1882: "Color Models"

This forum is for the individual discussion thread that goes with each new comic.

Moderators: Moderators General, Prelates, Magistrates

Vroomfundel
Posts: 95
Joined: Wed Sep 29, 2010 7:36 am UTC

1882: "Color Models"

Postby Vroomfundel » Mon Aug 28, 2017 1:14 pm UTC

Image

Alt text: What if what *I* see as blue, *you* see as a slightly different blue because you're using Chrome instead of Firefox and despite a decade of messing with profiles we STILL can't get this right somehow.

Come on people, I keep refreshing with the hope of a learned discussion on subjective colour perception and abstract multidimensional whatever but no one has started it for hours!
lexicum.net - my vocabulary learning platform

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 2492
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1882: "Color Models"

Postby Soupspoon » Mon Aug 28, 2017 1:56 pm UTC

I could readily write reams and reams of purple prose, but to have blew through and bring up inks and yell about tints or arrange the range of scientific terms for why two frequencies are perceived differently... That would be dull...

User avatar
Flumble
Yes Man
Posts: 1945
Joined: Sun Aug 05, 2012 9:35 pm UTC

Re: 1882: "Color Models"

Postby Flumble » Mon Aug 28, 2017 2:14 pm UTC

What you perceive as white and gold is actually blue and black.

User avatar
cellocgw
Posts: 1762
Joined: Sat Jun 21, 2008 7:40 pm UTC

Re: 1882: "Color Models"

Postby cellocgw » Mon Aug 28, 2017 2:14 pm UTC

This is without a doubt the perfect GOOMHR drawing for me.
I remember back as young as 7 or 8 wondering if other people saw lawns in what I thought of as bright crimson, and saw blood as dark green.

Later on, as an optical physicist, I learned all that technical junk.
Fortunately, I stayed out of sales and art, so never cared how each browser rendered each color on each monitor.

Which reminds me: if you happen across some interactive web page GUI whose colors suck, it may be the fault of a former acquaintance whose job is "User Experience" but, being colorblind, has certain problems setting the page contrasts. :P
https://app.box.com/witthoftresume
Former OTTer
Vote cellocgw for President 2020. #ScienceintheWhiteHouse http://cellocgw.wordpress.com
"The Planck length is 3.81779e-33 picas." -- keithl
" Earth weighs almost exactly π milliJupiters" -- what-if #146, note 7

User avatar
Old Bruce
Posts: 57
Joined: Tue Jun 28, 2016 2:27 pm UTC

Re: 1882: "Color Models"

Postby Old Bruce » Mon Aug 28, 2017 2:29 pm UTC

Anthony Burgess was colourblind and therefore his wife was given the job of "dressing" his characters because he knew the words meant different colours and wouldn't see anything wrong with a purple pin-stipe suit paired with a yellow bowler hat being worn by a banker.

morriswalters
Posts: 6904
Joined: Thu Jun 03, 2010 12:21 am UTC

Re: 1882: "Color Models"

Postby morriswalters » Mon Aug 28, 2017 2:31 pm UTC

Ok. You see in light and dark. And It's all about reflection. It's why they work so hard to get a really good black on displays. Nobody sees in red. Light a room with a real monochrome red. You can still see. You're seeing shades of red. What you think about that is up to you. So if I tell you something is red and we aren't using monochrome light you will know a range of colors to look for. On the other hand how do you describe red, when red is all you've got to describe it with?

You have a statistical model for red. As you converge to a point somewhere in the middle the differences become less distinct and then how we describe them becomes much more difficult. That describes the color wheel. Some people are good at this and some aren't. They sometimes call them artists.

Monet was excellent at this. Some artists paint from the shadows to color. You establish the darkest areas and move upward from there. Monet's work, for me, is a case where color trumps shadows. You can master it with practice. if you have that skill. I don't

I assume your statistical model is near mine, and as long as that is true I don't need to worry about it. I do need to worry that your model isn't. We call that color blindness.

Color models are an attempt to radiate, in a way so that both our statistical models understand it in a similar ways. Good luck with that. All displays are not created equal, any more than our visual skills are. And we won't talk about light bulbs.

If I need that exact shade of red I'll dye it until I get what I want, preferably in a well lit room with no direct sunlight. Northern exposures for me. Artists do this in paint.

Randall's strip is about where I'm at. If I don't like the colors in person I will ship it back, or grow to like the color, depending on how important color is to me. Otherwise it ain't my problem. Let's see. Have I covered my major points?

Vroomfundel
Posts: 95
Joined: Wed Sep 29, 2010 7:36 am UTC

Re: 1882: "Color Models"

Postby Vroomfundel » Mon Aug 28, 2017 2:52 pm UTC

Old Bruce wrote:Anthony Burgess was colourblind and therefore his wife was given the job of "dressing" his characters because he knew the words meant different colours and wouldn't see anything wrong with a purple pin-stipe suit paired with a yellow bowler hat being worn by a banker.


Interesting, I also don't see anything wrong with a purple pin-stipe suit paired with a yellow bowler hat being worn by a banker. I think I'm not colourblind though, I think I just like bright colours in all kinds of settings.

How do I know I'm not colourblind, you say? Well, I took a test when I was taking my driving exam but I was also asked by my flatmate how do I know the meat is cooked properly. "Well, easy - when it gets brown it's done" I replied but he said that's of no use for him; that's how I realized that he was colourblind.

Anyway, regarding the my-red-is-your-green controversy, I remember Hofstadter arguing about it in the following fashion: when you hear sound you can clearly distinguish between deep bass and high pitched tones and there's no way in hell somebody's individual perception of tones could be the other way round. The way colour is interpreted by the brain must surely come from some physical correlation of that kind that cannot be reversed in another individual; you might get some difference in the shades but the physics of it should not allow to go all the way from red to green.

Does anyone have any actual scientific insight on the above? It's sound reasoning but vision does not have work in the same way like auditory perception does, the images might actually be conjured out of thin air by the visual cortex with no relation to the actual stimuli. The way naming colours affects their perception, as explored by the BBC Programme Do you see what I see? certainly seems to suggest this.
lexicum.net - my vocabulary learning platform

User avatar
Old Bruce
Posts: 57
Joined: Tue Jun 28, 2016 2:27 pm UTC

Re: 1882: "Color Models"

Postby Old Bruce » Mon Aug 28, 2017 3:36 pm UTC

Vroomfundel wrote:
Old Bruce wrote:Anthony Burgess was colourblind and therefore his wife was given the job of "dressing" his characters because he knew the words meant different colours and wouldn't see anything wrong with a purple pin-stipe suit paired with a yellow bowler hat being worn by a banker.


Interesting, I also don't see anything wrong with a purple pin-stipe suit paired with a yellow bowler hat being worn by a banker. I think I'm not colourblind though, I think I just like bright colours in all kinds of settings.

How do I know I'm not colourblind, you say? Well, I took a test when I was taking my driving exam but I was also asked by my flatmate how do I know the meat is cooked properly. "Well, easy - when it gets brown it's done" I replied but he said that's of no use for him; that's how I realized that he was colourblind.

Anyway, regarding the my-red-is-your-green controversy, I remember Hofstadter arguing about it in the following fashion: when you hear sound you can clearly distinguish between deep bass and high pitched tones and there's no way in hell somebody's individual perception of tones could be the other way round. The way colour is interpreted by the brain must surely come from some physical correlation of that kind that cannot be reversed in another individual; you might get some difference in the shades but the physics of it should not allow to go all the way from red to green.

Does anyone have any actual scientific insight on the above? It's sound reasoning but vision does not have work in the same way like auditory perception does, the images might actually be conjured out of thin air by the visual cortex with no relation to the actual stimuli. The way naming colours affects their perception, as explored by the BBC Programme Do you see what I see? certainly seems to suggest this.

Naming colours... I submit for your consideration: Wine dark sea.
What the fuck?

rmsgrey
Posts: 3083
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1882: "Color Models"

Postby rmsgrey » Mon Aug 28, 2017 4:57 pm UTC

Vroomfundel wrote:Anyway, regarding the my-red-is-your-green controversy, I remember Hofstadter arguing about it in the following fashion: when you hear sound you can clearly distinguish between deep bass and high pitched tones and there's no way in hell somebody's individual perception of tones could be the other way round. The way colour is interpreted by the brain must surely come from some physical correlation of that kind that cannot be reversed in another individual; you might get some difference in the shades but the physics of it should not allow to go all the way from red to green.

Does anyone have any actual scientific insight on the above? It's sound reasoning but vision does not have work in the same way like auditory perception does, the images might actually be conjured out of thin air by the visual cortex with no relation to the actual stimuli. The way naming colours affects their perception, as explored by the BBC Programme Do you see what I see? certainly seems to suggest this.


When it comes to perception, there are (at least) three things happening:

1) Some physical phenomenon occurs in our sensory organs, triggering electrochemical responses. This is going to be pretty similar for most people - divergence at this point manifests as things like colour-blindness or actual blindness or the ability to see two distinct reds (or similar impairment or advantage to other senses)

2) Our minds interpret electrochemical patterns in the brain as subjective experiences of sensations. My subjective sensation of "red" may be the same as your subjective sensation of "green" - there's no real way of knowing for an individual sensation - and it doesn't really matter anyway. What's more significant is the relationships between various sensations. Reds and oranges should give rise to sensations that are more similar to each other than either is to the sensations of blues, and similarly for sounds and so on. A big hint here would be looking at how far synaesthetes experience the same sets of sensations as matching - or does one match red to a C, while another matches it to F-sharp?

3) We conceptualise those subjective sensations as symbols. This is the point where we can most readily compare, but it also offers minimal insight into the subjective experiences involved. When I look at a ripe banana, I experience something that I then call "yellow" (or "the experience of looking at something yellow") and, ideally, so does everyone else. If someone looks at the same banana, at the same time, and describes it as "purple" then we know there's something wrong - either we're speaking different languages, or at least one of us is suffering from a failure somewhere in the sequence - whether that's in the original sensory event, the subjective sensation, or the classification. That or there's some sort of weird optical illusion or lighting effect going on like with that dress...

User avatar
ManaUser
Posts: 284
Joined: Mon Jun 09, 2008 9:28 pm UTC

Re: 1882: "Color Models"

Postby ManaUser » Mon Aug 28, 2017 5:06 pm UTC

Now we know why XKCD is (almost always) black and white.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 2492
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1882: "Color Models"

Postby Soupspoon » Mon Aug 28, 2017 5:15 pm UTC

((Wine Dark Sea))
That was cultural. (And may rotate around such as https://eagereyes.org/blog/2011/you-onl ... u-can-name and related issues...)

I'm currently looking out of a window at a blue sky, a poster with two distinct blues (each graded, one a drawn sky, the other just 'background', neither overlapping), a building with a blue upper-layer of cladding, a shop hoarding with a blue-patterned background to the name, and a car parked (darn, just moved off!) that's also blue. And they're all different hues (or sufficiently different luminosities/saturations, to make the hues incomparable) of blue.

At a push, I'd have said the car had a bit of redness in there (might have been a irridescant metallicness to the paint job), the cladding is heading towards turquoise, etc, but I wouldn't feel comfortable taking away "that's purplish/turquoise blue" and cross-comparing with some other object later in the day. Even with a vague feeling that I sort of have the words. And reflectivity of wine(-rouge) under a blue sky might stir similar perceptions as a darkened sea under twilight skies, conceivably.

(And when I punned, above, and went searching for the Puce hexadecimal and found out it has two distinct points in different parts of the Anglosphere. Well, makes me wonder if Dudley Moore was talking about the one I know about these days, in my first ever exposure to the word "puce". A kind of purple, but not sure which kind of purple, now.)

User avatar
Steve the Pocket
Posts: 645
Joined: Mon Apr 23, 2007 4:02 am UTC
Location: Going downtuuu in a Luleelurah!

Re: 1882: "Color Models"

Postby Steve the Pocket » Mon Aug 28, 2017 6:24 pm UTC

I hate the concept of color profiles, and I especially hate that we invoke them at every single step of the process between the file's raw RGB data and what actually reaches our eyes. The file itself, the program that displays it, the video driver, the monitor—they all have color profiles that distort the source data a little bit more. And if we're very, very lucky, they'll even out.

I remember a time when it was virtually impossible to guarantee a match between PNGs being displayed inside a web page and blocks of solid color defined through CSS (background colors, borders, etc.) But GIFs, which didn't support embedded profiles, worked fine, so everyone was stuck using those even though every browser supported PNG just fine.
cephalopod9 wrote:Only on Xkcd can you start a topic involving Hitler and people spend the better part of half a dozen pages arguing about the quality of Operating Systems.

Baige.

User avatar
Znirk
Posts: 168
Joined: Mon Jul 01, 2013 9:47 am UTC
Location: ZZ9 plural Z α

Re: 1882: "Color Models"

Postby Znirk » Mon Aug 28, 2017 7:09 pm UTC

At least we're not stuck on Ohjann Golgo van Fontheweg. As Gofid Letterkerl described it: "Seine Romane sind ja schon schwer auszuhalten, aber seine Mineralfarbenlehre kann einem wirklich die Hirnhaut lösen."

("His novels are hard enough to bear as it is, but his Mineral Theory of Colour can really peel off your meninges.")

Spoiler:
Walter Moers, Die Stadt der träumenden Bücher / The City of Dreaming Books, alluding to the confident bullshit of Johann Wolfgang von Goethe trying to do physics.

morriswalters
Posts: 6904
Joined: Thu Jun 03, 2010 12:21 am UTC

Re: 1882: "Color Models"

Postby morriswalters » Mon Aug 28, 2017 7:36 pm UTC

The Deep Blue Good-By
One Fearful Yellow Eye
A Tan and Sandy Silence
The Lonely Silver Rain

His roommate answered his question. He perceives color differently.

Cameras pick up color, and because they are good at reproducing what they see, they produce pretty pictures. So we understand sight well enough. It like a camera or something close to it. The trick is putting it all together. And the brain is pretty good at inventing details out of little data. You have two blind spots in your eye, and you never see them.

Farabor
Posts: 147
Joined: Wed Apr 06, 2011 9:46 am UTC

Re: 1882: "Color Models"

Postby Farabor » Mon Aug 28, 2017 7:41 pm UTC

I can't believe we're this far in without anyone mentioning https://blog.xkcd.com/2010/05/03/color-survey-results/

User avatar
rhhardin
Posts: 51
Joined: Fri Apr 09, 2010 2:11 pm UTC

Re: 1882: "Color Models"

Postby rhhardin » Mon Aug 28, 2017 8:28 pm UTC

There are also philosophical colors, like "grue," which meant blue before the year 2000 and green thereafter.

It questioned whether there is sufficient evidence of sky color.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 2492
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1882: "Color Models"

Postby Soupspoon » Mon Aug 28, 2017 8:35 pm UTC

Surely grue is pitch black. (And likely to eat you!)

User avatar
chridd
Has a vermicelli title
Posts: 765
Joined: Tue Aug 19, 2008 10:07 am UTC
Location: ...Earth, I guess?
Contact:

Re: 1882: "Color Models"

Postby chridd » Mon Aug 28, 2017 10:07 pm UTC

Code: Select all

if(new Date().getFullYear() < 2000)
   print("It's pitch black.  You are likely to be eaten by a blue.");
else
   print("It's pitch black.  You are likely to be eaten by a green.");
~ chri d. d. /tʃɹɪ.di.di/ (Phonotactics, schmphonotactics) · they (for now, at least) · Forum game scores
mittfh wrote:I wish this post was very quotable...
flicky1991 wrote:In both cases the quote is "I'm being quoted too much!"

User avatar
Heimhenge
Posts: 167
Joined: Thu May 01, 2014 11:35 pm UTC

Re: 1882: "Color Models"

Postby Heimhenge » Mon Aug 28, 2017 10:18 pm UTC

Even before Randall's earliest impressions in grade school, to me, colors were what the labels said on my cheap box of 8 Crayolas. :D

kelly_holden
Posts: 32
Joined: Wed Apr 23, 2014 7:27 am UTC
Location: Rural New South Wales

Re: 1882: "Color Models"

Postby kelly_holden » Mon Aug 28, 2017 10:47 pm UTC

Steve the Pocket wrote:I remember a time when it was virtually impossible to guarantee a match between PNGs being displayed inside a web page and blocks of solid color defined through CSS (background colors, borders, etc.) But GIFs, which didn't support embedded profiles, worked fine, so everyone was stuck using those even though every browser supported PNG just fine.

I didn't know the technical details of what was causing it, but I remember struggling with this while taking a web design class, and eventually giving up and making the background "colour" a tiled 1px image.

rmsgrey
Posts: 3083
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1882: "Color Models"

Postby rmsgrey » Tue Aug 29, 2017 12:50 am UTC

rhhardin wrote:There are also philosophical colors, like "grue," which meant blue before the year 2000 and green thereafter.

It questioned whether there is sufficient evidence of sky color.


No, grue is green before and blue after. You're thinking of bleen.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 2492
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1882: "Color Models"

Postby Soupspoon » Tue Aug 29, 2017 1:19 am UTC

kelly_holden wrote:I didn't know the technical details of what was causing it, but I remember struggling with this while taking a web design class, and eventually giving up and making the background "colour" a tiled 1px image.

That sounds horrendous, actually. That you felt the need to do that, mostly, but that I know you did it at all makes me shudder, too.

(Still, better than the (loosely describable as) 'colleague' of mine whose answer to control-freaking the web-page layout was to create images sized to the (presumed!) browser portal size, while using image-map linking across the areas of the layout that were 'link text' - but still just image pixels, naturally. All except the "contact us" page that had actually working HTML <form> elements and <input>s of various kinds, placed upon the page background with absolute-coordinate positioning - but defined in an utterly illogical order (or, rather, as had arisen by the creation/copying/cloning, plus recycling and re-use, that had happened during design time), as anybody who was tempted to tab between fields instead of mousing it would have discovered. Argh! Definitely failed any Accessibility test you might have asked for.)

Obviously not every corporation is going to be content or able to start with (or resignedly resort to) only the web-safe pallette in their logos and other branding, but there comes a point where you have to accept that you do your best with any decent browser that complies with Acid3 (or relevant contemporary), doesn't horribly fall over in the noted still-popular legacy browsers and plus are still usable in Lynx, via various screen-readers/refreshable braille-'displays', and anything else you or your compliance department can think of at the time.

But too much effort is odten put into pixel-perfection. If your page doesn't make (some sort of) sense if the external .css doesn't load or the .js is blocked or jams up or the browser is told not to automatically download images, then you need to be happy that you're disenfranchising people. (Who might not have the option of "downloading the latest version of <recommended browser> here", however much you overlay/replace the page with your well-meaning admonishment for apparently being a technoluddite…)

/rant over. And just a generic rant, totally not aimed at you, Kelly. Just hit a nerve and aroused my long-standing annoyance with a 'certain kind' of web developer. (Some of whom may be reading this, and may have arguments in return, perhaps even valid ones in their own particular circumstances!)

User avatar
Pfhorrest
Posts: 3914
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1882: "Color Models"

Postby Pfhorrest » Tue Aug 29, 2017 1:55 am UTC

In my past jobs as a web developer (I do graphic design now instead), and searching for further work in it, my soul has been somewhat crushed by the overwhelming impression that nobody I would ever work for gave a flying fuck about any of that, and basically wanted me to make everything pixel-perfect despite any other goal (like the many good ones you lay out) that that might foil.

I want to (and believe people should) start a website with a plain text document, then mark it up for different levels of headers and so on until you have a basic unstyled HTML document that is usable for the site's intended purpose, and THEN apply styles to that to make it look pretty and be more easily navigable. Everyone else always seemed to want me to basically start in Photoshop and then work backward to figure out what code will render something that looks like that Photoshop image but load faster and be clickable. Ugh.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

User avatar
ucim
Posts: 5570
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1882: "Color Models"

Postby ucim » Tue Aug 29, 2017 3:13 am UTC

Does anybody remember Front Page?

(slinks out)

My work here is done.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

User avatar
JohnTheWysard
Posts: 88
Joined: Sun Feb 10, 2008 2:38 am UTC

Re: 1882: "Color Models"

Postby JohnTheWysard » Tue Aug 29, 2017 3:54 am UTC

Randall, you have NO business posting an xkcd comic on the subject of color. The strip is black and white and that's as it should be!

User avatar
Steve the Pocket
Posts: 645
Joined: Mon Apr 23, 2007 4:02 am UTC
Location: Going downtuuu in a Luleelurah!

Re: 1882: "Color Models"

Postby Steve the Pocket » Tue Aug 29, 2017 6:39 am UTC

Pfhorrest wrote:In my past jobs as a web developer (I do graphic design now instead), and searching for further work in it, my soul has been somewhat crushed by the overwhelming impression that nobody I would ever work for gave a flying fuck about any of that, and basically wanted me to make everything pixel-perfect despite any other goal (like the many good ones you lay out) that that might foil.

I want to (and believe people should) start a website with a plain text document, then mark it up for different levels of headers and so on until you have a basic unstyled HTML document that is usable for the site's intended purpose, and THEN apply styles to that to make it look pretty and be more easily navigable. Everyone else always seemed to want me to basically start in Photoshop and then work backward to figure out what code will render something that looks like that Photoshop image but load faster and be clickable. Ugh.

The design firm we were partnered with operated the second way, giving me a PSD with a zillion layers and wishing me luck. I don't blame them for it because that was the workflow they used for everything else—brochures and the like—and from a purely graphic design perspective, they did good work. I simply took the complexity as a challenge, and they never had an issue with the amount of time they were being billed for. I recall a couple times I intentionally put elements "out of order" relative to their positions on the screen, and then used a CSS trick (exploiting a quirk in the way it handles the flow of differently-sized <DIV>s, if I recall correctly) to switch them back around, because it would flow more logically on a CSS-free ("naked", I believe is the industry term) version that way.

I think my favorite moment was when I was handed a layout where the logo—which needed to be a PNG—had an outer glow set to Screen blending mode on top of a photo background. Naturally, I couldn't include the outer glow in the PNG because then the blending mode would break. On a whim, I Googled whether it's possible to split the outer glow (which was achieved using a layer effect) into its own separate layer. As luck would have it, yes, for some reason, that's a real feature. I ended up doing that and then making it part of the background JPEG.
cephalopod9 wrote:Only on Xkcd can you start a topic involving Hitler and people spend the better part of half a dozen pages arguing about the quality of Operating Systems.

Baige.

User avatar
Wee Red Bird
Posts: 141
Joined: Wed Apr 24, 2013 11:50 am UTC
Location: In a tree

Re: 1882: "Color Models"

Postby Wee Red Bird » Tue Aug 29, 2017 7:18 am UTC

Alt text: What if what *I* see as blue, *you* see as a slightly different blue because you're using Chrome instead of Firefox and despite a decade of messing with profiles we STILL can't get this right somehow.

Or when you have a second screen attached to your laptop and your Excel spreadsheet looks orange on one screen but yellow on the other.
There are settings on the monitor I can dick about with to make them match on both screens, but one if the monitor was right and the laptop screen was wrong, but it doesn't come with the same adjustments as the monitor.

User avatar
roderik
Posts: 22
Joined: Fri Apr 10, 2009 4:16 pm UTC
Location: Below sealevel

Re: 1882: "Color Models"

Postby roderik » Tue Aug 29, 2017 9:39 am UTC

Steve the Pocket wrote:I hate the concept of color profiles, and I especially hate that we invoke them at every single step of the process between the file's raw RGB data and what actually reaches our eyes. The file itself, the program that displays it, the video driver, the monitor—they all have color profiles that distort the source data a little bit more. And if we're very, very lucky, they'll even out.

I remember a time when it was virtually impossible to guarantee a match between PNGs being displayed inside a web page and blocks of solid color defined through CSS (background colors, borders, etc.) But GIFs, which didn't support embedded profiles, worked fine, so everyone was stuck using those even though every browser supported PNG just fine.

ai... but isn't the point of those profiles to preserve the actual colours of the file, so that you know for sure that the values given in the data, are the values for the colours intended, despite variations in monitors and all that crap.

morriswalters
Posts: 6904
Joined: Thu Jun 03, 2010 12:21 am UTC

Re: 1882: "Color Models"

Postby morriswalters » Tue Aug 29, 2017 10:15 am UTC

roderik wrote:ai... but isn't the point of those profiles to preserve the actual colours of the file, so that you know for sure that the values given in the data, are the values for the colours intended, despite variations in monitors and all that crap.
One might think so. If you had done this pre web, pages would have been presented in printers proofs, or galleys or some such. The art directors and editors would have been able to look at what they they were going to see on the newsstand. I have some sympathy for web developers.

User avatar
Pfhorrest
Posts: 3914
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1882: "Color Models"

Postby Pfhorrest » Tue Aug 29, 2017 3:24 pm UTC

roderik wrote:ai... but isn't the point of those profiles to preserve the actual colours of the file, so that you know for sure that the values given in the data, are the values for the colours intended, despite variations in monitors and all that crap.

The point of the profiles it to try to assure that the color seen in a file on one device looks the same for the same file on a different device.

Where that whole thing falls down is that it fails to understand that many people want (with good reason) to use RGB values as spot colors, and so nothing should be fucking with the colors, it should be preserved exactly as the numbers say.

If I went into a paint shop and picked out a Pantone 360 C green for the color I wanted the outside trim of my house painted, but the painters then took that and compensated for the difference between the warm lighting inside the store and the cool lighting outdoors and instead of using 360 used a yellower green that looked in cool outdoor lighting like the 360 looked in the warm indoor lighting, heads would fucking roll. When I tell you I want it Pantone 360 C you paint it that fucking color, not something else.

That's the equivalent of what color profiles embedded in files are trying to do, and they need to knock it the fuck off. Leave it up to the artist and the viewer to make sure that their monitors are correctly calibrated (the equivalent of making sure to look at your paint chips under diffuse neutral lighting to gauge their true color), and when the artist puts a 127,255,63 blob on their canvas, fucking display 127,255,63 on the viewer's screen however that color looks on their screen and if their screen makes it look wrong, that's their own goddamn fault.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

speising
Posts: 2070
Joined: Mon Sep 03, 2012 4:54 pm UTC
Location: wien

Re: 1882: "Color Models"

Postby speising » Tue Aug 29, 2017 4:48 pm UTC

127,255,63 and Pantone 360 C are completely different concepts. One is a coordinate in a colour space, and thus completely meaningless without defining that space. The other defines a specific colour.

User avatar
Sableagle
Ormurinn's Alt
Posts: 1201
Joined: Sat Jun 13, 2015 4:26 pm UTC
Location: The wrong side of the mirror
Contact:

Re: 1882: "Color Models"

Postby Sableagle » Tue Aug 29, 2017 5:43 pm UTC

Old Bruce wrote:Anthony Burgess was colourblind and therefore his wife was given the job of "dressing" his characters because he knew the words meant different colours and wouldn't see anything wrong with a purple pin-stipe suit paired with a yellow bowler hat being worn by a banker.


Image

...

Classic FM over here have at least one presenter with a bad habit of praising a composer's "use of colour" in a musical work.
It's not as bad as the Radio 3 presenter who used to tell us that "the performance ..... we're going to hear ..... today ..... is performed ..... by the Czech Philharmonic Orchestra ..... with the position ..... of second violin ..... filled by Anatolia Depravera Svetalova ..... the great-grand-daughter ..... of the sister ..... of Mikhail Tokarev ..... who conducted ..... the premiere ..... at the Pushkin Palace ..... in 1902 ..... on the occasion ..... of the Tsar's ..... second daughter's ..... fifth birthday," but it's still kind of annoying when you suspect he's talking shit.
Of course, Radio 3 is owned by the same company as Radio1 and Classic's owned by the same company as Capital, so I could cope with losing both just to have the other two silenced forever, but that's beside the point.
I looked up this idea of "colour" in music, wondering whether it could, perhaps, be a real thing, an actual term that afficionados would know, perhaps something about, well, something. Turns out it refers to the use of different tones.
Tones.
We used to call those notes.
Kind of hard to write good orchestral music without using them, really.
Wonderful use of different notes? Great. That's kind of why pianos have so many keys and guitars have all those strings and all those ridges on the neck and pipe organs have more than one pipe each and flutes have holes, you know. It's ...
Yeah, I got sick of hearing about "colour" in music. It's like a foodie praising a chef's "wonderful use of ingredients in this meal."
Oh, Willie McBride, it was all done in vain.

User avatar
ucim
Posts: 5570
Joined: Fri Sep 28, 2012 3:23 pm UTC
Location: The One True Thread

Re: 1882: "Color Models"

Postby ucim » Tue Aug 29, 2017 6:25 pm UTC

Sableagle wrote:Tones.
We used to call those notes.
No. Notes are distinguished by their fundamental frequency (and sometimes also by duration). Tones are distinguished by their harmonics. "Tonal color" refers to the distribution (in time and intensity) of the harmonics in the note. It's what makes a guitar F-flat sound different from a flute F-flat in the same octave.

Jose
Order of the Sillies, Honoris Causam - bestowed by charlie_grumbles on NP 859 * OTTscar winner: Wordsmith - bestowed by yappobiscuts and the OTT on NP 1832 * Ecclesiastical Calendar of the Order of the Holy Contradiction * Please help addams if you can. She needs all of us.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 2492
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1882: "Color Models"

Postby Soupspoon » Tue Aug 29, 2017 6:45 pm UTC

Sableagle wrote:I looked up this idea of "colour" in music, wondering whether it could, perhaps, be a real thing, an actual term that afficionados would know, perhaps something about, well, something. Turns out it refers to the use of different tones.
Tones.
We used to call those notes.
Kind of hard to write good orchestral music without using them, really.
Wonderful use of different notes? Great. That's kind of why pianos have so many keys and guitars have all those strings and all those ridges on the neck and pipe organs have more than one pipe each and flutes have holes, you know. It's ...
Yeah, I got sick of hearing about "colour" in music. It's like a foodie praising a chef's "wonderful use of ingredients in this meal."

Could this come from (if not within the commentator concerned, the ancestral influences upon his apparent terminological inexactitudes) a case or trend of synæsthesia, perhaps?

Similar also maybe to some chords in the western music system being keyed to certain emotions in their usual western audience, seemingly without explicit education. (Though as they do not elicit the same feelings amongst those used to non-western audiences, with reciprocal disjoints as well, that may well be due to general immersion in the base 'cultural environment'.)

Except that the harmonic qualities of a different instrument can make the same nominal note mean something different (as under control of the orchestral composer), and the 'fingering' or equivalent handling of the music by the musician (in their precise manner of blowing, hitting and/or scraping, to paraphrase the Young Person's Guide) can add nuance to the output beyond merely obeying a "fortissimo" mark upon the sheet music, after factoring in the additional interpretation demanded by musician or conductor in how to treat the work.

So "colour" can arise from those many and various opportunities to (if you'll excuse the alternate synæsthetistic term) 'flavour' the piece, leading to commentary for a piece that would be plainly different if played upon a guitar, Hammond organ, recorder, BBC Micro's SOUND command, etc, etc...

ilvos01
Posts: 1
Joined: Sat Jun 09, 2012 6:13 am UTC

Re: 1882: "Color Models"

Postby ilvos01 » Tue Aug 29, 2017 7:08 pm UTC

A lot of my theory work stems from the subjective interpretation of sensory phenomena. Ultimately it doesn't matter what color you see, so long as you have been raised to identify it. The only way one could convey a difference is as a relationship between two colors, if your particular color palette makes them more or less distinguishable to you than "baseline". That's why most forms of color-blindness are described with two colors. But even then, there's doubt. If your vision was entirely flipped to white-on-black, warm-to-cool, we would have no way to tell, as your "red" is equally far from your "blue" as everyone else's. Even two colorblind people with the same "type" might see different "colors", so long as their wavelength-palette is compressed in a similar fashion so as to be observable in testing. Because color is still, ultimately, "trained up", as we have fairly well assigned color-labels to ranges of wavelength. 475 nanometres is Blue, no matter if you see "blue" or not.
Maybe this has already been said, but I still wanted to say it.

User avatar
orthogon
Posts: 2697
Joined: Thu May 17, 2012 7:52 am UTC
Location: The Airy 1830 ellipsoid

Re: 1882: "Color Models"

Postby orthogon » Tue Aug 29, 2017 7:13 pm UTC

It's just an attempt to eff the ineffable, isn't it? Musical colour is just musical colour. The presenter is grasping for a visual metaphor for something, because we all at least somewhat agree on what colour is, that it can carry or stimulate emotion, and so on. In geek terms, I think it's at a higher "layer" than things like notes or timbre. It's like how you might talk about "tactics" in a game of sportsball, or "philosophy" of a company. We're talking about The Arts here, not physics.
xtifr wrote:... and orthogon merely sounds undecided.

User avatar
Soupspoon
You have done something you shouldn't. Or are about to.
Posts: 2492
Joined: Thu Jan 28, 2016 7:00 pm UTC
Location: 53-1

Re: 1882: "Color Models"

Postby Soupspoon » Tue Aug 29, 2017 7:41 pm UTC

Quarks are physics. Probably. ;)

User avatar
Steve the Pocket
Posts: 645
Joined: Mon Apr 23, 2007 4:02 am UTC
Location: Going downtuuu in a Luleelurah!

Re: 1882: "Color Models"

Postby Steve the Pocket » Wed Aug 30, 2017 1:10 am UTC

speising wrote:127,255,63 and Pantone 360 C are completely different concepts. One is a coordinate in a colour space, and thus completely meaningless without defining that space. The other defines a specific colour.

Well, the assumption is that it means the amount of "pure" red, green, and blue received by the rods (cones?) in the human eye. If cameras, scanners, or monitors aren't properly matching those values, then that should be treated as a deficiency of the equipment. If the manufacturers are capable of pinpointing that deficiency so precisely that they can program a color profile that compensates for it, they ought to be capable of just fixing it in the first place. But at the very least, either way, you can be assured that it's consistently inaccurate at reproducing color. Once you start assigning color profiles to different images and videos, let alone coding different programs to use different profiles by default, all bets are off and you're in crazytown.
cephalopod9 wrote:Only on Xkcd can you start a topic involving Hitler and people spend the better part of half a dozen pages arguing about the quality of Operating Systems.

Baige.

Eutychus
Posts: 431
Joined: Mon Jan 25, 2010 6:01 am UTC
Location: France

Re: 1882: "Color Models"

Postby Eutychus » Wed Aug 30, 2017 8:55 am UTC

The dress is blue.
Be very careful about rectilinear assumptions. Raptors could be hiding there - ucim

Kit.
Posts: 1020
Joined: Thu Jun 16, 2011 5:14 pm UTC

Re: 1882: "Color Models"

Postby Kit. » Wed Aug 30, 2017 1:20 pm UTC

Steve the Pocket wrote:I hate the concept of color profiles, and I especially hate that we invoke them at every single step of the process between the file's raw RGB data and what actually reaches our eyes. The file itself, the program that displays it, the video driver, the monitor—they all have color profiles that distort the source data a little bit more.

That's not "every" step.

The "distortion", ideally, should happen only twice: once at conversion from input space to working space, and once at conversion from working space to output space. Both conversions should be end-to-end calibrated.

In addition, there could be multiple steps happening in the working space itself.

Steve the Pocket wrote:And if we're very, very lucky, they'll even out.

That's typically even not theoretically possible, because of the differences between color gamuts of input and output devices. No matter which procedure you use to convert one space to another.

Steve the Pocket wrote:I remember a time when it was virtually impossible to guarantee a match between PNGs being displayed inside a web page and blocks of solid color defined through CSS (background colors, borders, etc.) But GIFs, which didn't support embedded profiles, worked fine, so everyone was stuck using those even though every browser supported PNG just fine.

GIFs only have 8 bits of color information per color sample. No one expects faithful color reproduction from them anyway.

Pfhorrest wrote:Where that whole thing falls down is that it fails to understand that many people want (with good reason) to use RGB values as spot colors

Oh, and some people want free energy from a perpetual motion machine.

Pfhorrest wrote:If I went into a paint shop and picked out a Pantone 360 C green for the color I wanted the outside trim of my house painted,

That means that you specify how the color should be produced. Not how you observed it in the paint shop. That might correspond to some proportions (numbers) of the standard pigments added to a standard base paint (usually not exactly 100% white by itself). However, if you apply the same numbers to slightly different pigments or to a slightly different base, you will end up with a different color.

Steve the Pocket wrote:Well, the assumption is that it means the amount of "pure" red, green, and blue received by the rods (cones?) in the human eye.

This assumption is based on a faulty understanding how human vision works.


Return to “Individual XKCD Comic Threads”

Who is online

Users browsing this forum: No registered users and 34 guests