To RIP or not to RIP

From ColorWiki

Revision as of 00:02, 2 June 2018 by Patrick (Talk | contribs)
(diff) ←Older revision | view current revision (diff) | Newer revision→ (diff)
Jump to: navigation, search
Reserved Article

This page is a
Reserved Article.
For more details see
Reserved ColorWiki Articles

ColorNews

This reserved article originally appeared in CHROMiX ColorNews Issue 15 on October 19, 2004.

Click here to see the original in its original context.
Email
colornews(at)chromix.com to subscribe to the ColorNews newsletter.

by Steve Upton


Contents

To RIP or not to RIP?

With the proliferation of inexpensive inkjet printers it seems everyone can now afford one and most professional imaging people have one (or more). Most inkjets on the market - even the wide-format variety - are available without a RIP and are accessed through the manufacturer's print driver. We receive a constant stream of questions about RIPs: what they do, how they work, are they required. It seems high time to put together an overview of RIPs' basic functions and uses.

Strictly speaking, a RIP is a Raster Image Processor. Not since the days of pen plotters have lines and curves in software actually been drawn as lines and curves on paper. All major printing technologies create pictures, linework and text using a grid of dots sometimes called a Raster. To create this grid of dots, software is required to convert line work (curves, text, and so forth) and images to a printable matrix of pixels and then screen them using a complicated pattern of cyan, yellow, magenta and black ink dots. With the advent of Adobe's Postscript page description language our software had the ability to create entire pages of text, artwork and images using raster and vector commands that could then be printed on any printer that could interpret Postscript. Postscript is not an open language but Adobe licensed it to many printer manufacturers so we can use the same printing language to print to our laser or inkjet printer that we use to print a final job on press. The RIP software engine that interprets the Postscript often resides in printers themselves but can also run on a purpose-built computer directly connected to the printer or (as is becoming more common) on a desktop or server computer running Windows or Mac operating systems.

Like most things in the computing world, RIPs have evolved mostly by adding feature after feature. Many of these features have nothing to do with rasterizing Postscript but make good sense to have on the same computer. RIPs now share printers across a network to multiple users, allow printing queue management for prioritizing of print jobs, nest jobs to save paper, and so forth.

A number of features that DO control or affect color are: linearization, ink limiting, screening & dot simulation, direct channel control and color management. Let's handle them one at a time.

Linearization

Straightening out the behavior of each channel makes a printer much more palatable to print profiling. RIPs have traditionally linearized based on density measurements but this is not necessarily linear as far as our eyes are concerned. Some RIPs are starting to allow linearization based on L* (the L in Lab) or Chroma (the C in LCH, a cousin of Lab). This type of linearization makes profiling much simpler and tends to increase accuracy as well. For printers that include light cyan and magenta inks, most RIPs allow control over light and dark ink blending to smooth the transition between inks.

Ink Limiting

Inkjet technology is all about the relationship between ink and paper. RIPs can allow limiting of the ink in each channel as well as the total ink limit allowed on paper (good RIPs allow both). There are two strategies to ink limiting each channel. One is to set the inking limit to the point just before the ink puddles on the paper. This allows for the maximum gamut possible from the printer. Unfortunately, when laid down heavily, inkjet inks can behave rather strangely by changing their hue. This causes "hooks" at the end of each channel (easily seen when graphed) and profiling software can struggle to capture this behavior properly, causing inaccuracies in saturated colors. Accordingly the second strategy is to limit the ink to a point just before the sharpest part of the hook occurs. This can be an effective trade-off between gamut and inaccuracy and is popular for proofing systems where the target gamut is defined by the proofing / reference profile. Setting the ink limit to be just high enough to contain the reference gamut is sometimes the only way to get the accuracy proofing systems require.

Screening and dot simulation

Once vector and line work has been rasterized into pixels it needs to be screened into dot patterns. Most inkjet screening is stochastic / chaotic or something similar. Randomizing the dot patterns can create beautifully smooth tones but they are certainly different from the patterns typically seen on press. Some RIPs will simulate the dot patterns on press but it is tough to do this effectively as each dot needs to be color managed to simulate press inks and then they require blending as is seen on press. While this type of proof can illustrate dot pattern problems that may occur on press (like moire) its limitations are slowing its acceptance and most proofing is accomplished with non half-tone screening.

Direct Channel Control

Controlling a 4 channel CMYK printer using 3 channel RGB numbers means giving up on influencing the black channel. You may be able to control black better than the "canned" separation that occurs in a printer driver. Controlling a (true) 6 channel printer using 3 channel RGB numbers is even worse. On the other hand, RGB profiling solutions tend to cost less than CMYK and considerably less than 6-channel profilers. Like everything in life, greater control tends to mean greater complexity and cost. If you want it however, it's typically there for a price.

Color Management

The CMYK to CMYK color transformation that is required for correct proofing can be accomplished in desktop applications but it is much simpler to perform it in-RIP. When the proofing transform is moved to the RIP, any application can print CMYK and create good proofs. All applications (and all users) will create the same quality proof and page elements such as EPS graphics - which are typically un-manageable in applications - will proof correctly in-RIP. Another in-RIP capability which is often overlooked is the substitution of vector colors. Important client logo colors can be carefully formulated when tuning the RIP so at print time they are pulled from the Postscript print stream or PDF and substituted with the tuned color for your RIP.

In Summary

So if a RIP is so powerful then why would we ever print without one? Well, with the power comes the price of complexity. In the past few years RIP interfaces and installation procedures have improved significantly but they still require more time and knowledge than printing from drivers. Also, drivers are included with the printer but RIPs can cost significantly more. They typically run on a separate computer so that also needs to be figured into the price. What are the limitations with drivers? The most important limitations are the inability to linearize and the lack of direct channel control. If you use third-party papers or inks in your inkjet and find that no matter what you do it clobbers your shadow detail then linearization may be just the thing you need. If you can't keep neutrals neutral, find saturated colors muddy or struggle with a lack of dark saturated colors then direct channel control and its ability for you to specify black generation may help you as well.

It's fair to say that a RIP will almost always give you better results than simply controlling your printer with standard drivers. It's certainly correct to say that a RIP will complicate your life and add to the cost of your printing system. If any of the above points rings true for you then it might be time to consider a RIP. If not, consider yourself lucky and forge ahead with your driver. If you are really not sure then try a good custom profile using your driver. It's a comparatively cheap experiment to see if the driver will do the job before heading down the RIP road.

Thanks for reading,
Steve Upton


Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox