#767 Profile review: iP4600 inherits brightness 0.7 and adds K to color patches

[Alexander Hübscher]

This is a model calibration review request, not a proven third generic
code bug or a claim of measured colorimetric accuracy.

The tested driver is bjc-iP4600-series, not the iP4300 driver. It inherits
PIXMA_iP4300_params. In current master
76b0f3571a1cf3d07fe9068e87840be94d28df5a,
src/xml/printers/canon.xml still defines Brightness 0.7, Density 0.6,
Contrast 1.7, Gamma 0.9 and BlackGamma 0.6 for those shared parameters.

After correcting the Canon density-index bug, K raster points remained
inside all six solid color patches of a CUPS test page. The model brightness
is multiplied with user brightness, so an explicit user value 1.0 still
produces effective brightness 0.7. Changing only effective brightness to
1.0 before color separation removed K points from the sampled interiors.

Using identical 40x40-pixel regions and the same input raster:

Region	K points at brightness 0.7	K points at brightness 1.0
C	117 / 1600	0 / 1600
M	118 / 1600	0 / 1600
Y	112 / 1600	0 / 1600
R	109 / 1600	0 / 1600
G	111 / 1600	0 / 1600
B	111 / 1600	0 / 1600
K	698 / 1600	667 / 1600

Black text and gray regions were retained. Antialiased color edges may
still contain black points. These are raster samples, not ink-volume or
spectrophotometer measurements. The user accepted the corrected physical
output, but no iP4300 or other Canon model has been tested.

Request: review whether the iP4600 should have separate model color
parameters, with brightness 1.0 as a candidate, after the density-index fix.
Do not change PIXMA_iP4300_params globally on this evidence. Our private
deployment copies the existing parameter set and changes brightness only
for the tested bjc-iP4600-series model.

The user's later preferred CMY multiplier 2.0 is deliberately excluded
from this proposed upstream correction. Further calibrated targets and
additional paper/quality-mode tests would be appropriate before choosing
new universal color defaults. Current physical testing is paused due to
ink depletion; this request does not propose bypassing ink detection.

[Solomon Peachy]

We (==I) have no way of acting on a "calibration request". If your experimental results show that this model needs to have its own profiling parameters (and what they should be), please submit a patch to that effect.

[Alexander Hübscher]

Thanks. Attached is a narrowly scoped candidate: it adds
PIXMA_iP4600_params, copying the five existing values but changing only
Brightness from 0.7 to 1.0, and selects it only for the tested
bjc-iP4600-series entry. The shared iP4300 parameters and untested aliases
are unchanged.

With the density-index fix applied, identical 40x40-pixel color-patch
interiors had 109-118 K points at brightness 0.7 and zero at 1.0. Black
text and gray regions remained; physical output was accepted. This is
plain-paper evidence from Gutenprint 5.3.4, not spectrophotometric
calibration or validation of other paper/quality modes.

The patch applies to master 79d9332ced; seven XML scope tests pass.
The local CMY multiplier 2.0 is not included. Please review this as a
model-specific candidate, not a universal color-calibration claim.