Re: Delta E 94 as new method to measure color variation
Re: Delta E 94 as new method to measure color variation
- Subject: Re: Delta E 94 as new method to measure color variation
- From: Robin Myers <email@hidden>
- Date: Fri, 8 Oct 2004 06:08:32 -0700
On 8 Oct 2004, at 05:12, Roger Breton wrote:
The current consensus among color scientists is that the dE94
equations
are better for small color differences, ie. less than 5 dE, but that
they are not very good for large color differences, with the dE76
equations still useful for these.
I was under the impression that DeltaEab was also meant to convey
measurement of small differences as well? Anything above 8 to 10
DeltaEab, I
seem to recall, falls under 'suprathreshold' differences.
The Delta-E(ab) (I will use the term dE76) metric was "intended" to be
a general method of conveying color differences, both large and small,
but what it "achieved" was different. The basic concept of the dE76
metric was that 1 unit would be a JND (just noticeable difference).
Research has shown that people can differentiate small differences at
finer values than one dE76 JND. dE94 is an attempt to provide a metric
that more closely agrees with people's experience. A unit of dE94 is
about half of one unit of dE76.
Even the dE94 unit can be too large when looking at color tints in
neutrals. It has been my experience that in the right conditions, some
people can see a much smaller color difference than one dE94 unit.
Color differencing can also be an industry specific topic. The CMC
difference metric (officially designated CMC(l:c)) has been used in two
versions, denoted CMC(1:1) and CMC(2:1), where the numeric designations
respectively refer to the "l" and "c" values, which are weighting
factors used in different industries. The "l" and "c" values are
divisors in the equations, so they weight the lightness and chroma
relative to the hue. The fabric industries tend to use the CMC(2:1)
where other industries use the CMC(1:1) metric, or other variants. For
more information on difference metrics I suggest you consult "The
Measurement of Appearance, Second Edition" by Richard S. Hunter and
Richard W. Harold. It has an excellent history of many difference
metrics and puts them in historical perspective.
Robin Myers
_______________________________________________
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