PapEye software

Measure print uniformity objectively from a flatbed scan

PapEye Print Uniformity turns scanned print samples into three ISO/IEC 24790-based measures of unevenness, mottling, graininess and banding, in one pass. The results correlate well with visual quality evaluation.

PapEye analysing a scanned print test sheet, with numbered measurement regions drawn on the scan and a mottling value listed for each region.

Judging print unevenness by eye is subjective. PapEye Print Uniformity gives paper, board and print professionals objective numbers instead, using computational image analysis of samples scanned with a standard flatbed scanner and algorithms from the ISO/IEC 24790 standard.

Each measure is reported as a single number describing the perceived visual evenness of the printed sample. Earlier PapEye versions measured mottling and graininess as separate products; PapEye Print Uniformity combines them, together with banding.

What it measures

Unevenness is not one defect. The human eye responds differently to variation at different scales, so print uniformity is evaluated at three scales.

Measure What it detects Typical causes Minimum region Scanner resolution
Mottling Blotching a few millimetres across, and similar marbling effects Absorption differences, coating variation and back-trap mottle 25.4 × 25.4 mm 300 DPI
Graininess Fine texture below approximately one millimetre Screening, ink-film break-up and base-sheet formation 12.7 × 12.7 mm 1200 DPI
Banding Periodic stripes running across the printed sheet Drum, roller and feed-step periodicity in the press or printer 150 mm across × 100 mm down 600 DPI

If the selected region or scan resolution does not meet a measurement's requirements, PapEye displays a dash instead of reporting an unreliable value.

How the analysis works

  1. 01

    Scan

    Scan the printed samples with a standard flatbed scanner. Use 1200 DPI to get all three measures; mottling and banding can also be analysed from suitable lower-resolution scans.

  2. 02

    Analyze

    Draw the analysis regions on the scan. For each measure, PapEye filters the region to a selected band of spatial frequencies, trims the margins, divides the area into tiles and calculates the result from the variation inside the tiles. The same regions can be applied to a whole folder of similar samples.

  3. 03

    Report

    Review mottling, graininess and banding for every region, choose which result columns to show, and export the results for the whole folder to CSV for comparison between production runs, suppliers and sample groups.

The scanner–autofeed-software combination has worked extremely well in measuring mottling. It is definitely the smartest way to carry out measurements, since there’s no need to open the scanner and change the sample after every sheet, whether you’re dealing with single sheets or 20 sheets.

— Pyry Erkkilä, Laboratory Engineer

Example results

Blue printed surface with visible mottling, showing the uneven colour distribution typical of poor print quality.

Poor print quality: mottling 1.79

A marbled print with clearly visible blotching.

Blue printed surface with a smooth, uniform appearance and no visible mottling.

Good print quality: mottling 0.40

An even print of the same colour.

The eye is relatively insensitive to very fine variation, but very sensitive to patches within a certain size range. The measurements use spatial filtering and tile-based variation analysis to emphasise the spatial frequencies the eye is most sensitive to, so higher values mean more visible unevenness and the ranking of samples corresponds well with visual ranking.

Graininess is calculated like mottling, but at finer detail and from high-resolution 1200 DPI images.

Mottling (1 × 1 inch) and graininess (0.5 × 0.5 inch) regions shown as the original scan, the greyscale image and the wavelet-filtered image used in the calculation.

Scanner calibration

Calibration is not required for repeatable comparisons made with the same scanner and settings; without calibration, PapEye assumes a default sRGB response. For comparable results between scanners and laboratories, calibrate the scanner's reflectance response (its opto-electronic conversion function, OECF):

  • Scan a greyscale target containing patches with known reflectance values.
  • PapEye fits a response curve for each colour channel and measures subsequent scans through that calibration.
  • Calibrations are stored by scanner name, so a laboratory with several scanners keeps one per instrument, and a calibration can be exported to another PapEye installation.

Requirements and outputs

Operating system Windows, Linux
Application Standalone executable with no conventional installation required
Scanner Standard flatbed scanner; a list of recommended scanners is available from Tapio
Input formats PNG, JPEG, GIF, BMP, TIFF and PDF
Result exports CSV and Excel-compatible exports, result thumbnails and sample-region files
Batch processing Analysis and export of entire folders containing similar samples
Command line Available for automated and scripted analysis
Standard Algorithms based on ISO/IEC 24790

Licensing and purchasing

PapEye Print Uniformity is sold as a permanent license for a single PC. Licenses are issued per PapEye product and per machine: a license for one PapEye product does not unlock another, and old Mottling and Graininess licenses do not automatically unlock Print Uniformity.

  1. Request a demo or quote from us.
  2. On first run, PapEye displays a Machine ID. Send it to Tapio Measurement Technologies.
  3. Place the license file you receive in the .papeye folder in the user's home directory.

Frequently asked questions

What is the difference between mottling and graininess?

The scale of the variation. Both measurements filter the scanned region to a band of spatial frequencies, trim the margins, divide the remaining area into tiles and calculate a result from the variation inside the tiles. Graininess measures fine detail below approximately one millimetre; mottling measures larger blotching several millimetres across. ISO/IEC 24790 defines separate analysis scales for the two.

Why does banding show a dash for my sample?

Banding needs a nominally uniform area of at least 150 mm across and 100 mm down, scanned at 600 DPI or higher. Patch samples are sometimes too small. Scan a larger uniform printed area to get a banding result.

Why does graininess show a dash at 600 DPI?

Graininess measures fine spatial variation reaching approximately 1.5 cycles per millimetre. From a lower-resolution scan, you would measure image resampling effects rather than the printed sheet, so PapEye requires a 1200 DPI scan for graininess.

Does the selected region have to be square?

For mottling and graininess, the region must meet the minimum size in both directions but does not have to be square; the software calculates the average within the rectangle. For banding, the directions are not interchangeable: the profile runs horizontally across the sheet and is averaged vertically, so the 150 mm minimum must be horizontal.

Can I hide measurements that I am not using?

Yes. In Settings → Show results, select which of the three measurements appear in the results table. All three are still analysed when their requirements are met, and all available results are included in exports; the setting only controls the columns on screen.

What scanner resolution should I use?

1200 DPI if you want all three measurements. The ISO/IEC 24790 measurement conditions specify 1200 DPI, and graininess requires it. Mottling and banding can be analysed from suitable lower-resolution scans; our tests have shown that mottling analysis gives equivalent results from 300 DPI scans of the same samples.

Do I need to calibrate the scanner?

Calibration is not essential for repeatable comparisons with the same scanner, but it is recommended. Without calibration, PapEye assumes an sRGB response; with calibration, measurements use the scanner's measured response curve, which improves comparability between instruments.

How do the values relate to visual assessment?

Higher values indicate more visible unevenness, and in practice the ranking of samples corresponds well with visual ranking. Compare absolute values using a consistent scanner, resolution, calibration and sample-placement procedure.

Can I analyse an entire folder at once?

Yes. Batch processing analyses every suitable image in a folder with the same sample-region definitions, and the results can be exported to CSV files.

Other PapEye applications

Measure print uniformity objectively

Evaluate mottling, graininess and banding from standard flatbed scans in one application. Ask us for a demonstration, discuss your measurement needs or request a quote.