RE: Colorimeter vs. Spectro
RE: Colorimeter vs. Spectro
- Subject: RE: Colorimeter vs. Spectro
- From: "Bertho" <email@hidden>
- Date: Tue, 4 Jan 2005 20:32:01 -0500
I hate to see a thread closed that is not providing correct information.
First, I am not claiming either type of instruments to be better or worse or
referring to good or bad brands, I just like the information to be clear.
What does a colorimeter actually consist of?
There are basically three detectors and three filters to simulate the eye's
response.
Why not talk about the possible error sources of it too. As previously
mentioned the filters are critical but that is not the only problem areas.
How stable are the three sensors and the attached amplifiers?
How stable are the three channels versus temperature? If the red channel for
example, changes sensitivity more than the other two channels, the result will
not be accurate.
What are the long term drift of the electronics and the sensors?
What is the wave length response of the detectors themselves, even if the
filters are perfect?
How linear are the sensors and amplifiers versus signal level?
As can be seen, there are many possible error sources. A final thought; how
can a user calibrate the instrument in the field.
Now, a good spectrometer can be made with extremely high resolution, spectral
peaks do not magically fall into some undetected space and more data is great,
not worse. The instrument can be field calibrated and provide general purpose
accurate data.
And the best part, not worst, is that the data can be mathematically filtered
to provide any desired filter shape with any desired mathematical accuracy, not
limited by the manufacturing tolerances of the filters and amplifier responses
and drift of the colorimeter. Similarly, noise can be better handled
mathematically then by hardware.
That does not mean that there are not error sources in a spectrometer. It too
has linearity, noise and spectral accuracy problems but they are often easier
to calibrate and control.
At the end, it depends on the quality and cost of the equipment, but
downplaying the spectrometer does not make sense.
Bertho Boman
> -----Original Message-----
> From: Marc Levine Sent: Tuesday, January 04, 2005 9:27
>
> In hopes of providing some closure to this thread, I would say this. The
> colorimeter and spectrophotometer have different missions in life. For a
> spectro, the system is geared to deliver a discrete set of spectral points
> that represent a color sample. Yes, I would agree that most spectophotmeters
> do see the data that is in between those discrete points. However, because
> the task is to report these discrete points, the data in between those
> points can be heavily "discounted". After all, a system does need to handle
> "noise", and needs to use some type of math filtering. In contrast to a
> colorimeter, a spectro is not geared to filter light like the human eye -
> only to capture/report these tiny spectral points. And yes, you might even
> be able to tune a spectro to handle the red phenomenon that occurs when
> measuring displays so that it works more like a colorimeter (and more like
> the human eye). The tough part about having a combination device for both
> display and print is that - if you "tune" it to yield a certain result with
> displays, you may compromise the way it reads prints (and vice versa).
>
> The colorimeter's mission is quantify color using the broad-band energy
> levels of a color sample. There is far less parsing out of what's good and
> what's bad about your data. In this type of case, if the filters are built
> correctly (a HUGE point that was made earlier), a colorimeter has the
> capability to capture color data that has even greater correlation to the
> human eye than a spectro.
>
> -Marc
> --
> Marc Levine
> Sales Guy
> Technical Guy
>
> X-Rite Incorporated
_______________________________________________
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