Original scientific paper
Visual estimation of colour intensity - Application in colorimetric determination of the concentration of air contaminants
Mirka Fugaš
; Institut za medicinska istraživanja i medicinu rada, Zagreb
Abstract
Factors were investigated influencing the accuracy of visual estimation of colour intensity of solutions, made by comparison with a limited number of standards. Investigations were carried out with the yellow coloured product of ammonia with Nesslers reagent and for the magenta coloured product of NO2 with Saltzman's reagent. Preliminary investigations have shown that: (I) the observers were not able to distinguish colour intensities if the optical density of the examined solution was greater than about 0,5, (2) there were marked differences among observers as to the ability of judging colour intensity, (3) most observers were inclined to identify the sample with a standard or to place it just in the middle between two standards, (4) in consequence of (3) relative errors for low concentrations were rather high. On the basis of these results and taking into account Weber's law stating that the minimal noticeable difference between two stimuli and the basic stimulus are in a constant ratio over a wide range of intensities, the assumption was made that a logarithmic sequence of standards might prove best for practical visual colorimetry. Therefore, a logarithmic series of standards was prepared and compared with two different equidistant series. Enabling the observer to make the best use of his ability to differentiate colour intensities, the logarithmic sequence of standards proved the best for this purpose. In addition, with the logarithmic series the relative error of estimation was of the same order over the whole range of concentrations covered by standards. It is advisable to test candidates before using them as observers for visual colorimetry. With the logarithmic series performed, good observers are those making an average relative error below 5.5% and no error of individual estimation above 15%. The observers making estimates deviating from the theoretical value for more than ±1/2 the interval between standards should be excluded.
Keywords
Hrčak ID:
180745
URI
Publication date:
25.3.1964.
Visits: 1.181 *