Textile Printing Techniques

Improving Color to Color Register on Press

Improving Color to Color Register on Press
by Mark Coudray

Color-to-color fit is one of my pet peeves. I hate out of register images, especially ones that have made it to the general public. It takes so little to make it right in the first place. With minimal effort, your images will fit and look great. There is no need for gross, or even slight misregister. This month I would like to outline some of the more common causes of misregistration resulting from the screen, and how you can avoid them. In the second part of this discussion, we will closely examine the mechanical and printing factors to cause the image to misalign.

Register issues fall into three primary categories. They are art/prepress, mechanical soundness, and printing technique. The art/prepress deals with how the artwork is generated, how stable it is, and how it is exposed and reproduced on the screen mesh. Mechanical soundness relates to the screen and printing press you are using. Finally, printing technique focuses on how you print and the major impact it can have on image alignment. Here I will address some of the major issues in the press/mechanical area.

When talking about registration accuracy, we need to define what is accurate. A printer doing athletic numbering will have a different idea of what is in register than a seasoned process printer. I like to use four classifications of registration accuracy. These are loose (±.030",) commercial (±.010",) and critical or tight (below .005".) Process Color or Simulated Process Color using halftones requires registration below .002". Very little of the work I see in the market falls into commercial or better. While almost all automated machines on the market today are capable of critical or process color level registration, few printers deliver what the machines offer. After our discussion today, you will have a shopping list of items to check.

We must make some assumptions before we can proceed. First, the art is perfectly aligned, and all positives fit in relation to each other. We also assume the positive material is stable and will not shrink or expand during exposure. Finally, registration marks are dead on, and not merely close. The art/prepress stage is your opportunity to make sure that your images are prepared properly at the very beginning. It only goes downhill from here. Any problems at this point multiply as they move down the production stream.


Transferring the image from the film positive to the screen is our first step in the registration journey. All along the way, there are potential points of failure as we fail to accurately translate information from one step to the next. Any failure will result in some degree of registration error. How much error you can handle is determined by your work. If you are printing process, almost no error will be allowed. The following six areas should be examined carefully, and DOCUMENTED so that you know what to expect.

Mesh Tension is the first, and most important factor. Properly tensioned mesh does more to accurately translate the image information than just about anything else. To achieve this, you must have a good calibrated tension meter. Whether you are using static or retensionable frames, the meter is what makes the difference. There are significant benefits to a retensionable frame over a static frame, but there is also a cost factor to consider. In the long run, I feel retensionable is the way to go for consistently maintained tension values.

There are all sorts of guidelines about mesh tension. It is better to have higher tension than lower, but it is a balancing act. The higher the tension, the more critical the rest of the printing system must be. At very high tensions, misalignment or parallelism issues of less than .010" will have a visible effect on your work. As you become more adept at this level of critical work, increase your tension. Until then, or if your presses are not if perfect mechanical condition, opt for a lower tension.

In the twenty-three years I have studied tension and its related effects, my personal conclusion is that tensions below 22 N/cm are of little value. If you can settle at a minimum level of 25-27 N/cm, you will derive significant benefits overall. The most important part of mesh tension is all of your screens are consistently within ±1 N/cm of whatever tension you choose. So, if you decide 24 N/cm is where you want to be, your screens should range between 23-25 N/cm, and all points on the screen should be within this range. If they are not, you will experience different amounts of image stretch and distortion during the print run, which results in out-of-register printing. As you become progressively move controlled, you can increase your tension for better results.

Screen Flatness is the next consideration. Whether you are using self-tensioning frames or static stretch and glue, the flatness of the screen is important. Any screen that must be forced to lie flat will have unnatural forces applied to it. These clamping forces, whether they are in the vacuum frame, or on the press, will distort your image. The torquing or racking of the frame twists the image and distorts it along a diagonal axis. This distortion usually results on one or more corners of the image being out of alignment, usually on a diagonal axis of the image.

Mesh Count is an indirect contributor to registration error. The ability of the mesh to sustain the desired mesh tension is our concern here. When mesh is woven, there are a limited number of thread diameters available. The mill will use one thread diameter over a number of different mesh counts. The problem comes when it is no longer possible to increase the mesh count using a specific diameter thread. At some point the next lower diameter thread is chosen, and the mesh count can now be increased. This is all very logical, but it results in some very unstable mesh counts. Examples would be 245, 255, and 265 where the thread diameter is 40 microns. I won't go into the math, but suffice it to say, these meshes are particularly bad at maintaining tension. As tension decreases with an image on the frame, the area of the image increases. Depending on how old the mesh is (older mesh is more stable,) you will end up with screen-to-screen variation in image size before you even print the screen. Particularly stable mesh counts are 230, 280, 300, and 355. Avoid 245, 255, 265, (40 micron) and 330 (31 micron) as unstable mesh counts. This may be somewhat of a temptation as the unstable mesh counts also have higher percentage open areas. This means ink will go though them easier, but you cannot count on the tension of the mesh. I have included a separate article on this site which dealsi specifically with the relationship between thread diameter, open area, and mesh count. The rules in this area are changing as more sopshisticated high density low modulus polyester compounds are being developed.

Stencil coating contains and controls the deposit of ink on the garment. If your coating technique is inadequate, the ink will grow beyond the stencil confines. If your coating is inconsistent, you will have different amounts of image gain. Image gain is just like dot gain, only bigger. You will experience this as fine detail that wants to close-up. This is often seen in fine type or reverse images. It applies to all image elements, not just halftone dots. Strive for good emulsion-over-mesh (EOM) of at least 10 microns.

Your emulsion supplier should be able to recommend a good coating technique to achieve the desired gasketing profile. I prefer a three-over-two technique, where the print side is coated twice, followed immediately with three coats on the squeegee side. The screen is then dried print side down. This coating is done wet-on-wet with a dull edged coating trough.

Block-out and taping of the screen can also cause problems. Excessive block-out can cause shrinking and stretching of the mesh. This usually occurs when a fast acting block-out is dried under heated conditions. The combination of heat and the shrinking block-out can cause localized distortion. This is really a big problem if you have low-tension screens. The higher the tension, the less chance this will occur.

Progressive block-out coats are also a problem. These multiple coatings can introduce irregularity and unevenness to the backside of the screen I have seen severely blocked out screens, prepared for long runs, where the surface texture on the print side resembles the mountainous topography of rolling foothills. The irregularity results in the need for differential squeegee pressure to achieve a uniform print. In order for the highest parts of the stencil to print, the lowest parts are over pressured. This causes excessive image gain and localized registration variance.

These are the basic screen related causes of misregistration. There are other factors to consider, as you become progressively better at holding registration. The more precise you desire to achieve, the more diligent you must be in isolating potential problem areas. In next month's continuation I will closely examine multiple factors on press that result in registration problems. I will also discuss the things you can do to improve a marginal piece of equipment.

How Your Press Affects Registration

Improve your color-to-color image placement by understanding how your press works and how registration is maintained. Let's continue the discussion by delving into the mechanical and press aspects of image registration. Registration is the accurate color-to-color placement of image elements to complete a design. This is particularly important as more printers make true process, simulated process, and index color a bigger part of their sales mix.

Whether you're a manual or automatic printer, the basic principles are the same. The mechanical aspects of controlling registration focus on four areas:
  1. Press mechanics
  2. Registration repeatability
  3. Printing environment
  4. Ink transfer and deposit, and print quality
Mechanical Condition

The main functions of the press are to hold screens accurately in place and repeatedly transfer ink from the screen to the garment in the proper place. To do this, there are a number of critical factors to consider. Overlooking small details can have a huge impact on registration accuracy.

A clean press is the starting point to good-quality work. I know there are those who are too busy to keep equipment clean and well-maintained. This catches up to you at some point. High-quality work requires well-maintained equipment, period. Accumulated dirt, lint, spray tack, ink, and old grease mix together and result in big problems over time. These problems include excessive wear of registration locking mechanisms, binding, thread stripping, and misalignment of register mechanisms. Wipe the press and important surfaces daily. Shield critical locking and register points like locators and forks from direct contamination and adhesive overspray.


The screen is held in the press with one of two types of clamping mechanisms: mechanical screw-type clamps or pneumatic clamps. In either case, the clamps must be smooth-acting and firm enough to grip the frame and not work loose during the run. To avoid stripping the threads and to minimize the chance of cross threading the clamps, be careful you don't overtighten screw clamps.

This is a problem when you have steel threads and tapped aluminum. These threaded aluminum clamps are much softer than steel and are prone to stripping. Overtightening runs the risks of "galling" and seizure as the metal-to-metal contact binds. Some of the softer aluminum is scraped off and actually bonds to the steel threads. This changes the thread tolerance, causing more galling. Eventually, the whole mechanism either binds up or strips out. To avoid this, look for presses that use steel inserts, or rethread them yourself using stainless steel Helicoil brand inserts.

Most presses use two clamps: one for the head and one for the tail of the screen. In graphic arts prepress, it is traditional that the top of the image is called the head and the bottom of the image is the tail. It doesn't matter if the image is on film, a screen, or a printing plate, it is still referred as head and tail. If one or more clamps strip, you have three choices:
  1. Do not use the stripped clamp, which results in a minimally secure screen and requires overtightening on the remaining clamp
  2. Use minimum clamp pressure on the weak clamp. This may result in the clamp working its way loose during the run
  3. Fix the stripped mechanism using inserts and new threaded clamps
Manual press design varies. If the screen is held at the back, without side clamps, the clamping must be rigid enough to withstand two types of continual force. The first is the lift-and-lower motion as the screen is raised and dropped during the print cycle. The second forces are the side-to-side motions created by acceleration and braking when the press head is spun from color-to-color.

Screens also must have strong corners to avoid frame distortion. The constant side-to-side forces cause the screen to twist from a rectangle into a parallelogram. This movement results in image shift and misregistration. This is most common with wooden frames. Automatic presses also are subject to vibration and continual lifting and lowering of the head or platen deck. And, there are stronger squeegee and flood forces than with a manual press requiring greater clamping force. The same guidelines apply to automatics as manual presses, with the exception of air clamps. Here, cleanliness is even more important as accumulated grime on the plunger causes it to stick and eventually cease functioning.

With either type of press, when the screen is locked into place, there should be no play in the press head when the screen is in the printing position. Grasp the screen and apply some force to it. You should not sense any movement or distortion. If you can, there will be registration variation during the print run.

When clamping round frame bars, the point of contact must be tangent to the roller. If the clamp is off-center, you risk having the image shift during the run, or the clamp slip off during the run. The reason is simple. There is not enough contact to secure the frame. There are adapters to help hold the frame more securely to avoid this problem.

When clamping frames, the screen must be flat. As wooden frames get old, they can develop a twist or torque. This also is common with retensionable frames that have not been properly tightened during stretching. This is called a "racked condition." When a racked frame is tightened, the image is twisted. There also can be residual error of a twisted image from exposure. This is another form of localized misregistration.

The correction is simple. Lay the frame on the exposure glass, and make sure it is flat. Discard racked or twisted wooden frames, and correct the rack on retensionable frames.


There are many sources of misregistration with platens. The starting point is parallelism. All platens should be parallel to the print surface of the screen. They also must be in the same plane. If they are not, there will be highs and lows between the platens, causing the screen to stretch to greater or lesser degrees at each platen. When the screen is stretched, the image is stretched.

If screen tension is not uniform, misregistration will result. This occurs with screens that differ as little as 3 N/cm². While you may not see a color jump out of alignment right away, you will experience a loss of image sharpness. Areas of the design that should be crisp and clean are fuzzy. This is caused by small amounts of ink touchoff left after a screen prints slightly out of register. Over time, the image moves out of alignment.

With automatic presses, platen parallelism should be checked at least semi-annually. I know good printers who check monthly, and sometimes weekly. Parallelism drifts are due to metal fatigue, mechanical slipping, exposure to repetitive printing forces, excessive flash temperatures (paltens above 160° F,) and the usual bumps and shakes of the printing process. Think of adjusting your platens in the same way as having the front end of your car aligned. The more aggressively you print, the more often you should check them.

Closely related to platen parallelism is off-contact. This is the distance the screens are above the platen at the time of imaging. The greater the distance, the greater the eventual misregistration. While uniformly tight screens may initially register, the constant fatigue cycle of mesh being stretched up and down eventually causes mesh tension to drop, and mis-register to occur.

Keep off-contact to around .040"to .0625", about the thickness of a penny if you are doing halftones. If you are printing general work, the thickness of a quarter is fine. Above this, you begin to experience problems, mostly in image fuzziness. It is much better to increase mesh tension than to raise off contact. Actually, increased off contact is increasing mesh tension.

Platen flatness is another cause of misregistration. This comes most often from repeated exposure to high flash temperatures. Check your platens the same way you check your screens. Lay them on the glass of your exposure unit. They should rest with all four corners in full contact. If not, there will be localized register shifts isolated in a specific area of the image. The amount of the error depends on how twisted the platen is. There is no easy way to reflatten a torqued platen; it must be replaced.

Platen-to-platen distance is one of the most important areas to evaluate when there are registration problems. To be sure, it can have a primary effect on alignment. This is a cumulative error. The register-locator cam followers on each platen fix the distance between platens. If these get loose, are moved, or wear out, it changes the distance between three platens: the one to the left, the affected platen, and the one to the right. This causes the image to jump back and forth as it moves around the press. Besides clearly noticing misalignment, this type of error has a disastrous effect on image sharpness and also results in extreme dot gain.

Unless you are mechanically inclined, adjusting platen-to-platen registration is best left to a trained factory tech. They have special fixtures and jigs equipped with dial micrometers. For excellent halftone work, you should be less than ±.001" between platens. For general work, ±.002" is usually acceptable. A well-maintained parallel press can hold tolerances below ±.0003", platen to platen.

The absolute fastest way to mess up platen-to-platen distance is by leaning on the press or platens. This happens when the press is stopped to look at the image. One or more people gather around and put their elbows on the platen, or rest their weight against the locked-up press. This can cause huge shifts between the platens. I have measured movement in excess of .015" from a single episode. Imagine what is happening if this is a constant occurrence. Make it a cardinal offense to lean on a platen, and your registration will stay solid for a long time.

Finally, let's talk about the register lockup, which is the fork and cam mechanism that assures proper alignment of the platens. There are three points to consider. The first is the cams that align the fork on some machines. These are located to the left and right of the fork. There are either two or four. They are often eccentric cams, meaning that they are egg-shaped. As the fork rises, the cams lock the fork as it mates with the register locator cams.

The second is the surface of the register fork. This is either a "V" configuration or a "U" configuration. Finally, there are individual register locator cams. There is usually one for each platen. Some older oval presses have them on the head and tail of each platen.

This area is often a mess. It is extremely easy to have adhesive overspray, lint, and grease accumulate on the fork and cams. When gunk builds up, each cam will locate in a different position. It also causes the platen to locate higher or lower than normal, resulting in platen parallelism issues. If gunk is on the fork, you run the risk of shifting register as the amount of buildup gets pushed around by the constant contact of the register cams.

More important than either of these easily fixed issues (simply keep the surfaces lubricated, but clean) is the long-term effect. If this mixture of adhesive, lint, grease, and dirt is allowed to accumulate, you lose the lubrication value of the grease. Now you have an abrasive mixture grinding into the locking surfaces thousands of times a day. You will quickly wear a groove into the fork surfaces. The locator cams may or may not fall within this groove. Eventually, you will have to replace the fork. This can be a major undertaking. It is like back surgery done poorly, and possibly crippling your registration from that point on.

If the timing of your press is off, and you do not brake to a smooth stop during the deceleration phase, the fork will take the brunt of the stopping force. When the press clunks or vibrates at the end of the transport phase, you are out of timing. This is another common cause for rapidly losing consistent register. The press should smoothly lock up as the platens locate into the fork mechanism. Either slow the press down or retime it to come in smoothly.

There are other issues that affect registration on press, but these outlined here, if corrected, offer the greatest return for consistent quality. To prevent misregistration, understanding how your press works, and how registration is maintained is critical. Take time to learn your machines. Carefully watch during the print cycle. Keep things clean and don't force clamps. This will ensure quality printing for months.

Mark Coudray

About the Author

Mark Coudray
Mark is founder and President of Coudray Serigraphics, a textile screen printing company and Coudray Graphic Technologies, a digital imaging and prepress supplier to the industry. He was inducted into the Academy of Screen Printing Technology in 1989 and has served as Chairman of the ASPT twice. In addition, Mark is the recipient of 18 SGIA Golden Image Awards, 2 Swormstedt Awards, Magnus Award, and the Parmele Award (2001). He is past SGIA Chairman (2000). He has served a total of 24 years as an SGIA Director. Mark is a noted industry author with Screen Printing, Impressions, Print Wear, Images, and other trade publications with over 275 articles, columns, and technical papers as well as an industry presenter at tradeshows, conventions, and technical symposiums. He is best known for his extensive work in controlling color halftone printing, quality, and production management of the screen printing process.

Previous Article Next Article