Zygomatic Color

Common inkjet printer drivers accept only RGB images, and it is a good design (theoretically)


Categories: Blogs Tags: Printer Drivers

Why we don’t have CMYK capable printers?

I’ve always wondered why common inkjet printer drivers don’t accept CMYK images. Inks are CMYK. Common inkjet printers can accept CMYK images, definitely. I couldn’t understand the design of inkjet printer drivers which accept only RGB images.

Before we proceed, let’s confirm the basics about printing.

So the basics about programming.

All non-trivial abstractions, to some degree, are leaky

Device RGB can be a perfect abstraction of color. Color is three-dimensional, and device RGB is three-dimensional too. They can be bijective for the printing gamut. (theoretically)

On the other hand, CMYK is four-dimensional. This is a bad abstraction of color. According to the abstraction, a perfect composition of CMY is the same color with K 5%. They are not bijective.


In reality, there is no such a perfect composition. CMY mixture cannot be genuine gray what we expect. So sometimes we need K 5% with noticeable dots instead of CMY mixture, right? Actually, expensive photo printers’ drivers offer black-and-white mode which ensures no CMY. Those printers don’t print any noticeable dots in the mode because they have gray ink, but we should have an option. Moreover, we should be able to print CMY mixture and K 5% both in a sheet, because our printers can do it! CMYK isn’t a bad abstraction of color, because, in the first place, it isn’t an abstraction of color. Device RGB is a good abstraction of color, but a bad abstraction for inkjet printing, isn’t it?

It makes sense in a point. But just a point.

If we need the option seriously, we should buy RIP, instead of adding more trouble to printer drivers. Common inkjet printer drivers are already deadly troublesome for most customers, and almost no one needs RIP or something like it. Only who needs RIP should pay the cost. This is ‘market segmentation.’

(BTW actually in such cases I recommend you to buy 700 USD photo printers instead of 1,000 USD RIP.)

“Common inkjet printers print color.” This proposition is already an abstraction. The abstraction leaks the detail about genuine gray and counterfeit gray. All non-trivial abstractions, to some degree, are leaky. Unless the leakage is serious for the usage, we can leave it. As far as we accept counterfeit gray, common inkjet printers print color. Device RGB is a good abstraction of color.

Of course, I will love stable, fine printing quality, open source RIP, if such a thing exists (Ghostscript is it for some printers). It never bothers the most customers of printers.


Did you find the word ‘theoretically’ in the title of this article?

In some printer drivers (EPSON’s macOS drivers for some all-in-one models), device RGB’s yellow (#FFFF00) contains magenta. I feel it is awful, but no one can assert it is a bug because there is no specification about ‘No Color Adjustment’ behavior. It may be a sales promotion of expensive printers… If the printer driver accepts CMYK, and M 0% contains magenta, everyone asserts the behavior is a bug. This is the reason why sometimes I want CMYK feature in reality. Set a thief to catch a thief. Life doesn’t go by theory alone.