Printer Resolution, Line Pairs per mm

[mcmorrison]

Hello,

Inspired by Ctein's thoughts on print resolution in his book, "Post Exposure", I did a crude test of the resolution capabilities of my iPF5000 on Epson Enhanced Matte and Premium Glossy.

I find that using the standard driver (not the plug-in), and the highest resolution settings (16-pass), I can resolve no more than about 5 line pairs per millimeter, or (if my math is correct. . .) 240 dpi.

I attach three files here: the tiff from which I printed the tests, and scans of the print out on gloss and matte. The tiff is 1-bit deep at 8096 dpi, and the scans are 8-bit grayscale at 2024 dpi.

Ctein tells us that 30 lp/mm is "perfectly sharp" and is at least theoretically achievable with film and enlargers, though 10 lp/mm and less is more common. We clearly see much improved color gamut and image control using inkjets, not to mention vastly increased paper options, but are we giving up  sharpness in the inkjet age? Do we care?

I don't see much discussion on increased resolving capabilities for inkjets, and wonder whether I have missed something, done an flawed test, or whether we are simply accepting a step down in sharpness in trade for all the other exceptional virtues of inkjets?

This also makes me wonder about the discussion of "native" resolution of our printers: if the printers can't resolve more than about 240 dpi, then what, if anything, changes when we send a 300, 360, or 600 dpi file to the printer?

Best Regards,

Michael Morrison

[attachment=1981:attachment]

[attachment=1982:attachment]

[attachment=1983:attachment]

[feppe]

I don't know - or care - about the math involved, but I imagine that current inkjet printers are capable of producing results sharp enough for normal eyesight at normal viewing distances. This meaning that any marginal improvement in resolution is lost on most people, except those who do academic research or commit to pixelpeeping.

[jerryrock]

Ink jets are stochastic printers, they do not print true halftone patterns which have an orderly arrangement of dots in lines. The frequency of dots in a halftone pattern is expressed in lines per inch. Ink jets printers print a scattered arrangements of dots called frequency modulation which enable the printer to retain smaller detail in the image and render smoother color tones.

While your lpi test would work on a typical lithographic press, it is not an accurate test for an ink jet printer.

[Steve Morton]

Ink jets are stochastic printers, they do not print true halftone patterns which have an orderly arrangement of dots in lines. The frequency of dots in a halftone pattern is expressed in lines per inch. Ink jets printers print a scattered arrangements of dots called frequency modulation which enable the printer to retain smaller detail in the image and render smoother color tones.

While your lpi test would work on a typical lithographic press, it is not an accurate test for an ink jet printer.
[{POST_SNAPBACK}][/a]

There is an interesting (if slightly biased) discussion of inkjet resolution here [a href=\"http://www.ddisoftware.com/qimage/quality/]http://www.ddisoftware.com/qimage/quality/[/url].  I downloaded the test image, and printed it on my Canon i9950 on my best quality glossy paper.  I convinced myself I could see the line pattern on the 600dpi test even though I had to use a loupe to do it!  I also seem to remember that Norman Koren has done some actual measurements of printer resolving power, which are reported somewhere on his web site at www.normankoren.com.

However, I also feel some sympathy for feppe's viewpoint - it is the end result that matters, and NO-ONE can see 600 dpi with naked eye!

Cheers


Steve Morton

[mcmorrison]

Hello Steve,

Thank you for the link to the Resolution Test Files—they are very illuminating, and answer much of my curiosity on the topic! It is also the first place I have seen that speaks with certainty about the native resolution of Epson, HP, and Canon printers. I hope they're information can be trusted in that regard!

The printed test charts clearly show the effects, and image quality value, of printing at full resolution (600 dpi in my case with an iPF5000).

Ctein's analysis suggests that while we may not be able to resolve 30 lp/mm, we do detect the quality of an edge at that level. This too is born out by the resolution test charts you suggested—I can tell the difference, at 3 feet, between the 600 dpi diagonal and the 300 dpi diagonal.

Thanks for a valuable link!

Best Regards,

Michael Morrison