Hi,
Roger got it right on all points.
Each T-Ref or Lab-Ref is dated and it is suggested that you use it for
one year. In practice, I have found that if you keep them in a dark
drawer, they last many years. I used to check my master instrument (an
X-Rite 938) against our BCRA tiles. I would then read the Lab-Ref and
T-Refs with this master instrument and then send them to my service
people in the field. The field personal would then check their X-Rite
528's, 530's, Spectrolino's, Spectro-eye's, and Eye-Ones against the
Lab-Refs and T-Refs. This allowed us to exchange measurements and know
the accuracy of the instruments. If you are only doing relative work
it does not matter. However, if you are trying to setup proofing
equipment to a ANSI-CGATS or GRACOL or SWOP standard you need some kind
of instrument metrology program. We also sent the master instrument
back to the manufacture for re-certification from time to time. This
system gave us an accuracy of about 1% or delta E of about 1.5.
When we had to exchange precise measurements with other color
scientists who made the imaging media we had to send around standards.
For example we wanted to standardize our measurements at our lab, the
media lab, and the media manufacturing lab. We made three sets of
imaged media and sent them to each site. We then took measurements
once a day for a week. We then rotated the samples on to the next lab
in the chain. We then took another weeks worth of readings. Next we
rotated the samples again and took another weeks worth of readings. We
then gathered all of the data. We did a statistical analysis of the
data. We then had correction factors for the master instruments at
each site so that we could adjust the readings to an agreed standard.
We also knew how much our instruments and media drifted over a week.
If you have strip reading instuments you have to read a strip and read
it again with a handheld instrument (like the X-Rite 938). You check
the hand held instument against the BCRA tiles or Lab-Ref. You then
compare your automated readings to your manual readings.
As you can see, making and exchanging accurate readings is non-trivial.
To do it right you need international standards, BCRA tiles, an
instrument metrology program, and statistical process control software.
This is what is needed to turn our industry from a trial and error
system to a toleranced manufacturing system. Almost all other
manufactures do this. We still have a ways to go.
Ray
Roger Breton wrote:
I went to the IDEAlliance site, but they don't show a picture of what the
Lab-Ref samples look like.
Actually, I stumbled upon one, on their site. It looks like a tiny series of
1/4" circles placed at 2" from each other on a piece of white material of
about 3" by 7". I think I saw 8 to 10 color samples in all.
Do they come with a list of the expected L*a*b
values for each of the patches?
Yes of course.
If not, how else do they function as
benchmarks for the integrity of a measuring device?
Right.
Also, I understand they come already laminated, and the site claims that
this doesn't affect the readings. Is that true in your own opinion?
I would trust that statement for, as far as I know, they are measured and
certified with the lamination on.
What is your experience with Lab-Ref? Do you use them, and how do they help
you? Would you deem them essential for keeping an eye on measuring device
behavior? Are they worth the $265 investment? And do they drift with time?
They will drift over time for sure. How fast? No idea. As for essential,
it's a function of what your measurement problems are, Marco. I think it can
help answering questions like "which instruments are telling the truth".
But, by design, it's geared toward portable instruments like a 538 or a
SpectroEye. I could conceivably use the Lab-Ref with the DTP41, as Grame
suggested. But I don't see how I could do it with an ICcolor. Also, bear in
mind that there is the issue of UV filtering. I'll bet the readings on the
Lab-Ref represents values *without* any filtering. So, if I have a DTP41
with UV filter or and EyeOnePro with UV filter or a SpectroCam with a UV
filter, well, you see the picture. The whole usefulness of the Lab-Ref, in
my view, hinges upon its stability over time. I know I want a stable
reference that will last me years. Otherwise, it's useless to buy them in
the first place. But it could be a much cheaper alternatives to a set of
Spectralon or BCRA tiles while playing more or less the same role.
It's still a work in progress. But, undoubdtedly, the quality of
measurements we get out of our instruments makes all the difference in the
quality of generated profiles. That is true for all instruments. And it's so
easy to see on comparative printouts. Hence their importance. Some
instruments are easier to track for accuracy and precision than others, like
spectrophotometers. But it should be an important concern, nonetheless. And
since all instruments drift with time, for those who like to think they do
the highest critical color work, periodic re-calibration by the instrument
manufacturer should not even be questioned. I know I do. It costs me but I
get peace of mind. It's an ISO9000 specifications anyway.
Marco Ugolini
|