Checking your linearization
20 janvier, 2009 par
Checking your linearization
Jon Cone

You can check the quality of your own linearization with a tool that I have provided in the form of an Excel spreadsheet called “Linearization_Checker.xls”. This spreadsheet has a form for entering in the measurements of a 21 step density chart that has values from 0 to 100% ink in 5% increments. You can enter either Luminosity values (L of Lab) or Density values if your instrument is a densitometer. After entering in the values, a chart is generated that shows the IDEAL linearization and your current linearization.

You must have Microsoft Excel or a compatible spreadsheet shareware or software in order to use this tool. You would also have to have a densitometer or a spectrophotometer or spectrocolorimeter in order to read the 21 step file. The 21 step file is printed through a K7 profile, measured and the resultant data checked in this tool. Instructions for use are next.

In illustrations 1 and 2 you can see the results of both a good linearization and a poor linearization. An ideal linearization will be within the pink horizontal line. The Ideal measurements are computed from your measured dMin (paper white) and dMax (black).

Illustration 1: A good linearization

Illustration 1: A good linearization

How far you let your system drift is up to you. But I believe the best output system will allow you to generate the best output. It is not a good idea to try and compensate through imaging. Piezography K7 is a very high fidelity system and should be in peak condition to get the best out of it.

Illustration 2: a poor linearization

Illustration 2: a poor linearization

The graph chart in illustration 2 is separated by the diagonal linearization line. The pink represents the ideal linearization. The black line represents the measured linearization. If the black line rises above the pink line it indicates that part of the grayscale when printed is too light. When the line tends to vertical or horizontal instead of diagonal, it represents a reversal and this is indicated in the print by posterization. If the black line falls below the pink line it indicates that part of the grayscale is printing too dark. Naturally, a uniform diagonal line is preferred, but some wobble within the pink is acceptable. Also, some users choose to have custom Piezography Type 2 profiles made which print the shadows darker and will be represented on a graph as dropping below the pink in the shadow region.

If you have Microsoft Excel and a spectrophotometer to measure, you can download the Linearization Checker (1016kb Excel and PDF reader required).

CUSTOM PROFILING

At this point you may have decided to have a custom profile created for your printer. A K7 or K6 curve is required for each different media/printer combination on which you wish to print. Most users have a single paper they like. Some photographers are sensitive to coating changes from batch to batch. My own practice in my own studio is to profile each new batch of paper and I do a density confirmation with each new box of paper. This way, I try to keep my own printers in sync with the paper manufacturers. Some papers have proven to be very on-and-off. One of them is my favorite, Hahnemuhle Photo Rag. This paper is very prone to ink bleed and coating changes that affect how Piezography inks will print.

Printers also age. It is the print head which has only a finite amount of drops of ink that it can form before the piezo electrics which generate drops of ink begin to fail. Epson has published the amount of drops of ink that its older heads can produce. One can only conjecture as to the amount of drops modern Epson heads can produce. Certainly, the ability of a head to form all of the multi-sized drops is essential to having a perfectly linearized printer. As a head begins to produce more or less dot gain on paper as a result of aging, new linearizations can be made gradually, or as needed.

If you want custom profiles, it may be time for you to read Custom Profiling.

Checking your linearization
Jon Cone
20 janvier, 2009
Partager ce poste
Archiver