I have made a change to Imatest in version 1.5.5 that affects dynamic range (DR) measurements made with the Stouffer T4110 transmission chart. I have realized that the "indistinct zones" (dark zones where edges between them are not clearly detected) represent valid data. I wasn't sure at first. As a result I have changed Imatest output to report the dynamic range with the indistinct zones as "Total dynamic range," and I changed the former (total) dynamic range to "DR distinct steps only."
This only affects the Total dynamic range display measured with the T4110, where "indistinct zones" are present. It does not affect the DR at the various quality levels (high,...), which may be the best indicators of useable DR.
I am upgrading the documentation to reflect a number of things I've learned. Some publications report the inverse of the f-stop noise as signal-to-noise ratio (SNR). The ISO 15739 noise standard has an extrapolated DR measurement whose intent is to find the dynamic range where SNR = 1, corresponding to "low quality" in Imatest. The ISO measurement, which uses a reflective target with a density of 2, is much less accurate than the Imatest/T4110 technique.
My apologies for the confusion this has caused and is likely to cause in the near future. It will make things clearer in the long run.
