[argyllcms] Re: Determining proper error value for -r

On 2006 May 26, at 3:39 PM, Gerhard Fuernkranz wrote:

> Ben Goren wrote:
>
>> Would that  be time and  resources that us  non-C-hackers could
>> help  with? If  it  means,  for example,  some  grunt  work  of
>> creating and measuring charts or hacking together Perl programs
>> or  spreadsheets to  analyze them,  I'd be  more than  happy to
>> volunteer.
>
> As it looks like you  are obviously pretty interested in finding
> the  best -r  value, probably  you could  do something  for this
> purpose.

I think I understand the approach you're suggesting. With a bit of
luck, I'll be able to hack  on something this week. But if I don't
get to it this week, the next couple are looking to be most hectic
at work--a couple people are going on vacation, one is expecting a
baby, etc. So it'll either be this week or a month or so....

However, it also  occurred to me: Argyll already  has the facility
to deal with  multiple readings of paper  white. Is this something
specific to paper white, or is it generic? That is, would there be
anything  to be  gained by  creating a  ti1 file  the normal  way,
reading  it multiple  times, merging  all those  multiple readings
into  a single  appropriately-munged  ti3 file,  and creating  the
profile from that?

Perhaps even a  combination of the two would be  in order. Use the
program whose  pseudocode you outlined  to figure out the  best -r
value,  but  generate  the  profile using  all  the  ``duplicate''
values.

Or,  how  about  averaging  the  values  an  generating  a  single
synthesized ti3 file from said average?

I'm sure  somebody who  knows much  more about  statistics, signal
processing, and  the like would  be better suited to  answer these
questions than I am....

Cheers,

b&

Other related posts: