Re: Step-by-step through a CLUT...
Re: Step-by-step through a CLUT...
- Subject: Re: Step-by-step through a CLUT...
- From: email@hidden
- Date: Wed, 17 Oct 2007 19:16:33 -0700
Some points to clarify regarding the nature of CLUTs.
> Ken Fleisher (Colorsync-users Digest, Vol 4, Issue 374) writes:
>
> The nodes in the CLUT are evenly distributed between the
> lowest and highest value allowable for each dimension.
The word 'even' suggests equal intervals between node points and this is
not
necessarily the case. A profile designer can choose to place these nodes
in
a non-linear way should he deem it beneficial. This placement is determined
by the shape of the input luts.
> Unless you are writing a CMM, you don't really need to
> worry about the interpolation methods.
Life would indeed be simpler if the above statement were indeed true
but, alas, its not. There are, in principle, instances when a given
profile
when paired with a CMM employing a particular interpolation scheme
can give spurious results. Case in point. Most B2Ax transforms to output
devices typically use Lab as the PCS. In that CLUT the neutral axis is
typically
represented by a column of nodes aligned along the L* dimension. Employing
either tri-linear (8 nodes) or tetrahedral (4 nodes) works fine in
evaluating to a
smooth set of neutral inkings. However, there exist some output profiles
whose PCS is XYZ and, in this case, the set of nodes containing the neutral
inkings lie along the diagonal of the CLUT cube. Employing tri-linear
interpolation for B2Ax transforms built using XYZ as the PCS yields
corrupted
inkings - significant scalloping of what should be smooth ink curves
occurs.
Whereas employing tetrahedral interpolation on that very same profile will
yield the expected smooth ink curves. Most robust CMMs these days employ
tetrahedral interpolation exclusively.
Harold
TOSHIBA AMERICA BUSINESS SOLUTIONS 2 Musick, Irvine, CA 92618
" The sunlights differ, but there is only one darkness. "
Ursula K. Leguin
This message (and any attached files) is secret, confidential and proprietary and is intended solely for specific addressee(s). If the reader of this message is not the intended recipient, please notify the sender immediately by reply email and delete it and all backup copies immediately.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden