Color blindness and it affects on GUIs
Tomas Johnson
- Statistics, do really one out of ten have problems with this? Are there different levels of the defect?
- What color combinations (as read on the screen) are to be avoided?
- Visual design rules due to color blindness?
Regards
/Tomas Johnson
Jon
>"Tomas Johnson" <tomas....@ericsson.com> wrote in message
>news:3C3C3AC8...@ericsson.com...
A comprehensive guide can be found at:
www.btexact.com/ces/colours
Try the PDF document for complete desciptions and a 'safe' colour palette.
Hope that helps, and sorry for top posting, something to do with the HTML
format of the original post???
Jon
Jon
news:a1hkkk$lhs$1...@pheidippides.axion.bt.co.uk...
Obviously I fixed the top posting problem.... :-)
Paul Schlyter
Tomas Johnson <tomas....@ericsson.com> wrote:
> "Don't use colors as the only information carrier since up to ten per
> cent of the male population has any kind of color reading defects".
> Quite often heard so I am looking for more inormation on this:
>
No. One of ten males makes one of twento of the general population.
> Are there different levels of the defect?
total color blindness - the latter is quite rare though. The most
common form of color blindness is inability to distinguish reds and
greens from one another.
> * What color combinations (as read on the screen) are to be avoided?
Red and green of course... :-)
In general, make sure you have a quite visible BRIGHTNESS difference
as information carrier -- i.e. design your GUI so it's usable also on
black-and-white monitor (yes, ther are still some of those). Then
use colors for aesthetic effects, or to enhance the information already
visible if the screen is viewed in black-and-white.
> * Visual design rules due to color blindness?
Simple: use BRIGHTNESS differences to display information.
--
----------------------------------------------------------------
Paul Schlyter, Swedish Amateur Astronomer's Society (SAAF)
Grev Turegatan 40, S-114 38 Stockholm, SWEDEN
e-mail: pausch at saaf dot se
WWW: http://hem.passagen.se/pausch/index.html
http://home.tiscali.se/~pausch/
Peter Hill
First of all the typical stats quoted for "colour blindness" are for
congenital (i.e., from birth) defects. These figures do not account for
acquired colour perception defects.
1) With respect to the following quotes:
> > "Don't use colors as the only information carrier since up to ten per
> > cent of the male population has any kind of color reading defects".
> > Quite often heard so I am looking for more inormation on this:
First of all, I don't think any designer worth anything would code
information solely with colour...it just makes no sense. Furthermore, it is
extremely to code solely on colour...it takes some calculation...you would
have to calibrate all other factors, including contrast, brightness, size,
etc.
The statistic misused above is for congenital (i.e., from birth) colour
perception defects...these only relate to pure colours...very few colours on
web pages and in the natural environment are so pure as to make
discrimination difficult or impossible. A sure way to avoid this issue, if
it is really an issue at all, is to redundantly code using a variety of
techniques, e.g., shape, brightness, hue, movement. The appropriate
combination of factors will be dependent on the particular
application...sorry, no simple answer here...that's why we have
designers...(smile).
The other thing to consider is that this statistic doesn't include people
with acquired colour perception deficiencies as occur in conditions like
macular degeneration.
> > Are there different levels of the defect?
>
> Yes - there are all levels from somewhat anomalous color vision to
> total color blindness - the latter is quite rare though. The most
> common form of color blindness is inability to distinguish reds and
> greens from one another.
Again, I would be careful with this quote. The poster is only referring to
congenital colour perception defects here...there are a myriad of acquired
conditions as well.
>
> > * What color combinations (as read on the screen) are to be avoided?
>
> Red and green of course... :-)
I strongly disagree with this statement...this assumes the colours are
equated along all other dimensions required for colour discrimination. The
issue here is discrimination of things...hue being one factor...brightness,
etc., etc., etc.
>
> In general, make sure you have a quite visible BRIGHTNESS difference
> as information carrier -- i.e. design your GUI so it's usable also on
> black-and-white monitor (yes, ther are still some of those). Then
> use colors for aesthetic effects, or to enhance the information already
> visible if the screen is viewed in black-and-white.
Again, what's important is ability to discriminate. Contrast (as opposed to
absolute brightness) is one of the good tools which can aid in facilitation
of discrimination between two objects, however, there are many people with
contrast sensitivity perception deficits in the population as well,
especially in the aging population. Spatial contrast sensitivity deficits
are way more common in the general population and are much more problematic
than colour perception deficits.
What you need is a good text on colour perception...I would recommend
searching for some good basic medical and psychology texts.
If you're interested in further exploring congenital and acquired colour
deficits, search for "Farnsworth-Munsell"...in particular...have a look at
the "FM-100" colour perception test.
If you have any other specific questions I'd be more than happy to answer
them if I can...or point you to a reference that will have the answer.
Regards,
Peter Hill
Peter Hill
First of all the typical stats quoted for "colour blindness" are for
congenital (i.e., from birth) defects. These figures do not account for
acquired colour perception defects.
1) With respect to the following quotes:
> > cent of the male population has any kind of color reading defects".
> > Quite often heard so I am looking for more inormation on this:
information solely with colour...it just makes no sense. Furthermore, it is
extremely to code solely on colour...it takes some calculation...you would
have to calibrate all other factors, including contrast, brightness, size,
etc.
The statistic misused above is for congenital (i.e., from birth) colour
perception defects...these only relate to pure colours...very few colours on
web pages and in the natural environment are so pure as to make
discrimination difficult or impossible. A sure way to avoid this issue, if
it is really an issue at all, is to redundantly code using a variety of
techniques, e.g., shape, brightness, hue, movement. The appropriate
combination of factors will be dependent on the particular
application...sorry, no simple answer here...that's why we have
designers...(smile).
The other thing to consider is that this statistic doesn't include people
with acquired colour perception deficiencies as occur in conditions like
macular degeneration.
>
"Tomas Johnson" <tomas....@ericsson.com> wrote in message news:3C3C3AC8...@ericsson.com...
Paul Schlyter
Peter Hill <peter...@sympatico.ca> wrote:
> There are several issues to consider with respect to colour blindness.
> First of all the typical stats quoted for "colour blindness" are for
> congenital (i.e., from birth) defects. These figures do not account for
> acquired colour perception defects.
>
> 1) With respect to the following quotes:
>
>>> "Don't use colors as the only information carrier since up to ten per
>>> cent of the male population has any kind of color reading defects".
>>> Quite often heard so I am looking for more inormation on this:
>
> First of all, I don't think any designer worth anything would code
> information solely with colour...it just makes no sense.
capacitors are usually coded with colored bands on the component.
Also, signals for trains and airplanes use a lot of color - and one
requirement for an engine driver or an airplane pilot is to not be
color-blind.
>>> * What color combinations (as read on the screen) are to be avoided?
>>
>> Red and green of course... :-)
>
> I strongly disagree with this statement...this assumes the colours are
> equated along all other dimensions required for colour discrimination. The
> issue here is discrimination of things...hue being one factor...brightness,
> etc., etc., etc.
small details. Try for instance to read a piece of text in saturated
red against a background of saturated green (or the other way
around), where the two colors have the same brightness - it's
impossible unless the text is really big. Color TV exploits this by
transmitting brightness information at high bandwidth but color
information at considerably lower bandwidth - that's part of the
secret about how three images (red, green, blue) could be squeezed
into just one TV channel originally designed for black-and-white TV.
>> In general, make sure you have a quite visible BRIGHTNESS difference
>> as information carrier -- i.e. design your GUI so it's usable also on
>> black-and-white monitor (yes, ther are still some of those). Then
>> use colors for aesthetic effects, or to enhance the information already
>> visible if the screen is viewed in black-and-white.
>
> Again, what's important is ability to discriminate. Contrast (as opposed
> to absolute brightness)
difference of, say, 10 lux, is a very large contrast if the two
brightnesses are 1 and 11 lux, but won't be noticeable if the two
brightnesses are 1000 and 1010 lux. In favourable circumstances,
the human eye is barely able to see a relative brightness difference
of about 10%.
> is one of the good tools which can aid in facilitation
> of discrimination between two objects, however, there are many people with
> contrast sensitivity perception deficits in the population as well,
> especially in the aging population. Spatial contrast sensitivity deficits
> are way more common in the general population and are much more problematic
> than colour perception deficits.
want it to be readable by an as large part of the population as possible?
The characters are assumed to be fairly, but not extremely, large.
Suzanne
>
> There are several issues to consider with respect to colour blindness.
> First of all the typical stats quoted for "colour blindness" are for
> congenital (i.e., from birth) defects. These figures do not account for
> acquired colour perception defects.
True. Age-related visual deterioration affects the blue-green area of the
spectrum more than other wavelengths.
> > > "Don't use colors as the only information carrier since up to ten per
> > > cent of the male population has any kind of color reading defects".
> > > Quite often heard so I am looking for more inormation on this:
>
> First of all, I don't think any designer worth anything would code
> information solely with colour...it just makes no sense. Furthermore, it
> is extremely to code solely on colour...it takes some calculation...you
> would have to calibrate all other factors, including contrast,
> brightness, size,etc.
I agree that *good* designers don't code solely with color; however, a lot
of *employed* designers do. That is, there is a significant number of
applications that do code information solely by color. Sometimes, the user
can reassign colors to ones he/she can differentiate, but not always.
>
> The statistic misused above is for congenital (i.e., from birth) colour
> perception defects...these only relate to pure colours...very few colours
> on web pages and in the natural environment are so pure as to make
> discrimination difficult or impossible.
It does NOT only relate to pure colors. In fact, for most people who are
color deficient, low saturation colors and colors which otherwise deviate
from bright primary hues are the most difficult to differentiate. For
example, while a person who has some degree of red/green deficiency may be
able to differentiate between pure, saturated shades of red and green, they
may be completely unable to differentiate between pastel shades and very
dark shades. Also, there might be severe difficulty with non-spectral
colors, such as brown and purple.
<snip>
> The other thing to consider is that this statistic doesn't include people
> with acquired colour perception deficiencies as occur in conditions like
> macular degeneration.
I don't think the statistic is really that important. It is pretty clear
that prevalence of some degree of color deficiency is high enough to
warrant the avoidance of color-only coding.
>
> > > Are there different levels of the defect?
> >
> > Yes - there are all levels from somewhat anomalous color vision to
> > total color blindness - the latter is quite rare though. The most
> > common form of color blindness is inability to distinguish reds and
> > greens from one another.
>
> Again, I would be careful with this quote. The poster is only referring
> to congenital colour perception defects here...there are a myriad of
> acquired conditions as well.
Out of curiosity, why is this such an important fact to you? I don't see
how that would change the recommendations.
> >
> > > * What color combinations (as read on the screen) are to be
> > > avoided?
> > Red and green of course... :-)
>
> I strongly disagree with this statement...this assumes the colours are
> equated along all other dimensions required for colour discrimination.
Red and green shouldn't, in general, be used for color coding, unless there
are redundant coding modalities. (size, for example) This is particularly
true for red on black and green on black, which may vanish into the background.
> Again, what's important is ability to discriminate. Contrast (as opposed
> to absolute brightness) is one of the good tools which can aid in
> facilitation of discrimination between two objects, however, there are
> many people with contrast sensitivity perception deficits in the
> population as well, especially in the aging population. Spatial contrast
> sensitivity deficits are way more common in the general population and
> are much more problematic than colour perception deficits.
What are spatial contrast sensitivity deficits?
With regards to color perception and color coding, there are really two
different issues: color differentiation and color identification.
If both the red and the green shade are visible on the screen at the same
time, then it may be possible, on the basis of different brightness levels,
to identify which is which. (Through rote memorization of brighter=green,
dimmer=red, for example.) This is what I mean by differentiation.
Sometimes, the colors are similar enough in brightness/saturation that it
is impossible to differentiate between the two colors. Also, the user may
forget whether the brighter one is blue or purple, to give an example.
If the user must actually identify the color of the target item on the
screen, without some sort of reference index present, brightness/saturation
differences may be useless. For example, if I have to decide whether an
item is green or red, to indicate status, and it will only be one or the
other, I may well be unable to perform the task.
Here's a pretty concise page on color vision:
http://www.firelily.com/opinions/color.html
Vischeck is a color vision simulator:
http://vischeck.com/showme.shtml
Suzanne
Color-deficient (anomalous trichromat) instructional media designer
Peter Hill
"Paul Schlyter" <pau...@saaf.se> wrote in message
news:a1im0l$oj2$1...@merope.saaf.se...
> In article <NU2%7.8047$p04.2...@news20.bellglobal.com>,
> Peter Hill <peter...@sympatico.ca> wrote:
>
> > There are several issues to consider with respect to colour blindness.
> > First of all the typical stats quoted for "colour blindness" are for
> > congenital (i.e., from birth) defects. These figures do not account for
> > acquired colour perception defects.
> >
> > 1) With respect to the following quotes:
> >
> >>> "Don't use colors as the only information carrier since up to ten per
> >>> cent of the male population has any kind of color reading defects".
> >>> Quite often heard so I am looking for more inormation on this:
> >
> > First of all, I don't think any designer worth anything would code
> > information solely with colour...it just makes no sense.
>
> That does actually happen. In electronics, values of resistors and
> capacitors are usually coded with colored bands on the component.
> Also, signals for trains and airplanes use a lot of color - and one
> requirement for an engine driver or an airplane pilot is to not be
> color-blind.
>
OK...I give...but I would doubt these are pure colours...also, there is
ample positional information available in this case for discrimination...
> >>> * What color combinations (as read on the screen) are to be
avoided?
> >>
> >> Red and green of course... :-)
> >
> > I strongly disagree with this statement...this assumes the colours are
> > equated along all other dimensions required for colour discrimination.
The
> > issue here is discrimination of things...hue being one
factor...brightness,
> > etc., etc., etc.
>
> As a matter of fact, all people become color-blind when looking at
> small details. Try for instance to read a piece of text in saturated
> red against a background of saturated green (or the other way
> around), where the two colors have the same brightness - it's
> impossible unless the text is really big. Color TV exploits this by
> transmitting brightness information at high bandwidth but color
> information at considerably lower bandwidth - that's part of the
> secret about how three images (red, green, blue) could be squeezed
> into just one TV channel originally designed for black-and-white TV.
I would say this isn't so much of a "colour blindness" as it is an acuity
issue. The fact that RGB pixels placed close enough together and small
enough to eliminate seeing the three colours is more acuity than colour
perception, per se.
The best colours to avoid issues with colour blindness are "black" and
"white" (smile)
Peter Hill
additional comments (smile):
"Suzanne" <sda...@uclalumni.net> wrote in message
news:3C3D35EC...@uclalumni.net...
> Peter Hill wrote:
> >
> > There are several issues to consider with respect to colour blindness.
> > First of all the typical stats quoted for "colour blindness" are for
> > congenital (i.e., from birth) defects. These figures do not account for
> > acquired colour perception defects.
>
> True. Age-related visual deterioration affects the blue-green area of the
> spectrum more than other wavelengths.
>
> > > > "Don't use colors as the only information carrier since up to ten
per
> > > > cent of the male population has any kind of color reading defects".
> > > > Quite often heard so I am looking for more inormation on this:
> >
> > First of all, I don't think any designer worth anything would code
> > information solely with colour...it just makes no sense. Furthermore,
it
> > is extremely to code solely on colour...it takes some calculation...you
> > would have to calibrate all other factors, including contrast,
> > brightness, size,etc.
>
> I agree that *good* designers don't code solely with color; however, a lot
> of *employed* designers do. That is, there is a significant number of
> applications that do code information solely by color. Sometimes, the user
> can reassign colors to ones he/she can differentiate, but not always.
>
First of all, I only employ good designers (ha ha)...second, I guess what I
was trying to say was that assuming a designer was to only code using
colour, it would be very unlikely that he/she would end up with colours
which are not discriminated along some dimension (e.g., saturation,
brightness, etc.).
> >
> > The statistic misused above is for congenital (i.e., from birth) colour
> > perception defects...these only relate to pure colours...very few
colours
> > on web pages and in the natural environment are so pure as to make
> > discrimination difficult or impossible.
>
> It does NOT only relate to pure colors. In fact, for most people who are
> color deficient, low saturation colors and colors which otherwise deviate
> from bright primary hues are the most difficult to differentiate. For
> example, while a person who has some degree of red/green deficiency may be
> able to differentiate between pure, saturated shades of red and green,
they
> may be completely unable to differentiate between pastel shades and very
> dark shades. Also, there might be severe difficulty with non-spectral
> colors, such as brown and purple.
> <snip>
I apologize...what I was trying to say was that the 10% figure only relates
to congenital defects and doesn't take into account a myriad of other colour
perception defects which exist in the population at large.
>
> > The other thing to consider is that this statistic doesn't include
people
> > with acquired colour perception deficiencies as occur in conditions like
> > macular degeneration.
>
> I don't think the statistic is really that important. It is pretty clear
> that prevalence of some degree of color deficiency is high enough to
> warrant the avoidance of color-only coding.
I strongly agree with this statement, although in the last decade or so of
design I don't believe I've ever come across a designer who has solely coded
on colour...I'm sure it does exist, however, I would take care of that
problem by just terminating those designers (smile).
>
> >
> > > > Are there different levels of the defect?
> > >
> > > Yes - there are all levels from somewhat anomalous color vision to
> > > total color blindness - the latter is quite rare though. The most
> > > common form of color blindness is inability to distinguish reds and
> > > greens from one another.
> >
> > Again, I would be careful with this quote. The poster is only referring
> > to congenital colour perception defects here...there are a myriad of
> > acquired conditions as well.
>
> Out of curiosity, why is this such an important fact to you? I don't see
> how that would change the recommendations.
It is important to me only because an outright recommendation of avoiding
reds and greens is outrageous. There are many variations of "red" and
"green" which are quite easily discriminated by people with severe red/green
defect. I've quite successfully launched mass-marketed products (1M plus
volume) in the consumer market which have performed at a superior level with
respect to usability which have encorporated these colours.
> > >
> > > > * What color combinations (as read on the screen) are to be
> > > > avoided?
> > > Red and green of course... :-)
> >
> > I strongly disagree with this statement...this assumes the colours are
> > equated along all other dimensions required for colour discrimination.
>
> Red and green shouldn't, in general, be used for color coding, unless
there
> are redundant coding modalities. (size, for example) This is particularly
> true for red on black and green on black, which may vanish into the
background.
I agree...the key is ability to discriminate...as long as you provide a
mechanism for discrimination given a colour perception deficit then there's
no problem (i.e., redundant coding). Given a very narrow range of colour
options I would have no problem using red and green or any other
"problematic" colour combination. I guess what I'm saying is that in most
instances there will be redundancy with very little or no effort. The idea
that people would equate position, brightness, saturation, etc., etc., just
doesn't seem like a very common sense thing to do. Again, if you have
instances of this in your design I would question the value of your
designer. There are plenty of human-saavy designers out there without
work...try the Ottawa (Canada)market, for example...several hundred of my
former colleagues are still without work.
>
> > Again, what's important is ability to discriminate. Contrast (as
opposed
> > to absolute brightness) is one of the good tools which can aid in
> > facilitation of discrimination between two objects, however, there are
> > many people with contrast sensitivity perception deficits in the
> > population as well, especially in the aging population. Spatial
contrast
> > sensitivity deficits are way more common in the general population and
> > are much more problematic than colour perception deficits.
>
> What are spatial contrast sensitivity deficits?
Spatial contrast sensitivity is a measure of acuity which takes into account
that people require varying degrees of contrast to see things of differing
sizes. Very large objects (low spatial frequency...e.g., less than one
cycle per degree of visual angle) and very small objects (e.g., high spatial
frequency...e.g., greater than 22 cycles per degree of visual angle) require
significantly more contrast for perception than do objects containing
spatial frequencies in the middle range (e.g., 10 cycles per degree of
visual angle). For example, very small text (containing high spatial
frequency information) requires more contrast to be readable than larger
text. Our visual system is tuned to be most sensitive to things in the
middle range with respect to visual information. An interesting example is
that even though a newborn baby has poor visual acuity (as measured by high
resolution power tests) they have very high acquity in the range required
for facial recognition and nipples on breasts. Another example, people with
macular degeneration have lost the ability to read "normal text", however,
magnify it (increase the spatial frequency) and they have no problem as they
have a high spatial frequency defect but their low and middle spatial
frequency perception is relatively intact.
>
> With regards to color perception and color coding, there are really two
> different issues: color differentiation and color identification.
>
> If both the red and the green shade are visible on the screen at the same
> time, then it may be possible, on the basis of different brightness
levels,
> to identify which is which. (Through rote memorization of brighter=green,
> dimmer=red, for example.) This is what I mean by differentiation.
> Sometimes, the colors are similar enough in brightness/saturation that it
> is impossible to differentiate between the two colors. Also, the user may
> forget whether the brighter one is blue or purple, to give an example.
YES...this is exactly what I was saying although you have said it in a much
more elegant manner.
>
> If the user must actually identify the color of the target item on the
> screen, without some sort of reference index present,
brightness/saturation
> differences may be useless. For example, if I have to decide whether an
> item is green or red, to indicate status, and it will only be one or the
> other, I may well be unable to perform the task.
Absolutely...agreed.
>
> Here's a pretty concise page on color vision:
> http://www.firelily.com/opinions/color.html
>
> Vischeck is a color vision simulator:
> http://vischeck.com/showme.shtml
>
> Suzanne
> Color-deficient (anomalous trichromat) instructional media designer
Thanks for rich point of view.
Peter (no problems differentiating colour) Hill
Suzanne
> > That does actually happen. In electronics, values of resistors and
> > capacitors are usually coded with colored bands on the component.
> > Also, signals for trains and airplanes use a lot of color - and one
> > requirement for an engine driver or an airplane pilot is to not be
> > color-blind.
> >
>
> OK...I give...but I would doubt these are pure colours...also, there is
> ample positional information available in this case for discrimination...
While colors that are actually a blend of colors (non-spectral colors) are
potentially more likely to be perceived, often the difference is so small
as to be negligible.
In the case of resistors, unless things have changed a lot since I last
took an electronics course, position gives very little information for
discrimination. With the exception of the tolerance band, the color bands
can be any of the colors, regardless of position.
Sometimes, position does make a difference, such as in vertically oriented
traffic signals. If the light that is lit is on top, it's red. If the one
on the bottom is lit, it's green. On the other hand, you don't want to be
in a car my father is driving if you're in one of the US states with
horizontally oriented traffic signals. Yes, he could learn whether the red
is on the left or right, but until he learns it and the reaction becomes
automatic, it can be rather an enlightening experience. <g>
>
> > If so, what colors should I choose for, say, text and background, if I
> > want it to be readable by an as large part of the population as possible?
> > The characters are assumed to be fairly, but not extremely, large.
>
> The best colours to avoid issues with colour blindness are "black" and
> "white" (smile)
True, but eye-candy sells software.
Suzanne
Suzanne
>
> Suzanne, thanks for your comments...very thoughtful...of course, I have
> some additional comments (smile):
My pleasure. You happened to touch on some of my pet topics. ;)
>
> First of all, I only employ good designers (ha ha)...second, I guess what
> I was trying to say was that assuming a designer was to only code using
> colour, it would be very unlikely that he/she would end up with colours
> which are not discriminated along some dimension (e.g., saturation,
> brightness, etc.).
While it may not be common in business software (not common, but also not
non-existent), it is very common in children's software. (Haven't I had
this discussion before? <g>) My son has several commercial edutainment
packages that use color without secondary coding, making it impossible for
him to use them without also running an application I found that tells the
color name of the pixel the cursor is over. (It doesn't, however, work with
all applications, so there are still some he can't use.)
I think that there are many otherwise good designers who are either unaware
of the prevalence of color deficiency or who feel it isn't their job to
cater to the minority with these special needs. The group I work in is
somewhat more aware of this subject than typical. Out of five designers and
three software technicians, there are two who are color deficient. (Not to
mention the fact that the majority are working on or have advanced degrees
in human factors.)
<snip>
> >
> > I don't think the statistic is really that important. It is pretty
> > clear that prevalence of some degree of color deficiency is high enough
> > to warrant the avoidance of color-only coding.
>
> I strongly agree with this statement, although in the last decade or so
> of design I don't believe I've ever come across a designer who has solely
> coded on colour...I'm sure it does exist, however, I would take care of
> that problem by just terminating those designers (smile).
It's fine to say that designers who solely code on color should be
terminated, but what about the fact that the software that they created has
to either be reworked or it will remain unusable to a significant
percentage of the population?
One place that color is used extensively for coding, and often in absence
of secondary coding is on the web. Not only are the color deficient blocked
from access to the information on these sites, so are individuals using
screen-reading software. This is really a different subject, so I won't
discuss further at this point.
> > Out of curiosity, why is this such an important fact to you? I don't
> > see how that would change the recommendations.
>
> It is important to me only because an outright recommendation of avoiding
> reds and greens is outrageous. There are many variations of "red" and
> "green" which are quite easily discriminated by people with severe red/
> green defect. I've quite successfully launched mass-marketed products
> (1M plusvolume) in the consumer market which have performed at a superior
> level with respect to usability which have encorporated these colours.
A couple of questions, which I hope aren't taken as argumentative.
1. Did you test the discriminability of the colors on a population of
severely red/green color deficient individuals?
2. Is the lack of complaints from users taken to be tacit approval or
evidence of absence of a problem? If so, this may be fallacious reasoning.
Individuals with color deficiency are so used to having this problem, and
being told that it is their problem to solve, there are very few who would
complain. (Especially when there aren't any better alternatives for the
application.)
Outright avoidance of red and green may not be necessary, but one must
think carefully about ways to avoid problems introduced by the use of these
colors for conveyance of information. I would definitely NOT recommend
using dark red or dark green on a black background. I can guarantee that it
will result in zero discriminability in a significant number of people.
Likewise, if the user has to differentiate between red and green, they
*must* be of significantly different saturation/brightness, and a key must
be present to allow the user to determine which is the red and which is the green.
> I agree...the key is ability to discriminate...as long as you provide a
> mechanism for discrimination given a colour perception deficit then
> there's no problem (i.e., redundant coding). Given a very narrow range
> of colour options I would have no problem using red and green or any
> other "problematic" colour combination. I guess what I'm saying is that
> in most instances there will be redundancy with very little or no effort.
Keep in mind that discrimination and identification may not be the same
thing for a person with color deficiency. Even if I can perceive something,
I may not have the same perception of meaning as you, because of the
difference in color perception. For example, if you are using a code to
indicate the criticality of status for something, and you have a code of
red = highly critical, yellow = moderate criticality, and green =
informative, but not critical, it may seem sufficient to have a very bright
red and a very muted green. After all, you have saturation/brightness
coding as well as color coding. The only problem is that either the red or
the green may be indistinguishable from the yellow, depending on the degree
and type of color deficiency. Likewise, if the red or the green is against
a black background, it may not be visible at all. I would recommend other
methods of redundancy, which don't involve color, such as size, shape,
position, etc.
> The idea
> that people would equate position, brightness, saturation, etc., etc.,
> just doesn't seem like a very common sense thing to do.
What do you mean by this?
> Again, if you
> have instances of this in your design I would question the value of your
> designer. There are plenty of human-saavy designers out there without
> work...try the Ottawa (Canada)market, for example...several hundred of my
> former colleagues are still without work.
With the dot-combustion, the same is true here in California. I am lucky to
be in a still-growing area of design--education.
Suzanne
Paul Schlyter
Peter Hill <peter...@sympatico.ca> wrote:
> "Paul Schlyter" <pau...@saaf.se> wrote in message
> news:a1im0l$oj2$1...@merope.saaf.se...
>> In article <NU2%7.8047$p04.2...@news20.bellglobal.com>,
>> Peter Hill <peter...@sympatico.ca> wrote:
>>
>>> There are several issues to consider with respect to colour blindness.
>>> First of all the typical stats quoted for "colour blindness" are for
>>> congenital (i.e., from birth) defects. These figures do not account for
>>> acquired colour perception defects.
>>>
>>> 1) With respect to the following quotes:
>>>
>>>>> "Don't use colors as the only information carrier since up to ten per
>>>>> cent of the male population has any kind of color reading defects".
>>>>> Quite often heard so I am looking for more inormation on this:
>>>
>>> First of all, I don't think any designer worth anything would code
>>> information solely with colour...it just makes no sense.
>>
>> That does actually happen. In electronics, values of resistors and
>> capacitors are usually coded with colored bands on the component.
>> Also, signals for trains and airplanes use a lot of color - and one
>> requirement for an engine driver or an airplane pilot is to not be
>> color-blind.
>
> OK...I give...but I would doubt these are pure colours...
appear outside physics labs.
> also, there is ample positional information available in this case
> for discrimination...
position of the colored ring only tells the position of the digit,
while the color of the colored ring tells the value of the digit.
>> As a matter of fact, all people become color-blind when looking at
>> small details. Try for instance to read a piece of text in saturated
>> red against a background of saturated green (or the other way
>> around), where the two colors have the same brightness - it's
>> impossible unless the text is really big. Color TV exploits this by
>> transmitting brightness information at high bandwidth but color
>> information at considerably lower bandwidth - that's part of the
>> secret about how three images (red, green, blue) could be squeezed
>> into just one TV channel originally designed for black-and-white TV.
>
> I would say this isn't so much of a "colour blindness" as it is an acuity
> issue. The fact that RGB pixels placed close enough together and small
> enough to eliminate seeing the three colours is more acuity than colour
> perception, per se.
matter whether you call it an "acuity issue" or "color blindness of
small details", the fact remains that the human eye can see
brightness differences in considerably smaller detail than
hue/saturation differences.
>>> is one of the good tools which can aid in facilitation
>>> of discrimination between two objects, however, there are many people
>>> with contrast sensitivity perception deficits in the population as well,
>>> especially in the aging population. Spatial contrast sensitivity
>>> deficits are way more common in the general population and are much more
>>> problematic than colour perception deficits.
>>
>> If so, what colors should I choose for, say, text and background, if I
>> want it to be readable by an as large part of the population as possible?
>> The characters are assumed to be fairly, but not extremely, large.
>
> The best colours to avoid issues with colour blindness are "black" and
> "white" (smile)
to be stored with only 1 bit per pixel.....
Paul Schlyter
Suzanne <sda...@uclalumni.net> wrote:
>In the case of resistors, unless things have changed a lot since I last
>took an electronics course,
It hasn't - resistors are still color-coded the same way. The reason
for color-coding he resistors is this: many decades ago, resistors
weren't color coded, instead the resistor value was printed in (small)
text on the resistor. This caused trouble when repairing circuits:
to read the value off a resistor you could have to turn it to make
the text visible - and that could be hard or impossible with the resistor
firmly soldered in place. The purpose of these color-coded bands was
to be able to read the value off the resistor without having to turn
it.
> Sometimes, position does make a difference, such as in vertically oriented
> traffic signals. If the light that is lit is on top, it's red. If the one
> on the bottom is lit, it's green.
That's the way with car traffic signals - probably because also color
blind people should be able to drive a car. But in railway signals,
the same position can sometimes have one color and other times
another (and to confuse matters further, green may mean "stop" !!!!).
And airplane pilots, when landing in the dark, may have to rely on
the colors of the lamps on the runway to decide where to land. That's
why railway engine drivers and airplane pilots are required to not be
color-blind.
Sharon Curtis
Suzanne <sda...@uclalumni.net> wrote:
>Peter Hill wrote:
<snip>
>>
>> First of all, I only employ good designers (ha ha)...second, I guess what
>> I was trying to say was that assuming a designer was to only code using
>> colour, it would be very unlikely that he/she would end up with colours
>> which are not discriminated along some dimension (e.g., saturation,
>> brightness, etc.).
>
>While it may not be common in business software (not common, but also not
>non-existent), it is very common in children's software. (Haven't I had
>this discussion before? <g>) My son has several commercial edutainment
>packages that use color without secondary coding, making it impossible for
>him to use them without also running an application I found that tells the
>color name of the pixel the cursor is over. (It doesn't, however, work with
>all applications, so there are still some he can't use.)
Yuck, I really can't think of many reasons (if any!) why a package would
have to do this. Even if colour is used (and it can be used in plenty),
there are so few times when distinction is *needed* and can't be
provided in any other way.
>I think that there are many otherwise good designers who are either unaware
>of the prevalence of color deficiency or who feel it isn't their job to
>cater to the minority with these special needs. The group I work in is
>somewhat more aware of this subject than typical. Out of five designers and
>three software technicians, there are two who are color deficient. (Not to
>mention the fact that the majority are working on or have advanced degrees
>in human factors.)
I teach an HCI course, and some of them do have this "*sigh* We don't
*really* have to cater to minorities, do we? Sounds like an awful lot of
work!". Besides telling them firmly that yes they do, and in ever so
many cases it isn't nearly as much work as they'd think, if they'd just
consider things properly beforehand, I also do the following: I point
out to them that in their class, there *are* people who are disabled,
in particular there *will* be people who are colour blind. I say I don't
know who, but I do know that they are in the class somewhere.
Somehow, having it brought home to them that this is *not* a far away
issue, that by not catering to people properly they are actually doing
the equivalent of saying "Sod off" to someone who is actually in their
class, well, it brings it home to them just how rude they are being by
not taking the right attitude. That works better than just telling them
what they should be doing.
>> > I don't think the statistic is really that important. It is pretty
>> > clear that prevalence of some degree of color deficiency is high enough
>> > to warrant the avoidance of color-only coding.
>>
>> I strongly agree with this statement, although in the last decade or so
>> of design I don't believe I've ever come across a designer who has solely
>> coded on colour...I'm sure it does exist,
Sure it does. I can think of oodles of examples. Like the bright red on
bright green text on the back of my Persil laundry tablets box.
>> It is important to me only because an outright recommendation of avoiding
>> reds and greens is outrageous. There are many variations of "red" and
>> "green" which are quite easily discriminated by people with severe red/
>> green defect.
Um, I agree with not having to avoid reds and greens, but I think your
latter statement is bordering on dangerous ground. I'd say fine, use
red and/or green, just don't make it *necessary* that someone needs to
distinguish between those colours. In other words, if I get to play with
your design and I get to change your red/green/brown/orange/yellows to
any colours I want to, can I successfully remove information? No?
If so, good.
>Outright avoidance of red and green may not be necessary, but one must
>think carefully about ways to avoid problems introduced by the use of these
>colors for conveyance of information. I would definitely NOT recommend
>using dark red or dark green on a black background. I can guarantee that it
>will result in zero discriminability in a significant number of people.
>Likewise, if the user has to differentiate between red and green, they
>*must* be of significantly different saturation/brightness, and a key must
>be present to allow the user to determine which is the red and which is the green.
Hmm...I'd say if you have a key with red and green on the key, well,
that would be an automatic no-no on my part - dare I say it would
throw up a red flag? :-)
>Keep in mind that discrimination and identification may not be the same
>thing for a person with color deficiency. Even if I can perceive something,
>I may not have the same perception of meaning as you, because of the
>difference in color perception. For example, if you are using a code to
>indicate the criticality of status for something, and you have a code of
>red = highly critical, yellow = moderate criticality, and green =
>informative, but not critical, it may seem sufficient to have a very bright
>red and a very muted green. After all, you have saturation/brightness
>coding as well as color coding. The only problem is that either the red or
>the green may be indistinguishable from the yellow, depending on the degree
>and type of color deficiency.
Absolutely! e.g. if someone is protanopic, and they just have green
cones to do the work in the red/green area, if that yellow just happens
to hit the green cones at the same intensity as the red ones, well
there you go. Two indistinuishable colours. Even though in terms of
light/dark, the yellow might be bright than the red.
Paul Schlyter
which helps red-green blind people to distinguish between red and
green. The invention is called "Seekey" and was presented at an
optical fair here in Stockholm. It's really quite simple: it consists
of one green and one red filter. If a red-green person looks through
one of the filters, and then the other, he'll soon be able to distinguish
red from green: through the red filter, red looks brighter and green darker,
and it's the opposite through the green filter.
Suzanne
>
> In a Swedish newspaper I read today about an invention (by a swede)
> which helps red-green blind people to distinguish between red and
> green. The invention is called "Seekey" and was presented at an
> optical fair here in Stockholm. It's really quite simple: it consists
> of one green and one red filter. If a red-green person looks through
> one of the filters, and then the other, he'll soon be able to distinguish
> red from green: through the red filter, red looks brighter and green
>
While the specific invention may be new, the concept behind it has been
around for a long time. About 15 or 20 years ago, my ophthalmologist
provided me with a contact lens that would serve very much the same
purpose. The lens was tinted, so that when a red object was viewed through
the eye with the lens, it looked enough different from the view in the
other eye that I was able to tell red from green, regardless of saturation.
It was, however, useless for such colors as brown or pastel shades. It was
also, a tremendous bother, and I decided very quickly not to continue using it.
Suzanne
Suzanne
> Yuck, I really can't think of many reasons (if any!) why a package would
> have to do this. Even if colour is used (and it can be used in plenty),
> there are so few times when distinction is *needed* and can't be
> provided in any other way.
One software package that comes to mind uses color as a sorting variable
for a logic game. The task is to determine the rules for whether an object
is or isn't placed into a category, and there are multiple variables on
which sorting occurs. For example, red triangles might go into one group,
while blue objects of any shape might be in a second, etc. Without color
vision, the task is impossible.
> I teach an HCI course, and some of them do have this "*sigh* We don't
> *really* have to cater to minorities, do we? Sounds like an awful lot of
> work!". Besides telling them firmly that yes they do, and in ever so
> many cases it isn't nearly as much work as they'd think, if they'd just
> consider things properly beforehand, I also do the following: I point
> out to them that in their class, there *are* people who are disabled,
> in particular there *will* be people who are colour blind. I say I don't
> know who, but I do know that they are in the class somewhere.
> Somehow, having it brought home to them that this is *not* a far away
> issue, that by not catering to people properly they are actually doing
> the equivalent of saying "Sod off" to someone who is actually in their
> class, well, it brings it home to them just how rude they are being by
> not taking the right attitude. That works better than just telling them
> what they should be doing.
I tend to try to find ways that following the correct practice is in the
self-interest of the individual I am trying to convince. It works well with
children, and, apparently, programmers and designers. <g>
>
> Sure it does. I can think of oodles of examples. Like the bright red on
> bright green text on the back of my Persil laundry tablets box.
Yuck! That would be enough for me not to purchase that brand.
>
> Hmm...I'd say if you have a key with red and green on the key, well,
> that would be an automatic no-no on my part - dare I say it would
> throw up a red flag? :-)
In general, I'd agree. If, however, a designer insisted that he absolutely
*had* to use both colors, I would sure hope that he selected hues of very
different saturation levels and provided a key to help those who have color
deficiency. It's not perfect, but it is better than nothing.
Suzanne
Paul Schlyter
must be an eyestrain, so I fully understand why you stopped using it.
The idea behind "Seekeye" is that you use it temporarily, i.e. only
briefly in situations where you need help in distinguishing red from
green. These two (one red, one green) filters can be folded into a
small container about the same size as a set of keys, and you pick it
up and use the filters only when you need them.
The idea of using colored filters must be quite old, as you point
out, so the invention behing "Seekeye" must be the "packaging" of the
filters themselves.
--------------------------------------------------------------------
BTW not only color-blind people can benefit from using colored
filters - people with normal vision can benefit too! A number of
amateur astronomers watch and make drawings of the planet Mars
whenever it's closest to the Earth (which happens once every 26
months or so). Now, the surface of Mars contains a lot of low
contrast features, many of them subtly colored in different ways.
These Mars observers have found out that by using a suitably colored
filter you can often see details on Mars you otherwise won't see.
This requires experience though - a person who views Mars through a
telescope for the first time usually don't perceive any surface
details at all on the (quite tiny) planetary disk.
---------------------------------------------------------------------
While most forms of deviating color vision is a deficience of color
vision ("color blindness"), a few people seem to instead have an
enhanced color vision. They are tetrachromats, and from their point
of view, all of us trichromats are color blind: they see all colors
we see, plus some we don't see. Or, to put it more properly, colors
which trichromats perceive as the same the tetrachromats can perceive
as different.
Now, if we trichromats equipped ourselves with a set of suitable
color filters, we could perhaps get some idea of the colors the
tetrachromats see? Just like red-green blind dichromats can use
the red and green filters of "Seekeye" to distinguish between
red and green....
Bradley K. Sherman
>
>Yuck! That would be enough for me not to purchase that brand.
However, the marketing team would consider it a major
victory for the graphic designers to have achieved
two mentions of the product brand name in this newsgroup
based purely on color choice!
--bks
Chip Wood
problem. Example, the PHBs decided on a dark blue background for all
PowerPoint presentations. However they also chose a shade of dark Red for
IMPORTANT and CRUCIAL text that is impossible for me to read in a moderately
lighted room. After bringing this problem up numerous times to various
presenters, nothing was done, and now I just suffer and try to learn the
IMPORTANT and CRUCIAL info off-line. Another difficult thing to see is the
laser pointer spot- too dim, too small, and too RED.
"Sharon Curtis" <s...@peseta.cs.stir.ac.uk> wrote in message
news:a1n89k$sn6$1...@peseta.cs.stir.ac.uk...