Color management myths and misinformation video

[Slobodan Blagojevic]

... If you have a thermometer that goes from 0-100º C and another that goes from 0-200º C, which one has more information?... If I take to two measurements with each, do I have more information if I use the 0-200º thermometer? No, you have exactly two measurements worth of information in each case...

You analogy is based on a presumption that the reading will be within the healthy human range, i.e., up to 98.6. In which case it does not really matter which thermometer you use. Just as it does not matter if you use sRGB or Adobe RGB if the subject's gamut falls within sRGB.

However, the analogy breaks down the moment you are sick and the temperature goes above 100. You will get more information with the 200 degrees thermometer. While the 100 degree thermometer will tell you are sick, it won't tell you how seriously, you will only know it is at least 100 and possibly above. If the other measures 104 degrees, you'll have more and better information and will call 911. Similarly, with sRGB, you will know there are colors at the edge of the triangle in the, say, green-blue zone, but you will not know how many or how much are outside of it. With Adobe RGB (or wider) you will. Just do not call 911.

[Slobodan Blagojevic]

I would strongly suggest we leave Gary out of this thread. If Andrew wants to create a video, it has to have a broader goal than just continue the bickering with one person.

[Royce Howland]

Andrew, I pretty much agree with a lot of these posts, including Tony Jay's just above. It may feel emotionally satisfying to do myth-busting and go head-to-head with disinformationists... lord knows I like to do it now and again. But it's more important I think to stay super constructive, pragmatic and actionable. Give early-stage photographers some clear, well-founded, accessible information they can use to make good decisions today. With pointers to where they may choose to go tomorrow.

This is how I try to structure my printing seminars and workshops, which target really entry-level folks at the widest end of the spectrum. I used to try to myth-bust some stuff at that entry level, but in my judgment it wasn't helping so I've stopped doing that. I now focus on providing a streamlined & simplified, but well-founded, workflow designed to produce good results in a variety of situations for people who want to print. And progress into more advanced areas for those who want to do so.

A key thing to do first is to frame the audience within the material, so they know whether it applies to them or not. This is what most armchair pundits don't do right at the beginning. Instead they hold forth with context-less, sweeping generalizations, silently making all sorts of assumptions and begging all kinds of questions. This doesn't help the reader / viewer really understand how or whether the info applies to them.

"This video is for you if you have a digital camera and you take ready-to-view JPEG photographs with it. I'll give you the essential information you need about good digital colour right now. You may have a couple of different choices to make if you only post small photos to Facebook, or if you get small prints made at a quick photo lab like Walmart, or if you want to make your larger prints on a photo inkjet printer with the best bang for your colour buck. Along the way, I'll point out a couple of things to help you avoid getting tripped up by common issues of bad digital colour, such as when you're looking at your photos online in a web browser. I'll also point out one or two of the most important consequences if you make certain choices now and want to change your mind later.

"If you're a photographer shooting RAW, or you have more advanced software like Adobe Lightroom or Photoshop, or you have a newer monitor that's advertised as having 'wide gamut' and which makes your photo colours look wonky, you may want to know more about controlling digital colour. If you're ready for more details, watch this short video to make sure you have the basics covered. Then go on to video XYZ where I'll talk more about colour spaces and how to make sure colour is consistent across your cameras, monitors and printers."

Or something like that...

The best rebuttal to bad info is, I think, not to get dragged into a tit-for-tat exchange with the disinformationists, tactically fighting problems on a case-by-case basis. I think you'd be better spending your time putting out quality information that engages the viewership it's meant to help, and then circulating that material around. Like immunizing a population against a common virus.

[digitaldog]

I would strongly suggest we leave Gary out of this thread. If Andrew wants to create a video, it has to have a broader goal than just continue the bickering with one person.

I'll start with this last thread first. I agree. There are numerous and equally (?) incorrect video's about sRGB vs. Adobe RGB (1998) on YouTube that it isn't necessary to single out one person.

[digitaldog]

You analogy is based on a presumption that the reading will be within the healthy human range, i.e., up to 98.6.

Well let's agree up to 105 or so, whatever agreed upon temp under which you're not dead. If the reason for taking the measurement is to see if you have a fever and if so, how high. As such, if we agree the max value is 105, the 200 degree meter isn't useful.
Isn't that the same as defining a dE of less than 1 not being a visisble difference between two colors and if so, not to be used?

[digitaldog]

The best rebuttal to bad info is, I think, not to get dragged into a tit-for-tat exchange with the disinformationists, tactically fighting problems on a case-by-case basis. I think you'd be better spending your time putting out quality information that engages the viewership it's meant to help, and then circulating that material around. Like immunizing a population against a common virus.

I agree.

[Simon Garrett]

Another temperature analogy: instead of comparing 0-100 or 0-200, how about 0-255F or 0-255C. 
As with Adobe RGB or sRGB jpegs, the numbering range is the same in these two cases, but they cover a different range of temperatures (colours). 

The Farenheit scale covers a little bit more (32F) at the bottom end, but the Celcius scale is larger overall, and covers 232F higher at the top end.  The same range of numbers covers a larger range of temperature in Celcius. 

Colour has 3 axes; sRGB and Adobe RGB are similar on the Red and Blue axes, but with the same number range Adobe RGB goes much further in the green axis (in crude terms). 

[David Sutton]

Andrew, in my humble opinion the issue has never been about an argument pitting sRGB and AdobeRGB against each other.

Instead Gary was committing acts of violence against a rational colour management-based workflow.
Far better to explain the place of the various colourspaces in a rational workflow.
How, depending on one's goals at shooting and the intended output, one can make very rational choices about assigning workspaces in camera (not applicable when shooting RAW - this should be explained too).
Some time should be spent discussing the rationale behind Adobe giving photographers no choice in their working colourspace - it is ProPhotoRGB for everyone - so many of the strange and incorrect misconceptions that Gary has can be put to bed here.
A good explanation of colourspaces on input, working colourspaces, and output colourspaces integrated into the above discussion.
How it is possible to take an image from one colourspace to another without doing the sort of violence to it that Gary demonstrated i.e. what happens with a colourspace conversion and the place of softproofing.

Thinking about it - I am not sure that any of us could adequately explain any of this in few minutes but the challenge is to make rational colour management-based workflows accessible to the very owners of "Digital Rebels" that Gary Fong believes cannot understand principles of colour management.
You may note that I have not included Gary's classification of individuals who are not interested in colour management - they will no more view Gary Fong's videos on colour management than they will Andrew Rodney's.
I was once one of those Digital Rebel owners who knew nothing about colour management but gradually became dimly aware that it may actually have some relevance - the journey started there.

So, I strongly believe that we should not get sucked into Gary Fong's misinformation agenda (the sRGB vs AdobeRGB non-issue) and concentrate on explaining rational and practical colour management-based workflow decision-making.

My 0.02c

Tony Jay

+1 (and Royce's post)

[Alan Goldhammer]

What is missed in the thermometer analogy is scale and precision.  My educational background is in chemistry and certainly we had to make lots of measurements in lab courses and for my graduate thesis work.  The scale of measurement is important as was already touched on in the case of a human running a fever.  A 0-100F thermometer is likely of little use if it reads 100.  100 is a modest fever but 104-06 is emergency room time.  Precision is critical as it gives you a better data point.  I can go to the grocery store and buy a meat thermometer 0-200F or a candy thermometer 0-400F.  these clearly encompass the scale needed to predict whether a fever is life threatening but useless in the real world as they don't have the precision.  I'm old enough to remember the days of the mercury fever thermometers and IIRC, their range was 90-110F and could read to 0.1F.  Of course everything now is digital.

Just trying to be cautionary when thinking of analogies.

[Slobodan Blagojevic]

If you look at the triangles, how can you NOT say that the bigger triangle (Adobe RGB) has more colors?

[Eyeball]

Here are a few quick observations responding specifically to what you asked for (regarding a "common myths" type of video):

• Be careful with consistency regarding how you express the "myths".  For example, "AdobeRGB produces duller colors" is (largely) a myth but "There are no sRGB printers" is the correction to the myth that "There ARE sRGB printers".
• For the duller colors, I would stress that this is the result of a bad conversion from one color space to another.  That bad conversion could be caused by the photographer (due to incorrect color management practices), by the printing company (failure to recognize non-sRGB or bad color management in general), or by the browser/viewer (in the case of web presentation).
• For the "sRGB printers" myth, I think you run the risk of getting into confusing territory if you mix a gamut discussion with RGB vs. CMYK.  It might be safer to not bring up CMYK at all and just emphasize that output devices have their own gamuts.
• For "printers with a color space larger than sRGB", besides showing gamut plot differences, it would be great if you could somehow show what those differences translate to in an actual print.  I know it is like judging a high-end stereo over the telephone but I think giving your audience a hint of what to look for (increased saturation, more detail/less hue shift/less blockiness/posterization in saturated areas) would be helpful.
• The "more colors" one is tricky.  I think you can certainly say without any equivocation that the best way to say it is that AdobeRGB permits a wider range of color to be expressed by your image, if your image actually has that range of color.  You might add that the reason we try to avoid talking about "number of colors" is that it just isn't very useful.  Here is an example:  I have one box of 20 crayons with a variety of colors; I have another box of 20 crayons with 20 different shades of yellow; Which box has the greater number of colors?

Now, beyond that feedback oriented to what you asked for, I'll add the following:

I agree with much of what others have said.  A "rebuttal" is going to add more fuel to the fire and calling those things "myths" is almost giving them more credit than they're due.  There is also the false dichotomy of "AdobeRGB vs. sRGB" that others have pointed out - something that I think originated from a doubt of whether to use the in-camera sRGB/AdobeRGB setting and that has largely been lost in all of this.

Perhaps a more positive alternative for a video would be to provide a down-to-earth value road-map for how a beginner might see their journey in terms of learning and using color management.

For example, starting out the cost is high (need to study, need to take extra care, print labs or viewers who may not interpret non-sRGB correctly) and the benefit is low (better representation of bold color on SOME media/devices).  Over time, the costs start to level out (better understanding, more comfort) and the benefits may go up (satisfying more demanding clients, satisfying a more demanding YOU).

It might also be worthwhile mentioning some of the key things that they will probably want to consider if they continue on their "color management journey":
- Am I shooting subjects with bold colors that stretch beyond sRGB?
- Will I be using an output device/media that can show those bold colors?
- What is the likelihood that something will go wrong between photographer and client/viewer if I use a non-sRGB color space?
- How important is it to me and/or my clients and viewers that I produce those strong colors as accurately as possible?
- Even if I don't find it necessary to output wide-gamut today, do I want to keep that option open for the future?

[Eyeball]

If you look at the triangles, how can you NOT say that the bigger triangle (Adobe RGB) has more colors?

Remember your geometry?  A big triangle and a small triangle both have an infinite number of points.  One does not have more than the other.

The statement also fails from a practical standpoint since the nature of digital imagery caps the points/colors for images in both color spaces to the same maximum.  You have a larger range/gamut for one but the same number of potential colors as dictated by the number of bits used.

[mrenters]

For "printers with a color space larger than sRGB", besides showing gamut plot differences, it would be great if you could somehow show what those differences translate to in an actual print.  I know it is like judging a high-end stereo over the telephone but I think giving your audience a hint of what to look for (increased saturation, more detail/less hue shift/less blockiness/posterization in saturated areas) would be helpful.

What I've done in the past is to print the same image in sRGB and the printer's full gamut and then place the two prints on top of each other with holes cut out in those areas where the gamut exceeded sRGB to show the differences.

Martin

[digitaldog]

Lots going no, where to start.

In terms of the color numbers and color space, I think Mark pointed out something that is important:

AdobeRGB and sRGB are just spaces, they don't inherently have any information (other than specifications for primaries, white point, etc). Until you actually have a pixel, there isn't any color information.

I believe that is quite true.
If I make a 300x300 pixel document with a white bkgnd in Photoshop using sRGB then Adobe RGB (1998) does one have more colors?
Next he wrote:

Those measurement could be specified by how accurate they and this could be expressed as bits.

We should consider bit depth and encoding. If the two PS documents above are 8-bit per color or 16-bit per color, any differences?
We also need to consider what we mean by a color. Let's skip the color space for a second and say we have a pixel with a value of R55/G55/B55 and one with a value of R55/G56/B55 and we can't see any difference. Is that a different colored pixel? We can in theory define 16.7 million colors with 24 bit precision. We can't see 16.7 million colors. This was discussed in the Steve Upton article about the PowerBook display.
I would submit that if we are talking numbers, if we can't see a difference between one versus the other, they can't be different colors.

If you look at the triangles, how can you NOT say that the bigger triangle (Adobe RGB) has more colors?

And here's the rub. The gamut is larger, no question. The range of colors is greater in one than the other. Does that mean one has more colors? Isn't that an attribute of the encoding of the pixles? Which goes back to gamut volume. Adobe RGB (1998) does have a larger gamut volume than sRGB but more colors?
Back to encoding. Now we have R55.009/G55.009/B55.009 and R56.001/G56.001/B56.001. Can we see a difference? If not, are these two legal color values?

[Slobodan Blagojevic]

Remember your geometry?  A big triangle and a small triangle both have an infinite number of points.  One does not have more than the other.

The statement also fails from a practical standpoint since the nature of digital imagery caps the points/colors for images in both color spaces to the same maximum.  You have a larger range/gamut for one but the same number of potential colors as dictated by the number of bits used.

HuH!? I might have a rusty geometry memory, but never heard of "infinite number of points." Sure one has a larger area that the other?

Also, that fact they might have the same number of potential colors, does not imply they have the same number of actual colors. Again, I am speaking from a layman's perspective.

[digitaldog]

Also, that fact they might have the same number of potential colors, does not imply they have the same number of actual colors.

And that's a critical distinction! 

[fdisilvestro]

The issue I don't like about a debate between sRGB and Adobe RGB is that it might give the false impression of that's all it is about color management.

I would suggest to start first with the foundations of color vision and color management and then get into virtues and limitations of current practical approaches, being sRGB and AdobeRGB some of the options

Colors are not physical properties, they are perceptions in our brain. I myselft could not make sense of color management without first having a basic understanding on color vision.
Just the opponent process is fascinating!

Something I just learned a few days ago: human photoreceptors are sensitive to UV light. It is the lens in the eye which filters it, so we normally do not perceive UV. It has been reported of people undergoing cataract surgery being able to see into the UV region. (BTW No, I will not consider undergoing an unnecessary intervention to see UV light) 

[Eyeball]

HuH!? I might have a rusty geometry memory, but never heard of "infinite number of points." Sure one has a larger area that the other?

OK.  You have a triangle that is 3 inches on each side.  How many points DOES it contain?
The answer depends on the area or volume that a point occupies.  Since in geometry, a point is generally considered to be dimensionless (no height, width, or depth), it occupies no space and therefore an infinite number can fit inside a given area or volume.

This is the theoretical reason that it doesn't make much sense to talk about quantity of discrete colors.  Colors are essentially points.

Also, that fact they might have the same number of potential colors, does not imply they have the same number of actual colors. Again, I am speaking from a layman's perspective.

Exactly.  Another reason why it's best NOT to say that AdobeRGB "has more colors".

At a practical level, the number of colors is constrained by the bit depth and by the color originally captured in the camera.  The number of colors is NOT constrained by the gamut of the color space UNLESS you start defining an arbitrary volume to what a "color" is.  This is basically what Chromix is doing when it uses a Delta-e value to define the area of a "color".  Just be aware though that this is placing a human perceptual limitation on what a color is.  A camera sensor can differentiate more finely than that and probably a lot of output devices as well.

If you really wanted to push it at a practical level, I guess you could say that a larger-gamut color space "has the potential to contain more colors" if:
- if the maximum precision of the color recording device can be determined
- the bit-depth being used to digitally capture the image exceeds that maximum precision

But you can see just how far-out we have to go to even get close to making the statement "has more colors" legitimate.  We are almost literally into "number of angels on the head of a pin" territory.

I just think that the clearest and least-misleading way to talk about it is by sticking with words like gamut, range, and volume and NOT talking about quantities of discrete colors.

[Eyeball]

Colors are not physical properties, they are perceptions in our brain.

Well, I would say "colors are not JUST physical properties ..." but I completely understand your point.

This is one of the reasons that I think color theory and color management can get so complex.

- First, you have the physics, which is complex enough.
- Then enters human biology and how it reacts to the physics.
- Then you have the hardware and software creators, who can implement things in a certain way due to other constraints or down-right mistakes that have nothing to do with either the physics or the biology.

I think that is why the subject can sometimes take on a life of its own and leave photography in the dust.  A lot of the science and engineering can make an interesting hobby by itself.

[Slobodan Blagojevic]

... We are almost literally into "number of angels on the head of a pin" territory.

I just think that the clearest and least-misleading way to talk about it is by sticking with words like gamut, range, and volume and NOT talking about quantities of discrete colors.

Oh, your triangle explanation is definitely in the "number of angels on the head of a pin" territory.

If "sticking with words like gamut, range, and volume" you are DEFINITELY weaseling out of using plain English. You are hiding behind another layer that needs explanation (gamut, range...). All this just to avoid saying simply that Adobe RGB does have more discernible, actual colors, while at at the same time having the same number of potential colors as sRGB.