[argyllcms] Re: D65 color temperature + low DE on the i1 ?
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Fri, 11 Jan 2008 14:41:56 +1100
lee scratchy wrote:
I'm trying to get as close as possible to 6500K color temperature, and also get
a very low DE.
I managed to do it with the s2, but this is getting complicated with the i1 :(
I'm using the color.HCFR application to check that, which is regarded as pretty
accurate on the HCFR forum.
It's not clear to me why you think this. I'd imagine that absolute accuracy is
more
determined by the instrument than anything else, and that without a better
reference
instrument to check it against, it's not actually possible to know this.
You might well be able to make an evaluation of the repeatability of the
particular
instruments, since it will be one factor that influences the verification
results.
I don't succeed to get perfect 6500K color temp and a low DE.........where it
was possible on the spyder2 :(
I would guess that the i1 display is not as repeatable. It might be possible to
improve this by tweaking the driver (ie. increasing integration time, or
repeating measurements and averaging them at low light levels), but the cost
will
be in overall measurement time. Another possible solution that won't worsen the
measurement
time is to get more light into the instrument by having the projector and
screen closer together.
if you look at 's2 L HQ.chc', it was made and measured with the s2 on the same
CRT.
the color temp is almost perfect, and the DE is very low
The repeatability is good. The accuracy is unknown.
so what's going on ?
it's the same display........I appreciate the fact that the i1 is more
sensitive, but if I can't reach
the same level of excellence as with the s2, it's pretty much useless to me.
It's got a different measurement time/repeatability trade-off. The Spyder 2
driver in
Argyll has been tweaked to improve low light repeatability, since the naive way
of using it was noticeably worse than other instruments (including the
i1display).
It depends on the display a great deal as well. CRT's are particularly difficult
due to the refresh flicker. Some quick checks of my i1d2 indicate that typically
it's repeatability is 0.25 - 0.5 delta E, but occasionally it will be
as bad as 1.0 delta E. Other instruments may be better, but I haven't
done a comprehensive survey.
the color.HCFR app does the unit precalibration as required per the Gretag SDK,
and they told
me it was mandatory to get the utmost precision...........is that related ? or
is my i1 faulty ? :o
This is not as I understand it. I understood that the i1display 1 needed black
level calibration,
but that the i1display 2 does not. It does need to be synchronized to the CRT
refresh though
(which dispcal and dispread do automatically).
Now it wouldn't be hard to add black calibration back into the i1display 2, and
it's pretty
easy to see if it in fact would be necessary. Place the i1d2 on a matt black
surface
(and even then cover the whole thing with something opaque), and do a spotread.
What you hope to read are numbers close to, but not quite zero. For my i1d2 I
get:
XYZ: 0.000039 0.000021 0.000055, D50 Lab: 0.000193 0.000730 -0.000715
which looks excellent. None are zero (which would indicate that the offset is
too much), while the numbers are smaller than the black you will typically
read from a display.
The i1display 1 was nowhere near as good, and did need black calibration
for best accuracy.
Graeme Gill.
Other related posts:
