Thomy H. Nilsson

Flicker adaptation shows evidence of many visual channels selectively sensitive to temporal frequency

Temporally specific visual channels were proposed by Levinson (1959) to account for complex effects on flicker detection when two frequencies of intermittency were alternated. The existence of visual channels which are specifically sensitive to spatial frequency has been demonstrated in humans by measurements of sensitivity to luminance modulation at various spatial frequencies after adaptation to a given spatial frequency (Blakemore and Campbell, 1969; Graham, 1972). These measurements show that the decrease in sensitivity resulting from adaptation is largely specific to the spatial frequency at which the observer adapted. Using a similar procedure but employing luminance modulation at various temporal frequencies, Smith (1970) discovered a temporally specific adaptation effect. A subsequent study by Smith (1971) found evidence of temporal specificity at more than one frequency, but the use of only three widely spaced adapting frequencies gave little indication of how temporal specificity varied as a function of frequency. A similar study by Pantle (1971) employed more frequencies but obtained less definitive results-possibly due to the use of square-wave stimuli which introduces a broad range of frequencies. Evidence by Smith (1971) indicated that the effect was restricted to the cone system and he mentions that transfer of adaptation occurred between a red and a blue stimulus. Since all evidence of more than one temporally selective channel was obtained with full color-spectrum stimuli in both the Smith and Pantle studies, the possibility remained that temporal specificity at more than one frequency represented changing contributions from color pathways of different temporal responsivity (Krauskopf and Mellon, 1971). Inherent in the concept “temporal adaptation” are two distinct system properties, short term frequency sensitivity and a longer term adaptation effect. Since these coexist in the time domain, a transitory interaction such as the drifts in flicker sensitivity reported by Smith (1970) and Akos and Akos (1971) might be

Two-Pulse-Interval Vision Thresholds*

Data on visual discrimination of two-pulse-interval differences are presented. Experiment I measured difference thresholds for pulse intervals ranging from 0 to 75 msec; luminance being varied from 50 to 2000 mL. Luminance had no significant effect on these temporal discriminations. Difference thresholds decreased linearly as pulse interval increased. Experiment II measured increment and decrement difference thresholds separately, for pulse intervals ranging from 0 to 90 msec. Increment thresholds decreased as pulse interval increased from 0 to 30 msec, but increased as pulse interval was increased beyond 30 msec. Although the shape of the decrement-threshold function was not defined, increment- and decrement-threshold functions differed markedly. These discrimination functions are thought to result from the temporal-response characteristics of definite visual processes.

Hue shifts produced by intermittent stimulation.

Abstract A matching technique was used to measure the shifts in hue produced in ten narrow-band spectra between 425 and 650 nm by eight rates of intermittency ranging from 0 to 15 Hz. The direction of hue shift did not vary with intermittency rate, but magnitude of hue shift did. The intermittency rate producing maximum shift varied with wavelength. Stimuli at 425, 500 and 574–600 nm, were relatively invariant at all rates. Maximal hue shifts were observed with stimuli at 525–550 and 650 nm towards longer wavelengths. Direction of shift of red targets may, however, depend on observer criteria, since strong desaturation effects accompanied the hue shifts. The data indicate that the hue shifts produced by intermittent stimulation differ from those produced by luminance changes. It is suggested that intermittency hue shifts involve an interaction between stimulus intermittency and a temporal coding of colour in the visual system.

Effects of pulse duration and pulse rate on hue of monochromatic stimuli

Abstract The effects of pulse durations from 10 to 1000 msec and pulse rates from 1 to 37 Hz on hue were measured for stimulus wavelengths of 425–650 run by a stimulus matching technique at illuminance levels of photopic threshold plus 2 and plus 4 log. The results show that varying pulse duration and pulse rate produces hue changes which vary in magnitude and direction with wavelength and illuminance level. These temporally induced effects appear to represent two dichotomous hue shift processes on an illuminance dependent continuum. At high illuminance levels, hue shifts are similar to the Bezold-Brucke effect; while at low illuminances a nearly contrary hue shift phenomenon is encountered. An explanation suggests how these effects arise from an interaction of spectral and temporal response characteristics of receptor pathways.

Visual temporal discriminations of brief pulse intervals

Abstract Determination of the transient response of a system as complex as vision when the sensory characteristics of that response elude definition may be accomplished by measuring the difference threshold for increases in the interval between double pulses at various pulse intervals. Such temporal difference thresholds were measured at pulse intervals from 3 to 140 msec with stimuli ranging in luminance from 1.5 to 3.5 log above foveal threshold. These thresholds were a general U-shaped function of interval, a minimum detectable difference in pulse interval occurred when the pulse interval was about 80 msec: increasing luminance decreased the interval at which this minimum difference threshold occurred. The fine temporal resolution with which these thresholds were measured revealed additional complexities in the threshold function, notably a secondary fall and rise in threshold occurring at intervals from 20 to 50 msec. These results are interpreted as an indication that the response to a brief input is a complex pattern of activity perhaps reflecting the time constants of various characteristics of visual processing whose composite effect may still be describable as an impulse response function.

Randomization without replacement using replacement without losing your place

Lehman (1977) describes a procedure for sampling without replacement from an array of N variables that does not require reshuffling the entire array N times. I have been using a different sampling-withoutreplacement procedure that avoids such reshuffling and also avoids knowing the number of samples remaining. The procedure may be more efficient for on-line applications, since it does not wait upon a randomization routine each time a value is requested. The procedure also facilitates storing data nonrandomly and handling replications. Start with a one-dimensional array, XI-N, of the various values to be selected. (If R replications are wanted, repeat each value R times in the array and set N = N X R.) For 1= I-N, select a random number between 1 and N then interchange the value of the Ith element with the value of the element in the position of the random number. The result is an array of the original values replaced in random order. With this procedure, one randomizes the array beforehand and then takes values from the array in succession. The randomized values are immediately available. The procedure can be written in a single program line using either FOCAlrRT (Siegel, 1972):

IMPROVING PERCEPTION OF LETTERS AND VISUAL STRUCTURE OF LANGUAGE

Information about letters and the physical structure of language printed in Roman characters was given to children beginning to read. Experimental investigations coupled three alternative graphic modes of printing upper- and lower-case letters with an instructional intervention termed “Alpha-Beta” which provides practice in letter sorting, matching of letters, associative matching, and memory matching. In respect to graphics, Mode A letters were in standard alphabet form. Mode B provided standard letters with each backed by a unique half-tone (Visually Stippled Alphabet); Mode C provided standard letters with each backed by a unique visual texture (Visually Patterned Alphabet). Pre-posttest change in reading readiness was measured using the Metropolitan Readiness Test. In the first study 224 English-speaking 5- to 6-yr.-old children were tested. In the second there were 158 Spanish-speaking girls and boys 6 to 7 years old. It was predicted that Alpha-Beta intervention involving visually patterned alphabet would lead to the greatest increases in readiness scores. This is confirmed in both studies for children low in reading readiness preexperiment. Children high in reading readiness are less affected. The second experiment involved Spanish-speaking children and investigated intervention by Alpha-Beta against a no-intervention control. This confirms the value of Alpha-Beta per se. Possible explanations for the improvements are identified.

Reference light source for luminance and illuminance photometry

The rationale for a constant source of both luminance and illuminance to check photometer calibration is discussed. The inexpensive construction of such a reference source is described, together with some performance data.

Transistorized timer for sensory-behavioral research

A timer circuit is presented that can produce pulse durations ranging from 1 to 20,000 msec. The timer has high reliability, is inexpensive, and simple to construct.

Observer positioning device with simple, three-axis adjustment of rotation and translation

A rod clamped in a rotatable, vertical vise forms the basis of a sturdy, inexpensive, and convenient bite-board holder for vision research.