A Curious Thing

[J B Lansing]

I have been having trouble with a printer who cannot seem to match a partular blue I used in the background of a panel. I even gave him the CMYK values: C=72.5, M=72.2, Y=31, k=9 with out clearing up the problem. So I was playing around with the color. I created a patch of that color and then made a duplicate and gave it a pantone color that was close to my original color. The I bought up the color dialog box and set the CMYK values to the same as the original. The two colors did not match! Here are the examples.
I would very much like to hear a someone explain this.
Thanks
J B Lansing

[J B Lansing]

I have been having trouble with a printer who cannot seem to match a partular blue I used in the background of a panel. I even gave him the CMYK values: C=72.5, M=72.2, Y=31, k=9 with out clearing up the problem. So I was playing around with the color. I created a patch of that color and then made a duplicate and gave it a pantone color that was close to my original color. The I bought up the color dialog box and set the CMYK values to the same as the original. The two colors did not match! Here are the examples.
I would very much like to hear a someone explain this.
Thanks
J B Lansing

[jens g.r. benthien]

J B,

color printers are not suited for proofs. A professional printing process is different to what the ink jet printers offer. In addition you would need to calibrate your scanner (film or negative/slide), monitor and printer. The densitometric tools are very expensive - too expensive for 'soho' printing.

And even if you would have calibrated all devices, the colors might differ: the ink cartridges will be from different lots in the factory, and they don't calibrate the inks - there is a relatively large variation in color pigments.

Finally: to proof your colors you need to find a printing company with proof equipment. A letter or A4 size proof costs between US$ 80 and 200, depending on the equipment they use.

Part of the problem is based on the different color models:

your monitor display shows colors that are 'active light', that means the CRT or TFT pixels emit light in the specified color.

A printed color is a passive one - it only reflects the environmental color. Examined in bright sunlight the printed colors will definitely look different than in artificial light in an office or home (bulbs and flourescent lamps emit different wave lenghts - bulbs more read, the 'neons' more blue).

Complicated? You bet!

If you want to achieve a certain color on a specific paper, there is no way around a test pattern. Then you pick the color you want and use that one, no matter how it will look on your screen... a purple on your screen might be the perfect blue on the print!

jens

[gwpriester]

Where did you get the CMYK values from?

If you used a CYMK swatching system such as TruMatch or Pantone CMYK colors, and your printer (I assume you mean a professional printing company and not a desktop printer) could not match the colors, then the problem is with your printing company. They should be able to reproduce a color if you have specified the color correctly from a reliable swatching product.

Pantone spot colors, are not intended to be printed in CMYK. They are intended to be printed as solid colored, specially mixed inks. Pantone spot color inks are mixed to a special formula and should print the same regardless of who is printing them.

About 1/2 of all Pantone spot colors can be matched using CMYK. About 1/4 of the colors will be close but not match exactly. The other 1/4 are not even close. Deep blues are not easy to reproduce in CMYK. Reflex blue is a good example. There is no way to match this intense deep blue with CMYK.

Finally, as Jens has pointed out, colors you see on the screen are RGB colors and have a much greater gammut than CMYK colors. So many beautiful colors you can create on the screen are just not in the gammut of CMYK printing.

This is why Xara, and most graphics software, offers a Simulated CMYK color space (Window > Show Printer Colors). In this color space, the usually vibrant screen colors appear dull and muted. But in reality, these colors are closer to what the printed output will be.

Hope this helps answer your question.

Gary

[J B Lansing]

This may sound grumpy and I don't mean it to be. But, while you both gave me very good information, it does not answer the question. Why do the two samples shown look different when they have the same CMYK values?
The one on the left looks significantly less saturated than the one on the right. If I look at their RGB or HSV values they ARE significantly different. It seems to indicate that CMYK does not provide a complete definition of a color. Incidently I did finally supply the printer with a PMS color number that he matched.

[Soquili]

Hi J B Lansing,

Gary's information about the simulated printer colors is the key to the answer. The attached image shows the same variance in 'brightness' as your first post.

[gwpriester]

You don't sound grumpy at all. Getting accurate color is a very frustrating thing. And it has been from the day we started trying to get color from the computer to print like what we see on the screen.

CMYK color is not intended to be viewed on the screen. CMYK color is one color model and RGB is a very different color model. If you look at an image printed in CMYK on a sheet of white paper, and then look at the same image on a color corrected light box on a 4 x 5 or 8 x 10 transparency, the transparency will look much brighter and the colors will appear more vibrant. There is no way you can capture the same vibrancy in CMYK printing.

Red is a very good example.

Several years ago, I worked on a printing project with a professional photographer who had taken a spectacular studio photograph of a red sports car. The 8 x 10 transparency was rich and saturated on the light box. And yet the printed version of the image was dull and the colors were muted.

We finally used a "touch plate", a 5th color, Rubine Red which we ran along side the magenta which made the printed red almost as rich and bright as the transparency.

Hexachrome printing, which adds an orange and green to the CMYK is capable of producing more vibrant and saturated colors. CMYK is just limited by the bandwidth of colors it is able to accurately reproduce.

Gary

[Ross Macintosh]

From reading this thread, it seems JB's question still isn't directly answered. Is it possible for what appear to be two different colours to have identical CMYK values??

Regards, Ross

PS - thanks Jens & Gary. Your explanations on this complicated subject are very educational.

[tad]

What Color Really Is....

Don’t forget folks, color requires:
1) A light source
2) A reflecting or transmitting object
3) A beholder - eye, phototube, diffraction instrument, etc.

CMYK only affects #2; and, theoretically if two colors have the same CMYK values they will look the same to one person under the same light source.

Unfortunately, I don’t think any printer on earth can print the same color twice. AND, If I’m dissing your pantone swatches or color calibration equipment, hey, go visit a large print shop. They put trees in one side of the building, and with pressmen running up and down the line slapping gallons of ink on the press, colored brochures come out the other end - neatly folded.

Rgds,
tad

[J B Lansing]

Soquili,
The color shift in my example is not due to a change in the display. What you see is exactly what is displayed. I believe it is a coincidence that it looks like your example. The point is that it seems CMYK is not a complete definition of a color. While I recognize that different color media and different color standards (not the right term, I mean RGB vs CMYK) can result in color shifts, This example is displayed in a common standard.
I did not use trickory to arrive at these colors, they occured following the procedure I outlined. Try it and see if you get the same result.
J B