IMATEST comparison of Pentax Photo Laboratory and Adobe Photoshop CS RAW converter

[Herb Chong]

i've been playing a little bit with IMATEST, Norm Koren's program to analyze 
color, resolution, noise, and linearity of digital cameras. you can read 
more about it at www.normankoren.com. i've posted a set of results from 
doing the color test at various ISO settings and looking at color fidelity, 
noise, and linearity. you can read about the test and how to interpret the 
graphs at http://www.imatest.com/docs/colorcheck.html. the details are 
pretty technical so i posted all of the output graphs instead of trying to 
explain too much. here are the results 
http://users.bestweb.net/~hchong/IMATEST_Results/ as output by IMATEST. all 
of the EXIF details for my six test images are contained in 
http://users.bestweb.net/~hchong/IMATEST_Results/PentaxPhotoLaboratoryImageData.csv. i 
shot all of these outdoor with direct sun on the color chart. this reduces 
contrast by having the darker zones coming out light. all images were 
defocused a fair bit to remove surface texture noise from the chart itself.

i always shoot in RAW. i used IMATEST to do a comparison of the output from 
Pentax Photo Laboratory TIFF16 to Adobe Photoshop CS RAW TIFF16 from the 
same set of RAW files. both conversions were done with with automatic white 
balance and exposure. Photo Laboratory conversion used all default settings. 
i have the Sharpening slider in Photoshop RAW set to 50 and otherwise 
everything else was default. white balance was left at As Shot.

the conclusions i draw from doing this comparison, which mainly shows the 
difference between the Pentax and Photoshop converters is that the Pentax 
converter is significantly worse at color accuracy and noise reduction. the 
Photo Laboratory conversion rates as very good for color fidelity while 
Photoshop CS RAW rates as excellent or better. the noise from Photo 
Laboratory at high ISO shows significant chroma noise while the noise from 
Photoshop CS RAW is mostly luminance noise. the noise spectrum of both 
converters is really different too. it suggests that Photo Laboratory does 
little or no noise reduction and that the noise grows rapidly with 
underexposure. Photoshop CS RAW seems to do the same types of noise 
reduction that most digital cameras do.

i didn't shoot enough images to be sure, but the sequence of images using 
ISO 200, 400, 800, 1600, 3200, and 200 was interesting enough to make me 
want to try something. the second ISO 200 exposure tested worse than the 
first one. i wonder if i shoot a full burst of 6 images at ISO 200 whether 
there will be a progression of decreasing quality from the first image to 
the last.

i need to do this test with the Thumbs+ Digicam RAW 1.5 plugin to see how it 
fares. i wonder how other converters will do.

Herb....