There have been several notes on the general subject of the reproduction accuracy of photocopiers and laser printers.  Some years ago I had product management responsibility for IBM's copiers and laser printers so I have a bit of background in this general area.  Technologies change, but the principles seem to be the same.  Neither office copiers or desktop printers claim to be high precision devices and are really designed to reproduce printed text efficiently and with good resolution.  They are not designed to provide accurate reproduction of high precision graphics, although both do a pretty fair job.
 
Basically, almost all copiers use an optical system to transmit the image from the paper to be copied to the reproduction engine.  The optical system, which consists of several lenses and mirrors, can be calibrated and adjusted to make perfectly accurate copies but the process is very involved and time consuming so many field engineers do not try to make all of the required adjustments.  Most copiers from reliable manufacturers are calibrated at the factory to provide images that are very close to 1:1.  There will be some variation in the field because some of the motors in the system may be slightly voltage sensitive which may cause variations in the speed of paper in the paper path and may change the rates that the optical system moves in relation to the photoresistive drum.  If you are making copies at a copy center, all you can do is try several copiers and use the one that does the best job.  Usually, the big, fast machines do a good job.  Small copiers often use a different imaging system that suffers from the same problems as laser printers.
 
Small laser printers use an imaging system that usually consists of a single row of LED's that expose the photoresist.  The latent image on the photoresist is transferred to the paper in the developer where the toner is added.  The resulting image is then fused by a heated platen.  Small laser printers often have trouble printing in exactly the same resolution in both the horizontal and vertical directions due to the variations in the speed of the paper through the machine, which may vary depending on the line voltage, paper weight and surface finish.  This manifests itself in elliptical circles and unsquare squares. The error is often too small to see with the naked eye, but can be measured with an accurate rule.  To compensate, you can try adjusting vertical and horizontal scales for a particular paper weight (which I do on the Personal Astrolabe).  Even that may not be perfect because papers change size differently when they are heated depending on the paper grain and moisture content.  PostScript has many features that allow you control this sort of thing and the microcode parameters can be adjusted for most printers to provide some compensation.  You are then in the position to start all over if very high accuracy is needed when the humidity is very high or very low or the line voltage varies .  I use a scale divided into 1/64 inch and try to estimate the center of lines with a 10x loupe.  I consider it close enough if it seems accurate to this precision, which may not be sufficient for more demanding applications.  I use my test sheets for scratch paper.  The pile is currently about two feet high.
 
Summary; You can't rely on either copiers or small laser printers to produce exact figures that require high two dimensional accuracy without detailed attention to a lot of variables.  However, either will produce images that are accurate to within about .5 mm if properly adjusted.  Laser printer images can be adjusted by software to compensate for variations in paper path speed.  One thing that I have never been able to find a solution for is page to page registration.  You must print registration marks if pages must match as in color separations or photomasks.  Professional print shops have very expensive printers that will reproduce anything with very high accuracy, but the cost is high, particularly for a short run.
 
Getting accurate images on paper is kind of like getting to Wink, Texas; you have to want to get there to get there.
 
Best regards,
 
Jim
 
James E. Morrison
Astrolabe web pages at: http://myhouse.com/mc/planet/astrodir/astrolab.htm

Reply via email to