that having data about the light conditions would be good. I opened a thread on the QS forum about the search for a proper tool.
The huge problem is that we lack vocabulary to talk about unique qualia. Our words come from talking to other people and if nobody around us has the same qualia as we are, nobody gave us a word.
At the moment I’m learning to distinguish colors better via an Anki deck. I use the CSS color name definition. Seeing the difference between navy and midnightblue is still hard for me but I’m confident that I can learn it with practice. Some day I will hopefully even be able to tell apart snow from floralwhite.
I like the particular deck and if someone wants to train their color perception I’m happy to share it. It’s build in a way that you get progressively more difficult decisions and provides years of fun at 5 new cards per day.
I would also like to create a deck to train sound perception. Does anyone know of a good tool that can automatically produce sound files with a specific pitch for pitch training? At best a tool that can be used via the command line.
Regarding color you might want to have a look at the as usual funny and detailed XKCD color survey: http://blog.xkcd.com/2010/05/03/color-survey-results/
Even better than that is this series of blog posts, which talks about color identification across languages, the way that color-space is in a sense “optimally” divided by basic color words, and how children develop a sense for naming colors:
http://www.wired.com/wiredscience/2012/06/the-crayola-fication-of-the-world-how-we-gave-colors-names-and-it-messed-with-our-brains-part-i/ http://www.wired.com/wiredscience/2012/06/the-crayola-fication-of-the-world-how-we-gave-colors-names-and-it-messed-with-our-brains-part-ii/
How well calibrated is your monitor? My wife recently had a “That’s not what I wanted!” experience after discovering that the brightness levels which are optimal for watching movies are not optimal for designing artwork which looks the same after printing. Unless you’ve done some careful tweaking you might not be learning the colors you think you’re learning.
For sound generation, what operating system are you using? Alsa (at least on Ubuntu 13.10) comes with “speaker-test”, a command line program which can be used to play specified frequency sine waves, and alsa itself can be configured to allow you to save sound output to files.
..And that’s pretty much the story of how at work we ended up with a hideous orange conference table instead of the nice warm brown our department chair envisioned
Calibration is something I don’t learn, at the moment I can more about fine color distinction.
I use windows, but I might write a script and run it on a linux computer to get files, I will think about it.
You’re teaching yourself to associate labels (e.g. “midnight blue”) with particular colors.
You seem to be doing the colors on your monitor and I doubt your monitor is color-calibrated. This implies that you’re looking at a somewhat-randomized version of a true color. Your monitor also has a limited gamut (the set of colors that it can display) and for some monitors the gamut is very limited to the extent that they have to simulate some colors by dithering. Most uncalibrated monitors are too bright and too blue.
Basically, display of proper colors on a screen is complicated and you must take active steps (e.g. color calibration) to make it happen.
Actually the css label is “midnightblue” without a space. For my purposes I don’t think it’s a problem to treat the word true as meaning whatever the hardware interprets it too mean.
I want to get better at precision of color distinction. Having accurate labels is secondary.
Your sentence doesn’t make sense, but keep in mind that we’re talking about what your local, particular, individual hardware interprets it to mean. My hardware (my screen) will likely interpret it differently. Screens of random people will interpret it differently again.
Also a lot of LCD screens use what’s known at TN (twisted nematic) panels and the great majority of them are 6-bit panels, that is, they drive pixels at only 2^6 = 64 levels. So, for example, they can only show you 63 levels of blue (0 is off/black) even though the standard software treatment of color is 8 bits per primary for 256 levels. The TN screens mitigate this problem through dithering.
Human vision is highly adaptive to ambient light, in particular its color temperature. We perceive colors differently in the sunlight, in the shade, or under a tungsten lamp. And these are full-spectrum light sources. Fluorescent lights are not, they have spikes and voids in their spectrum and that affects certain colors under them as well.
As I said, it’s complicated.
Yes, there should have been the word “problem” when I wrote the word “probably”.
I know that human vision is complicated.
I however also know my bit about measurement theory. The important thing isn’t that a measurement is “true” but that that it has features like sensitivity and specificity. Accuracy and precision are other words to speak about measurements.
When dealing with a complex subject the important thing is whether your map of reality is good enough for the purpose for which you want to use it. I think the way my computer models colors is good enough to produce a valid training effect that helps me to get better at distinguishing colors.
I could create a randomly generated filtered deck out of 100 mature Anki color cards and test it at the computer of someone else and see whether the particularities of my specific computer monitor produce a problem. It’s a question that has an empiric answer and in a year when I have put more training into the colors I probably will do something like that. At the moment I have one day 70% correct cards and the next day 85% right cards, so it’s not stable enough for good test.
My notebook has a slightly different color profile than my 24″ monitor. Switching between the two doesn’t produce issues I can perceive. It doesn’t feel like everything is suddenly wrong. My brain seems to be able to make the necessary adjustments that I can still answer the cards correctly.
What do you mean by that?
Let’s use meatspace examples. Do you want to be able to look at a wall and say “this is color X”? Do you want to to be able to look at two walls side by side and say “These are different colors”? Do you want to be able look at a wall, look at another wall the next day and say “This is the same color as I saw yesterday”?
Being able to answer the question: What color is this isn’t very useful. It’s not what separates the person who rather wants to be deaf from the person who rather wants to be deaf. I want is a cure for my partial blindness. I want to perceive more bits of information through the visual channel.
I don’t care for ‘is’ or ‘true’. I’m post- aristotelean. To quote the constructivist Heinz von Foerster: Truth is the invention of a liar. Just in case you think, I’m off-topic, I’m not. Those Anki cards are a result of among other things reading Korzybski’s Science and Sanity.
To go back to questions I want to be able to notice if a website I visit changes their color scheme in a way that exchanges navy with midnightblue just as I’m now able to notice a change from red to green.
Ah. Well, this is useful to know.
Given that your brain’s processing bandwidth is severely limited I think it’s mostly an issue of controlling your attention but experimenting is good.
I don’t think there a good reason to believe that. I think quite often the limiting factor is time spent in deliberate practice and not lack of neurons or similar hardware problems.
Sounds like you want sox. To make an mp3 that plays an A 440 for 1 second you would do:
But note that most real world sounds are a combination of many frequencies, so training on sine waves may not be what you want.
I would think that training on them provides useful skills that generalize more broadly. It’s probably not perfect but it’s easy to create cards with binary choices that can get progressively more difficult.
The goal is getting to a point where I engage into deliberate practice of distinguishing sounds and using Spaced Repetition to do it.
If anyone who’s good at sounds has better ideas about creating a Anki deck to train distinguishing sounds, I would be happy to hear ideas.
I also try to train phonemes, but creating good cards for it proved to be hard. The first cards I created where simply to hard for myself as I’m not good at audio perception.
I can hear a lot more in a Salsa song than I could hear 5 years ago. I think that it’s worthwhile to invest significant time in getting to perceive more bits of reality in daily life. I’m still at the phase of experimenting about how to train myself and others to have richer qualia, but I think it’s an area worthy of further investigation.
Pitch seems to me like a very straightforward concept, but I’m also willing to learn other ways of distinguishing sounds.
Instead of phonemes in isolation, it should work to train on them in words as minimal pairs. For example, to train the difference between /b/ and /d/ you would test discrimination between /bog/ and /dog/, /cab/ and /cad/, /cabby/ and /caddy/, etc.
Under different light contexts an reflective object might be closer to either midnight blue or navy. Have you attempted using paint chips or something to test yourself under sunlight versus florescent light or anything?
Also, for sound perception: sox(1)
I know at the moment it seems to me that the colors are far enough apart that light conditions at my PC are not the main problem.
My notebook is slightly differently configured and I didn’t judge many Anki cards wrongly when answering on the notebook instead of my main monitor.
Most things don’t produce their own color though, they reflect varying amounts of the incoming spectrum. If that incoming spectrum is different, the outgoing spectrum is different. You can take advantage of that in various ways, but it might also confound the question of what color “is” this object.
Or maybe you automatically take that into account by using the ambient light as a reference, I was wondering whether you had tested for that or not?
At the moment I haven’t tested. I spent a total of 7 hours with the average of 5 minutes per day on the Anki cards and I seem to be getter better at color distinction.
Every card provides a binary choice. I have cards that present me with two color words and a large circle that’s filled with the corresponding color. The difference of the colors is at the beginning stage where I’m still at least a total 32 different hex values. I also have cards that ask for the hex value of the colors.
There were some days were I traveled and used my laptop in other light conditions. They weren’t a problem.
For the time in 7 years I created cards to distinguish 4dc636 from 4dc637. That might run into issues with light conditions. I don’t know whether it does and whether the human mind is trainable to distinguish colors as finely, but I will find out if I continue spending my 5 minutes every day.
I don’t like “is” anyway for the reasons Korzybski layd out.
I want to increase the amount of bits I perceive through the visual channel. It’s an open experiment.
The outcome might be that I have color distinction in a few years that allows me to impress people with stunts. I might learn something valuable about colors that can be made into scientific paper or blog post. I also expect that I will get better at usability design even if I don’t get superhuman color perception abilities out of the project.
But you are right that having data about the light conditions would be good. I opened a thread on the QS forum about the search for a proper tool.