spectral sensitivities, but this information is not directly available from normal color matches (except for the S-cone fundamentals; see [15, 16]). Direct measurements of the cone spectral sensitivities made using special, color-deficient, observers not only yield estimates of the cone spectral sensitivities themselves, but also allow estimates of the transformation from the CMFs (which is sometimes preferable to the raw spectral sensitivity data).

CONE SPECTRAL SENSITIVITIES

Introduction

Since the establishment of trichromatic color theory (e.g., [17–19]), a central goal of vision science has been the accurate determination of the L-, M-, and S- cone spectral sensitivities. Studies of human cone spectral sensitivity have encompassed many fields of inquiry, including fundus reflectometry (e.g., [20]), microspectrophotometry (e.g., [21]), suction electrode recordings (e.g., [22, 23]), electroretinography (e.g., [24]), and absorption spectroscopy [25–28]. Visual psychophysical experiments in human observers still provide the most extensive and accurate spectral sensitivity data.

Arguably, the first plausible psychophysical estimates of the cone spectral sensitivities were obtained by König and Dieterici [29] (see symbols, Fig. 3, below). Since then, many other estimates have been made, notably those by Bouma [30]; Judd [31, 32]; Wyszecki and Stiles [33]; Vos and Walraven [34]; Vos [35]; Estévez [36], Vos, Estévez, and Walraven [37]; and Stockman, MacLeod, and Johnson [16]. These have been discussed elsewhere (e.g., [38, 39–41]). Until recently, the estimates by Smith and Pokorny [42] have been widely used in science and research as a de facto standard. However, newer estimates by Stockman and Sharpe [2] have improved on these and have now been proposed as the Commission Internationale de l’ Éclairage (International Lighting Commission, CIE) standard for “physiologically relevant” fundamental primaries (CIE Technical Reports: Fundamental Chromaticity Diagram with Physiological Axes, Parts 1 and 2, CIE, Vienna).

Most estimates of the cone spectral sensitivities depend on the use of dichromat observers, missing one of the three normal cone types (see the sections on protanopia and deuteranopia and on tritanopia) and the assumption—known as the “loss”, “reduction,” or “König” hypothesis—that their remaining cone classes are normal [29, 43], an approach that now has a firm foundation, molecular genetically [44, 45] and psychophysically [46]. S-cone (or blue-cone) monochromats [47, 48], which are doubly reduced in the sense that they lack both the L and M cones, are particularly useful for measuring S-cone spectral sensitivity (see the section on cone monochromacies).

Cone Spectral Sensitivity Measurements

The cone fundamentals can be estimated by comparing dichromatic and normal color matches [19, 49] and then deriving a transformation matrix based on the locations of the confusion points of the three varieties of dichromacy. These confusion or copunctal points are the chromaticities at which the color confusion lines of the dichromats converge (see the section on congenital color vision deficiencies). They correspond to the chromaticities of the missing fundamental primaries.

Fig. 2. Mean spectral sensitivity data. L-cone data from 17 L(ser180) subjects (filled squares) and 5 L(ala180) subjects (open circles) and M-cone data from 9 L1M2/L2M3 protanopes (gray diamonds) measured by Sharpe et al. [46] and S-cone data from 5 normals and 3 blue cone monochromats (filled triangles) measured by Stockman, Sharpe, and Fach [1].

The most straightforward method of estimating the cone fundamentals, however, is to measure the three cone spectral sensitivities directly. Since the three cone spectral sensitivities overlap extensively throughout the spectrum, trichromatic observers are not particularly useful for such measurements. In them, the isolation of the response of a single cone type over substantial regions of the spectrum requires special procedures to favor the desired cone type and disfavor the two undesired ones. A now classical approach is to use selective chromatic adaptation (e.g., [50, 51]) and to present a target of variable wavelength on a larger adapting or background field of a second wavelength (or mixture of wavelengths) that selectively suppresses the sensitivities of the two unwanted cone types.

However, cone isolation in normal observers is still difficult to achieve throughout the spectrum. One way of overcoming this difficulty is to use selective adaptation in genotypically identified dichromats, who lack one of the three cone types. With the S cones disadvantaged or suppressed, L- and M-cone spectral sensitivities can be directly measured in deuteranopes who lack M-cone function and in protanopes without L-cone function. Figure 2 shows the mean spectral sensitivity data obtained from 9 L1M2/ L2M3 protanopes (green diamonds), from 17 single-gene L(ser180) deuteranopes with serine at position 180 of their L-cone photopigment opsin gene (red squares), and from 5 single-gene L(ala180) deuteranopes with alanine at position 180 (open circles) (for further details, see [2, 46]). The two mean L-cone functions, which are separated by about 2.5 nm in λmax [46], reflect the two commonly occurring L-cone photopigment

Prereceptoral filters

0 400 450 500 550

Wavelength (nm)

Fig. 3. S-, M-, and L-cone 2° spectral sensitivity estimates (short-dashed, long-dashed, and solid lines, respectively) of Stockman and Sharpe [2] based on linear transformations of the Stiles and Birch [58] 10° red, green, blue (RGB) color-matching functions (CMFs) using the mean spectral sensitivity data shown in Fig. 2 as a guide. They are compared with the historical S-, M-, and L-cone estimates of König and Dieterici [29] (filled triangles, gray diamonds, and open squares, respectively). The lower inset shows the mean macular density spectrum for a 2° field (dashed line) based on measurements by Bone, Landrum, and Cains [68] and recommended by Stockman and Sharpe [2, 41] and the mean lens density spectrum of van Norren and Vos [59] and slightly adjusted by Stockman, Sharpe, and Fach [1] (solid line).

polymorphisms (see the section on protan and deutan defects). An overall L-cone mean was also derived (not shown) to reflect the proportions of the two polymorphic variants in the population [2]. Tritanopes, who lack the S cones, are less useful in this context because their closely overlapping L- and M-cone spectral sensitivities are hard to isolate by selective adaptation.

S-cone spectral sensitivity is most easily measured throughout the spectrum in S-cone monochromats (e.g., [48, 52–57]). In defining a mean S-cone spectral sensitivity, Stockman, Sharpe, and Fach [1] measured S-cone spectral sensitivities in three blue cone monochromats known to lack L and M cones on genotypical as well as phenotypical grounds and combined them with normals’ S-cone data obtained at short and middle wavelengths on an intense yellow background field that selectively adapted