[argyllcms] Re: Averaging measurements

2008/4/22 Klaus Karcher:
>  As Gerhard mentioned, you can simply combine several ti3 files and let
> profile do the rest -- but I can understand your demand to do this
> beforehand (e.g. to check for measurement errors / inaccuracies, ...).

For now, I use verify to get dE numbers and output a vrml to see if
there's some visual patterns (probably to subtle to spot)

> argyll's profiles are simply better :-)

:-)

>  MeasureTool is great to quickly compare and average Measurement: It's
> convenient to do such things in a GUI app and it's nice to get a visual
> impression of the colors and differences you are dealing with.

True.

>  One of MeasuerTools shortcomings are the missing 3D visualization functions
> and the lack of tools to evaluate the *quality* (not only the *quantity*) of
> differences: E.g. it's nearly impossible to discover or analyze even the
> simplest systematic errors like hue, chroma or lightness shifts with
> MeasureTool.

This applies to verify too, I guess.

>  I use R <www.r-project.org/> to perform my analyses. I wrote a very simple
> function to import CGATS files and implemented some basic color related
> stuff (e.g a DE00 fuction). Of course averaging functions in all flavors are
> already present in a statistical environment like R. When I'm done, I export
> my data as text tables and add the headers manually (whereas a simple CGATS
> export function could also be written with ease.)

Ahh, I used R analysing flow cytometry data. Maybe I'll try R or
Octave for this purpose.

Martin Weberg

Other related posts: