This page was design with the purpose of making available bispectral measurements for anyone who wants to analyze them. The database is made up of sampling different printing processes under normal reproduction conditions of solid colors and a variety of fluorescent materials. Throughout the experiment a bispectral spectrophotometer (BFC-450) manufactured by Labsphere was used to measure the samples

**Links to the individual bispectral files are located in the tables were the samples are listed.
All of the data files can be downloaded here. This is a 6.5MB zipped file. It may take a long time to download, and occasional connection failures may occur. The relevant thesis can also be downloaded: SergioThesis.pdf. This is a 2.8MB PDF file.

Below there is brief discription of bispectral measurement and how to obtain colorimetric values out of the files.

Bispectral spectrophotometric instruments can make colorimetric measurements by taking into account the contribution of both the fluorescent and the reflected component to the total radiance of a sample. For the bispectral method one monochromator is located between the instrument light source and the sample to be measured. The function of the monochromator is to separate the radiation from the instrument's light source into its spectral components before it reaches the sample. The second monochromator is located between the sample and the photodetector, which separates the radiation leaving the sample surface into its spectral components.

This section is devoted to describing the process in which colorimetric values are obtained (CIELAB) from the spectral radiance factor of a fluorescent sample. Two arbitrary samples were chosen to exemplify how the values are calculated. The samples chosen were a fluorescent orange golf ball and a green fluorescent plastic sample.

The work in this investigation was done in a bispectral spectrophotometer (which gives data of complete spectrum of light emitted by the sample at each wavelength of irradiation). The procedure will have as starting point the output data from bispectral measurements.

The instrument output is a matrix with wavelength by wavelength contribution of light excitation and emission. The columns in Figure 5 correspond to the excitation while the rows correspond to the emission wavelengths, the values within the diagonal correspond to the reflected component while the values off-diagonal correspond to the fluorescent contribution. Figure 6 shows the graphical representation of the matrix form. The xy plane corresponds to the excitation and emission wavelengths while the z-axis represents the radiance factor.

After having the matrix representation of a fluorescent sample Equation 1 can be rewritten into Equation 2, where the total radiance factor is in terms of both the emission wavelengths (irradiating light) and the excitation wavelengths (light coming out of the sample) as well as the reflected radiance factor and fluorescent radiance factors. Now the total radiance will be describe with the symbol "b" instead of the "I", previously used because the total radiance now is a function depending on two variables.

The calculation of the tristimulus values starts with the bispectral radiance factor (b_T(m,w) matrix form) which is expressed in function of the excitation (m) and the emission (w) wavelengths.

The b_T(m,w) is multiplied by the specified light source (F^I_l(m)) for colorimetric calculations as shown in Equation 6. Then the resultant matrix is summed over the excitation wavelength to obtain an array, which becomes emission wavelength dependent.

Also by using equations 4 and 5 the stimulus function for the reflected and the fluorescent components can be obtained separately. Then by solving Equation 7 the tristimulus values of any stimulus function can be derived. The sum of each tristimulus value from the reflected component with the fluorescent component must equal each tristumulus value of the total radiance factor (example X_F+X_S=X_T).

For a better understanding how the math works an analogy for non-fluorescent samples can be made using bispectral nomenclature. Equation 9 shows the stimulus function for a non-fluorescent sample. In this case, the sum symbol is replaced with the integral symbol because a continuous function replaces the discrete function of the fluorescent sample. Both equations have the same function to sum over the excitation wavelengths. Since there is no fluorescent component the equation is simplified because there is no excitation dependency. Then by applying Equation 7 the XYZs can be derived.

The main difference of the bispectral instruments from the conventional spectrophotometers is the incorporation of two monochromators into the instrument. This way the measurement becomes light-source independent and the full bispectral radiance factor can be obtained in a matrix form as a function of the excited and emitted wavelengths.

On the first stage the analysis was based on: seven prints (paper with color patches of 100% CMYK and 50 % CMYK), one print (paper with patches of 100% CMYK and 40 % CMYK), and one print (paper with patches of 100% CMYK). In total they were 76 measurements. They were measured with the intention to analyze the effect of fluorescent component in color determination.

Among the different printing process used to generate the samples were: two color proofers (3M & Epson), two thermal printers (Kodak XLT 7720 & Fujix Pictography), two RIT Lithographic presses, and a combination of inkjet printers with different quality papers. Table 1 shows the complete list of printed samples used on the first stage of measurements.

In the second stage 60 samples were measured, which were considered to have fluorescent properties under normal conditions seen by an average person as well as under black light. The list of samples is shown in Table 2. The origin of the samples is broad, it ranges from textiles, plastics, crayons, highlighters to color catalogs, etc. The main purpose of this second stage is to build a small database of fluorescent materials for future study and research.

Extracted from the thesis:

Gonzalez,Sergio "Evaluation of Bispectral Spectrophotometry for Accurate Colorimetry of Printing Materials" ,RIT, Jun 2000, pp 14-23

Any comments or Questions at sgcheco@usa.net