I'm about to buy an instrument to measure the colors emitted by my
computer's CRT monitor, and the colors printed by my printer.  I can
then put those numbers into established software, to create standard
mechanisms to ensure that my prints look as much as possible like what
I see on my screen.

I have a choice between buying a "colorimeter" (measures colors via
red-blue-green components, like the human eye) or a
"spectrophotometer" (measures actual spectrum distribution curves,
from 400 to 700 nanometers of wavelength, the visual spectrum).

Generally, a spectrophotometer is much more advanced and gives much
better data.  But in one place, I read that "colorimeters are much
better than spectrophotometers for the most common CRT phosphor types
'P22' and 'EBU', because these phosphors have multiple spikes at
wavelengths that confuse a spectrophotometer."  I found out that my
monitor has a different phosphor type, "B22".

So my question for you is:  "Can a 32-band spectrophotometer give good
results in calibrating a CRT monitor using the 'B22' phosphor type? 
Or should I use a colorimeter for this monitor instead?"

---

Clarification of Question by magic_mark-ga on 24 Jul 2002 20:52 PDT

My CRT monitor is a Philips "Brilliance" model 107P.  My thanks to
"actinide-ga" and "west-ga" for their help.

This is a tough one to answer using publicly available web resources. 
To start with it does appear that P22, B22 and EBU are discrete
phosphor types, e.g.:

<http://support.microsoft.com/default.aspx?scid=KB;EN-US;Q140902&>

lists many monitors each listed as having one of the three cited
phosphors.

But looking further, the PDF file "Color accuracy - problems and
solutions"

<http://support.necmitsubishi.com/nec/common/library/Coloracc.pdf>

has a table on page 2 titled "Example:  B22 (P22)" which strongly
implies that B22 and P22 are equivalent, or at least have very similar
CIE colour coordinates.

Similarly,

<http://www.guardian.com.au/monitors/mitsubishi.htm>

refers to "Phosphor: B22(EBU)," implying these two terms to be also
closely related.

Looking at commercially available color calibration systems now,

<http://www.xrite.com/Products/Product.asp?Show=Description&id=11>

descibes a colorimeter that "boasts improved monitor calibration
performance over spectrophotometer-based solutions," providing support
to the idea that spectrophotometers are generally less suitable for
monitor calibration.  The next page

<http://www.xrite.com/products/industry.asp?ID=87> 

links to product descriptions implying spectrophotometers to be more
normally used for measuring printed output rather than monitors, at
least in high-end systems.

The optical emission security FAQ 

<http://www.cl.cam.ac.uk/~mgk25/emsec/optical-faq.html> 

notes the following: "The vendors of colour computer monitor vendors
currently do not provide any useful information about what screen
phosphor type is used. The designation "P22" found sometimes in
datasheets is mostly meaningless, as it describes any combination of
red/green/blue phosphors that fulfill the NTSC TV colour
specification."

None of this answers your question though.  I would put the question
to the manufacturer(s) of the products you are considering, or
alternatively try asking it on an appropriate usenet group e.g.
sci.engr.color

<http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&safe=off&group=sci.engr.color>

seems to have particularly knowldegible participants in this field, as
shown by the two threads turned up by the following google search:

<http://groups.google.com/groups?hl=en&lr=&ie=ISO-8859-1&safe=off&q=phosphor+spectrophotometer+colorimeter+group%3Asci.engr.color>

The same newsgroup has a thread addressing the question "what is p22"

<http://groups.google.com/groups?q=phosphor+group:sci.engr.color&hl=en&lr=&ie=UTF-8&safe=off&selm=353E55D9.A62%40philabs.research.philips.com&rnum=5>

which supports the opinion above.  I'm sure if you asked on the
newsgroup sci.engr.color

<http://posting.google.com/post?cmd=post&enc=ISO-8859-1&group=sci.engr.color&gs=/groups%3Fhl%3Den%26lr%3D%26ie%3DUTF-8%26safe%3Doff%26group%3Dsci.engr.color>

you'd find plenty of informed and opinionated answers to your
question.  From what I've seen though, I suspect the consensus answer
will be that colorimeters are a superior option for measurement of
monitors using any phosphor system, for the reason you state in your
question (unless you want to pay an enormous sum of money for a
high-end spectrophotometer).

Hello magic_mark,

According to the manufacturer 'X-rite', abridged spectrophotometers
that measure in 32 bands are not adequate for monitor measurement. The
reason is the errors resulting when measuring the spiky spectral
curves of the two most popular red phosphors used in color monitors.
 
X-rite states that if a spectrophotometer is used it must be capable
of measuring at wavelength intervals of 1nm (or, at worst, 4nm
intervals).

You can access the X-rite technology backgrounder article with a full
explanation of the above by using the following link and downloading
the document " XTB_OnScreenColor.PDF"

http://www.xrite.com/HelpDesk/DocDetail.asp?DocNr=367

There appears to be no reason why the above advice would differ
between P22 and B22 phosphors. The B22 phosphor seems to be used in
CRTs with an aperture grille, whereas the P22 type is found in shadow
mask CRTs. I believe that the red phosphor in both types has a
spectral curve with two prominent spikes.

I hope you find this information helpful.

Regards,   West