Geometric and radiometric evaluation of the DSW 300 roll film scanner 1

Geometric and radiometric investigations performed with two LH Systems DSW300 scanners are presented. Their performance evaluation was carried out using good quality test patterns and accurate processing methods. The geometric tests include global and local geometric errors, misregistration between colour channels, geometric repeatability and determination of the geometric resolution. Efforts were made to separate the contribution of various error sources (especially mechanical positioning, vibrations and lens distortion) on the total error. The radiometric tests include investigations of noise, linearity, dynamic range, spectral variation of noise, and artifacts. After a brief description of the scanner, details on the above investigations, analysis and results are presented. Regarding the geometric accuracy, the RMS was 1.3–1.9 μm and the mean maximum absolute error 4.5–8 μm. The errors are bounded, i.e. on the average the 3¦ (99.7%) values are 3 RMS, and the maximum absolute error 3.7 RMS. The co-registration accuracy of colour channels was about 1 μm. The short and medium term repeatability was very high. With a linear Look Up Table (LUT) the radiometric noise level is 1 and 1–1.5 grey values for 25 and 12.5 μm scan pixel size, respectively. The dynamic range is 2 D (D : : : density) with a very good linear response up to this value. One of the major remaining radiometric problems is dust. In both geometric and radiometric tests, no significant differences between R, G, B and B=W scans has been observed. These results show that the geometric and radiometric quality of the DSW300 has been very much improved as compared to the DSW200 and also other scanner models. This test was part of a long and fruitful cooperation between the manufacturer, a major user, and an academic institution and shows that honest and critical behaviour, as well as thorough understanding of the problems and a desire to search for solutions, can lead to significant improvements to the benefit of all.  1998 Elsevier Science B.V. All rights reserved.