Re: On the effect of fluorescence
Re: On the effect of fluorescence
- Subject: Re: On the effect of fluorescence
- From: Graeme Gill <email@hidden>
- Date: Tue, 01 Mar 2005 12:34:42 +1100
- Organization: Argyll CMS
Roger Breton wrote:
I'm comparing two sets of measurements made from the same litho press sheet.
Up until now, I was under the impression that the UV-cut filter found on the
DTP41UV and, possibly, the Spectroscan (and other UV-cut fitted instruments)
would mostly affect colors with L* > 70, because colors in that range become
progressively affected visually by the hue of the paper which is the vector
of fluorescence excitation.
My experience is that UV has an almost equal effect (in terms of Delta E),
irrespective of the L* level. This implies that it will affect almost
any colorant.
But, no, it does not seem that way in the face of some new evidence which
never occurred to me. Based on the two sets of data I'm working on, it's
plain to see that fluorecence seems to have practically no effect on solid
cyan (100%C with L*=56), surprisingly little effect on solid yellow (100%Y
L*=88), barrely detectable effect on a solid black, but what a difference on
on solid magenta (100%M L=46)! Almost 2 DeltaE, in fact. That's not just
random instrument variation.
The maximum I've measured (in rather limited experiments) was between
illuminants A and D65, and a level of 90% black colorant, which had
a delta E of 5. This was with a well known chemical proofing process.
This was a very simple experiment comparing the L*a*b* values read by
a Gretag Spectrolino, without and with it's D65 filter.
And say that I thought that the net effect of fluorescence would generally
be to fool the instrument, and the profiler, to see more blue into colors
than there 'perceivably' is. Hence, the typical compensation of profilers to
throw more yellow into the mix to compensate for all that extra blue. But I
realize, now, I was wrong.
You can't generalize like that. The change will depend on the spectral
difference between the illuminant in the instrument for the input profile,
the illumination in the instrument for the output profile, the
level of FWA (Fluorescent Whitening Agent) in the input paper, the
level of FWA in the output paper, and the illumination for viewing the input
and output prints.
Note that (very roughly) light sources relative UV component
ranks something like this:
Incandescent (A) with UV filter Low
Office Fluorescent light Low
Incandescent (A) Medium
Daylight (D50, D65) High
and that most popular instruments use an Incandescent source (A).
So, if you have some paper with a high level of FWA, and you measure
it with an instrument using A with a UV filter, and then view it
in an incandescent illuminant, it will look "bluish" by comparison
to the instrument reading.
If you have some paper with a high level of FWA, and you measure
it with an instrument using illuminant A, and then view it
in office fluorescent illuminant, it will look "yellowish" by comparison
to the instrument reading.
If you have some paper with a high level of FWA, and you measure
it with an instrument using illuminant A, and then view it in a
daylight illuminant, it will look "bluish" by comparison
to the instrument reading.
If you have some paper with a low level of FWA, your measurements
and visual appearance will tend to agree, irrespective of the levels
of UV in the illuminants of the instrument or viewing conditions.
Now, depending on which of these situations are the input profile
situation, and which is the output profile situation, you will
get opposite effects.
Note that the amount of UV a Fluorescent light emits is under the
control of the designer of the lamp. Most common ones have relatively
low UV, because the phosphor is sucking it all up to produce
visible light. Some of the main wavelengths from the mercury may
punch through, depending on the thickness of the coating
(not much one hopes !). Fluorescent lamps designed to be D50
simulators may have some long-UV emitting phosphors added to
them, which give a UV to visible balance closer to real daylight,
but the response of paper with FWA to D50 simulator Fluorescent
lights may vary wildly. (ie. I suspect not enough attention is
paid to this aspect).
Incidently, would anyone know the difference between X-Rite and
GretagMacbeth UV-cut filtration? Is there an ASTM standard on the
specifications of UV filtration, by any chance, or is it to the good
judgement of an instrument manufacturer?
I don't have the definitive answer to this, but an impression I got was
that UV cut filters came out of the paper measuring camp, where they
wanted to be able to measure the underlying paper characteristic,
without the FWA getting in the way. It's use for colorimetry could
be rather doubtful.
Graeme Gill.
_______________________________________________
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