You are here: http://www.DAD-opto.de --> home/english --> index --> sample texts --> Light Measurement Essay --> Para. 4.4.1


M E A S U R I N G   L I G H T (continued)

4. RADIOMETRIC MEASUREMENTS (continued)

4.4 Spectrally Resolved Measurements(continued)


4.4.1 Why Use Spectral Resolution?
----------------------------------

Key words:
FLAT RESPONSE | WAVELENGTH LIMIT | BUTTING PASS-BANDS

As stated in para. 4.0, you always have to keep in mind the WAVELENGTH LIMITs when looking at radiometric measurements.
Fig. 4.4.1-a shows an example: The detector has wide bandwidth. But anyway, its spectral sensitivity characteristics shows two sharp indents: at 600nm and at 1000nm. While the dent at 600nm will affect measurement result, the bigger one at 1000nm will not. This is because at 600nm, the signal to be measured is strong. And at 1000nm, it is zero.

    fig.4.4.1-a: detector broad-band but not flat (14 kByte)

For deriving the measurement result mathematically, we would have to multiply the spectral characteristics of illumination and of detector on a by-wavelength basis and then integrate over the entire spectrum. But in fact, multiplication and integration are done deep down in the hardware of the radiometer; you cannot access the two component spectral characteristics. You only get one resulting number which equals the ready integral value of the product curve. And you won't notice that there is some indent affecting measurement integrity.

This is why general broad-band radiometric measurements would take detectors
* not only with a bandwidth containing the bandwidth of the radiation to be measured plus some safety margin;
* but also with FLAT RESPONSE within this bandwidth.

This is not easy to get. But we can replace this ideal detector. We take a set  comprising a big number of narrow-bandwidth detectors  with BUTTING PASS-BANDS (fig. 4.4.1-b). Though the envelope of all the individual detector sensitivities varies in the same way like in fig. 4.4.1-a, this is no longer a problem. Every single radiometer knows its own detector and hence, can use its own scale factor. Which will correct every single radiometer to 100% sensitivity in its respective pass-band.

    fig.4.4.1-b: multidetector compound bandwidth (30 kByte)

This way, we would get a "big number" of radiometers. Where each radiometer gives  100 %  of the correct reading  d(Phi)  that belongs to its own narrow wavelength interval  d(lambda) .

Now let us look at the histogram-like appearance of diagram of fig. 4.4.1-b in a different way:
* Correction by individual scale factors shall be completed so that every element radiometer has  100%  sensitivity;
* an unknown kind of light (spectral characteristics indicated by green dotted line) shall be received by this batch of element radiometers;
* and the columns of the diagram now show the radiometers' responses.

Here, each column's area(!) represents  d(Phi) .
And each column's width represents  d(lambda) .
And hence, each column's hight represents the derivative  d(Phi) / d(lambda) .
That's why at the y-axis, I have written spectral(!) sensitivity and spectral(!) flux.
For the term "spectral", I remember to (1), page 3, where we find:
The prefix "spectral" is an abbreviation for the complete expression "spectral concentration of", which is defined as the quotient of the radiometric quantity taken over an infinitesimal range on either side of a given wavelength, by the range.

For completing the measurement, we sum up the areas  d(Phi)  across all the individual wavelength intervals:

    (Phi) = (sum of) [d(lambda) * d(Phi) / d(lambda)]   ,

                                  from the shortest up to the longest wavelength measured.
This is a procedure that you can call "numerical integration".

Sad to say: The "big number" of radiometers in our thought experiment would be expensive and additionally, would need uncomfortable handling.
But people developped a variety of devices that use only one set-up of hardware and/or only one handling procedure to do the complete - though spectrally resolved - measurement job.

Let us consider these devices in the next paragraph.


Literature and Link
-------------------

Subject used in source
defining the term
"spectral"
text reference (1) Publication CIE No. 63 (1984)
"The Spectroradiometric Measurement of Light Sources"
Commission Internationale de l'Éclairage
(= International Commission on Illumination),
Kegelgasse 27, A-1030 Wien, Austria

Continued: 4.4.2 Instrumentation

Contents of entire essay

Contents of entire web site


Last modified April 29th 2004 23:59