[EXPERT: CONSTRUCTED EYE] Day 3 — Psychophysics: Measuring Experiences That Cannot Be Observed Directly

Day 3 – Psychophysics: Measuring Experiences That Cannot Be Observed Directly

The Constructed Eye: Visual Illusion, Perception Science, and the Work of Akiyoshi Kitaoka and Beau Lotto

Strips with controlled luminance contrast, as could be used in classical threshold experiments (Hecht et al., 1942). Such stimuli are fundamental to measuring perception absent any direct access to neural responses.

Expert Objective

The aim today is to master how vision scientists measure subjective sensory experiences—contrast, brightness, motion, color—using psychophysical methods, and to critically examine how works like Akiyoshi Kitaoka’s illusions and Beau Lotto’s colour studies draw from, and sometimes complicate, these quantitative techniques. We will explicitly query: How do we reliably quantify something we cannot observe directly?

Observed Effects

• Luminosity Thresholds: Classical experiments (Hecht et al., 1942; Watson & Pelli, 1983) show that observers report detection of a stimulus at well-defined, quantifiable physical intensities—even when neural events remain unobservable without imaging.
• Illusion Strength: Kitaoka’s moving snakes and Lotto’s adaptive colour contexts are measured in terms of response times and subjective matches (Lotto & Purves, 2002; Kitaoka, 2021).

Hypothetical matches in psychophysical brightness scaling: observers rate these spots as "equally bright" despite different spectral compositions. Such matches reveal the constructiveness of perceptual experience (Stevens, 1961).

Supported Mechanisms

• Signal Detection Theory (SDT) quantifies the probability distributions of sensory events relative to observer criteria, giving psychophysics mathematical power (Green & Swets, 1966).
• Adaptive Staircase Procedures (Levitt, 1971) allow efficient threshold estimation; e.g., adjusting brightness until an observer can barely detect a difference.

For illusions: mechanisms implicate spatial filtering, lateral inhibition, and hierarchical prediction—not merely “filling in gaps" but specific receptive field and contextual computations, as demonstrated in primary visual cortex (Carandini et al., 2005).

Typical adaptive staircase data: stimulus intensity approaches the observer’s detection threshold. Artists and researchers alike can adopt these methods to calibrate perception in their own visual experiments.

Evidence and Competing Explanations

• Classical psychophysics: is robust for simple features (luminance, orientation, localization). However, even with meticulous controls, observers’ reports reflect not only retinal sensitivity but also cognitive criteria—e.g., decision biases or learning effects (Kingdom & Prins, 2016).
• Competing Theories: Bayesian perception posits that experience reflects statistically optimal integration of noisy input and prior expectation (Knill & Pouget, 2004), but some illusions (e.g., Kitaoka’s anomalous motion) reveal limits of such optimality.
• Unresolved Questions: How does introspective report relate to neural-level thresholds? For artists, the crucial question remains: can we trust cross-observer equivalence in perception indices, or do personal priors and local adaptation undermine standardization?

Digital Experiment

Design: Use a stable display and lighting environment. Present a sequence of gray patches against a neutral background, gradually increasing contrast in steps of 2%. Record the lowest contrast at which you reliably detect the patch (5/6 successes at a position = threshold).

Controlled Variables: Luminance and contrast, patch position, viewing distance.

Observation Protocol: Spend at least 3 seconds with eyes closed between trials to reduce adaptation. Repeat with background hues to test for context effects.

Limitation: This test estimates your behavioral threshold under specific contextual setups; it does not directly trace retinal or cortical spiking activity.

Retrieval Question

Question: How do adaptive psychophysical procedures clarify the difference between neural sensitivity and perceptual decision criteria? Give an example involving context-dependent illusions.

Further Discussion

Studio insights from recent illusions by Kitaoka and digital context manipulations by Lotto illustrate that while psychophysics provides a rigorous toolkit, artists routinely push perceptual boundaries beyond what threshold measurements alone reveal. Understanding these metrics allows advanced practitioners to design more potent, testable visual experiences.

Sources

• Hecht, S., Shlaer, S., & Pirenne, M. H. (1942). Energy, Quanta, and Vision. J Gen Physiol. Read
• Watson, A. B., & Pelli, D. G. (1983). Quest: A Bayesian adaptive psychometric method. Perception & Psychophysics.
• Green, D. M., & Swets, J. A. (1966). Signal Detection Theory and Psychophysics. Wiley.
• Levitt, H. (1971). Transformed up-down methods in psychoacoustics. J. Acoust. Soc. Am.
• Kingdom, F. A. A. & Prins, N. (2016). Psychophysics: A Practical Introduction. Academic Press.
• Kitaoka, A. (2021). "Moving snake" illusions: Mechanisms and models. Lab page
• Lotto, R. B., & Purves, D. (2002). The Empirical Basis of Color Perception. Consciousness and Cognition.
• Carandini, M., Heeger, D. J., & Movshon, J. A. (2005). Linearity and normalization in simple cells of the macaque primary visual cortex. J. Neurosci.
• Stevens, S. S. (1961). To Honor Fechner and Repeal His Law: Science.
• Knill, D. C. & Pouget, A. (2004). The Bayesian brain: the role of uncertainty in neural coding and computation. Trends in Neurosciences.