Re: Transmissive Spectrophotometers
Re: Transmissive Spectrophotometers
- Subject: Re: Transmissive Spectrophotometers
- From: "Wolf Faust" <email@hidden>
- Date: Thu, 1 Feb 2001 09:31:44 +0100
- Priority: normal
On 31 Jan 2001, at 15:46, colorsync-users-request@lists wrote:
>
Can the Spectrolino do this? Using the SpectroScan table? And can one
>
"trust" the XYZ data? I know once, LightSource's Colortron II was able to
1. I would strongly recommend he Barbieri Spectro 100 or Spectro 100xy
(see
http://www.BarbieriElectronic) over the Spectrolino/SpectoScan for
transmission measurements. The price is pretty much the same but the
Barbieri does offer 3nm steps from 380nm to 780nm, automatic measurement,
a 3 or 0.5mm measurement diameter, repeatability <0.02dE(!). The main
problem is software support and the lack of features like emissive
measurement like the Gretag offers. But IMHO for transmission measurement
definitly the better device in my oppinion.
2. Trust the measurement??? Well, typical inter-instrument agreement
between spectrometers of the same type are usualy specified arround dE 0.5
and 1.
I have done some test here with an X-Rite Swatchbook, Gretag Spectrolino
and the Barbieri Spectro 100xy. 10 measurements were made of an IT 8.7
target (Kodak D3 paper, black backing, D50, 2 degree observer,...) and
averaged. For most colors, the difference between the spectrometers was
slightly below dE 1. But than there were colors with differences up to dE 3.5.
For instance, yellow (patch L15) is a problem with the Swatchbook. I assume
mainly because of the larger spectral resolution of the Swatchbook. Colors
with spectral responses >700nm seem to be different. Also, repeatability is a
contributing factor. Here I found the Gretag Spectronlino to be rather bad with
differences of up to dE 1 compared to max 0.2 (typical 0.03) of the Barbieri.
So measurements of different "consumer" spectrometers can be pretty
different and most likely they are all off from the correct value by a similar
fault. I think TARGA made a comparison of different spectrometers and their
faults some time ago and found errors up to dE 6.
And so far we have not talked about measuring paper with optic whiteners and
other problems.
To finish answering your question: Trust? Yes, but only up to a certain limit
that is most likely higher than you would first think off...
--
Wolf Faust Tel: +49-69-5486556
mailto:email@hidden Fax: +49-69-95409598
http://www.coloraid.de Mobile: +49-179-6924769