We see things not as they are, but as we are.” — Anaïs Nin
Due to the nature of my work, I am often asked, “How do you get your paintings to look so realistic?” I genuinely love when a colleague or aspiring representationalist opens this door in a conversation, as it usually leads to great exchanges about art, visual perception, and a host of related topics. When people ask about making work look realistic, they often instinctively associate it with a direct accordance with reality, assuming it means adding more detail, increasing resolution, or refining rendering. While these factors can contribute to a realistic representation, the most crucial accordance is not merely with physical reality but with how a viewer’s perceptual experience of the representation aligns with their past experiences of reality.
To clarify, as I stated in my 2019 paper What Does Realistic Look Like?, “realistic” for the representational artist is often best understood in terms of what people actually mean when they describe something as realistic:
“Realistic is the degree of relative similarity between a perceptual response to a surrogate, simulation, or other representation and the past perceptual responses to the stimulus, stimulus components, or experience that is being represented.”
This view of “realistic” not only accords with what is the case, but it can have a significant impact on how representationalism is approached. For me, it has led to extensive consideration of what I call the A1 Problem. My focus on this issue has profoundly influenced my assumptions and, in turn, the decisions and strategies I employ in my painting process.
“The A1 Problem is the challenge of constructing a representation (R) from the artist’s perceptual experience of a subject (A1)—the artist’s only visual access to the external subject (A)—so that the viewer’s response to the representation (A2) closely resembles that viewer’s past perceptual responses to the subject (A), its components, or the experience being represented.“
Here’s how it works: An artist cannot access objective reality directly through the visual system. When interacting with a subject (A), the artist experiences the internal perceptual response generated by their visual system (A1). It is therefore A1—not A itself—that guides the representation (R) that the artist creates.
When a viewer encounters R, they do not gain direct access to either A or the artist’s A1. Instead, the viewer’s visual system generates a new perceptual response to the representation: A2. This response to R is shaped by evolved visual architecture, prior experience, expectations, attention, and context.
Because perception is inherently non-veridical—not a direct measurement or objective mirror of physical reality—achieving a stronger perceived realism is not simply a matter of copying the artist’s A1 as literally as possible. The artist must construct R so that A2 (the viewer’s response to R) corresponds as closely as possible to the viewer’s own past perceptual experiences of A.
The non-veridical nature of human visual perception makes direct replication of an object’s “true” properties via that system impossible. As George Berkeley (1709) pointed out several centuries ago, sources underlying visual stimuli are unknowable in any direct sense—a fact repeatedly confirmed by modern research. The retinal projection of an object does not contain explicit information about size, distance, or orientation, meaning that vision must “construct” these properties using evolved responses rather than directly capture reality. Interestingly enough, in a 2013 review, Frédéric Perdreau and Patrick Cavanagh examined experimental evidence concerning whether visual artists possess enhanced access to low-level retinal information or “more veridical” perception than non-artists. The study found that artists do not have special perceptual expertise that allows them direct access to their retinal image or any learned strategies to recover that image. Instead, the advantage of artists may arise from more robust representations of object structure in memory, which are more appropriate to the visuomotor demands of drawing.
Artists working within the framework influenced by the A1 Problem should recognize that making something ‘realistic’ is not about mechanically reproducing visual input, but is rather about creating a surrogate that strategically aims to maximize consistent correspondence across viewers’ perceptual responses. The artist is not merely painting A1 (their own perception of the subject) but constructing a representation that evokes a percept (A2) intended to align with the collective past perceptual experiences associated with A (the subject). Achieving this requires an understanding of perceptual mediation and the cognitive mechanisms that shape visual experience.
Why Not Just Paint What You See?
There has long been extensive discussion about whether it is better to paint what you see or what you know. Both approaches can offer significant advantages as well as a few problems for the observational representationalist. The mindset I have found most advantageous—one that dovetails with the A1 Problem—is to paint what you wish viewers to experience.
On the surface, one might argue that an artist can bypass concerns about perceptual mediation simply by ensuring that a painting “looks the same” as the subject to the artist’s own eyes. However, this overlooks several important features of perception.
- Individual Perception Is Not Objective: As we have already established, visual perception is not a direct reflection or measurement of external reality. It is an internal construction shaped by many interacting factors, including adaptation, color constancy, opponent processing, attention, context, prior experience, and evolved visual architecture. Even when an artist believes they are faithfully matching what they see, their visual system is already doing much of the work involved in constructing the experience that guides them.
- Viewers Will Not See the Painting Exactly as the Artist Does: A viewer’s perceptual experience of an artwork—A2—is generated by that viewer’s own visual system. It is shaped by individual learning and experience, species-typical visual architecture, expectations, attention, and viewing context. If an artist relies exclusively on copying isolated components of their own A1, the resulting A2 may not align as closely with the viewer’s previous perceptual experiences of A as it would if the representation were strategically constructed to recreate the relevant perceptual conditions. The goal is therefore not necessarily to maximize physical correspondence between the subject and the representation, nor to reproduce every component of the artist’s A1. It is to construct R so that the viewer’s response to it—A2—corresponds as closely as possible with the viewer’s accumulated perceptual experience of A.
This approach does not prioritize a direct visual match between the subject and its representation but instead optimizes the potential alignment between A1 and A2, ensuring the viewer’s perceptual response more closely reflects the artist’s intended experience of A.
Experiencing the A1 Problem
A particularly vivid way to experience the A1 Problem is through Hinnerk Schulz-Hildebrandt’s fixation-dependent “Purple Dots” illusion. The image contains nine physically identical purple dots on a bluish background. When a viewer looks directly at one dot, it appears more ‘reddish’ than the others, while the surrounding dots shift toward blue. As the viewer moves their gaze, the more reddish-purple appearance follows the point of fixation. The physical colors remain unchanged; what changes is the perceptual response to them.
Why Does the Color Appear to Change?
The exact contribution of each mechanism has not yet been established experimentally, but Schulz-Hildebrandt proposes that the illusion results from several interacting features of human color vision. Purple is a nonspectral color: unlike a spectral green or yellow, it does not correspond to one continuous band of visible wavelengths. Its appearance is constructed largely from a combination of long- and short-wavelength signals—what we might respond to as red and blue—without a correspondingly strong middle-wavelength (green) signal. Because its appearance depends on the balance between these signals, relatively small changes in that balance can shift purple toward either red or blue. That balance also varies across the retina. Short-wavelength-sensitive cones are sparse and nearly absent at the very center of the fovea, while the yellowish macular pigment near the center of vision absorbs more short-wavelength light than regions farther from fixation. These factors can reduce the relative contribution of blue-sensitive signals at the point of direct gaze, allowing the dot’s redder component to become more prominent. Away from fixation, greater relative sensitivity to the short-wavelength component can shift the same physical purple toward blue. Contextual interactions between the dots and the bluish background may further modify that balance.
The illusion presents the representational artist with an interesting problem. Imagine looking directly at each dot in turn and painting it according to the appearance it produces at that moment. Because every fixated dot appears reddish-purple, an artist committed only to copying each immediate appearance might paint every dot that color. Each individual choice could truthfully record one aspect of the artist’s A1. Yet, if one’s process unfolds in this way, something would be lost: the dynamic perceptual relationship among the dots. A viewer encountering the altered representation would no longer experience the shifting distinction between the fixated dot and those seen peripherally. In attempting to reproduce a series of isolated appearances, the artist would have eliminated the conditions that gave rise to the original experience.
To bring the viewer an experience closer to that produced by the original stimulus, the artist would need to reproduce the physically identical bluish-purple dots rather than paint each one with a colorant matching its reddish appearance under direct fixation. This preserves the relationships that allow the viewer’s own visual system to generate the changing colors. The representation may therefore need to depart from the artist’s immediate experience of each isolated component in order to produce an A2 that more closely aligns with the viewer’s experience of A.
Importantly, this does not mean that painting one or more dots redder would be artistically invalid. Such a decision could communicate a genuine perceptual moment, direct attention, reveal something about the artist’s experience, or serve another expressive purpose. That flexibility is one of the great strengths of visual art. However, under the present definition of realism, the altered version would be less realistic insofar as the viewer’s response no longer reproduces the distinctive perceptual dynamics encountered in the original stimulus.
The Software-Hardware Metaphor: Art as Neural Programming
To conceptualize this further, we can use a software–hardware analogy. An artwork functions somewhat like a structured set of instructions: its colors, shapes, edges, spatial relationships, and other physical features set patterns of neural processing in motion within the observer. Those features interact with the observer’s visual system, prior experience, beliefs, expectations, attention, and the context in which the work is encountered. Together, these factors contribute to the perceptual and cognitive experience we describe as an encounter with art.
The analogy is somewhat imperfect, as an artwork does not contain a fixed program that executes identically in every observer, nor is the brain merely static hardware. Neural systems are adaptive, state-dependent, and continually shaped by experience. Nevertheless, the metaphor highlights an important point: the artwork is an objective physical entity, but the experience it produces is constructed anew within each observer. As I have discussed elsewhere, as viewers differ, an artist cannot predict with absolute certainty how a representation will be experienced by every individual. However, artists can draw on an understanding of perceptual processes and broadly shared features of human vision to increase the likelihood of consistent responses and to deliberately influence how realistic a representation appears.
In this context, success is not simply a matter of physically duplicating the subject. It involves constructing the representation (R) so that the viewer’s perceptual response (A2) aligns as closely as possible with that viewer’s past perceptual experiences of the subject (A). The greater that alignment, the more realistic the representation is likely to appear.
Let’s Review
Understanding the A1 Problem highlights the importance of visual-perception research for artists seeking a “high degree of realism”. It carries several important implications.
Awareness of Perceptual Mechanisms Can Improve Realistic-Representation Strategies
Simply copying what appears to be “seen” does not necessarily produce an image that corresponds with the perceptual experience an artist intends to evoke. Visual experience is shaped by mechanisms such as adaptation, color constancy, contrast, opponent processing, depth cues, attention, and contextual expectations. An awareness of these processes can help artists construct representations that more effectively engage the visual system and better align with viewers’ accumulated perceptual experiences of the represented subject.
Strategic Representation Can Enhance Perceived Realism
Artists regularly amplify, adjust, simplify, or suppress visual information in order to shape the viewer’s experience. These choices are not necessarily departures from realism. In some cases, they may be required to preserve it across differences in scale, medium, illumination, contrast range, or viewing conditions. Without an understanding of how perception contributes to the experience of a representation, artists may overlook opportunities to produce a more convincing result.
One useful strategy is to engage the viewer as a more active participant in the representational process. Rather than specifying every feature completely, an artist can provide information that allows the viewer’s visual system to organize, complete, and interpret the image to a “greater” degree. Lost-and-found edges, for example, can encourage perceptual completion when the missing boundary is consistent with plausible conditions such as low contrast, shadow, glare, atmospheric interference, or occlusion. Under those conditions, the viewer’s own visual processing helps construct the represented form. The result may feel more realistic because it more closely resembles how boundaries are often experienced in the visual world.
Viewer Experience Should Be Considered
Different viewers may bring different perceptual histories, expectations, abilities, and cultural experiences to a representation. Their prior experiences of A will therefore never be completely identical. Nevertheless, viewers also share many features of human visual architecture and extensive experience with recurring properties of the physical world. Artists can draw upon these shared regularities to increase the likelihood that a representation will produce broadly similar responses within an intended audience.
The goal is not to eliminate individual variation (which is impossible) but to construct R so that the resulting A2 remains compatible with perceptual expectations likely to be shared by many viewers.
Medium and Context Influence Perception
Every representational medium—whether painting, photography, print, or digital display—introduces its own constraints and transformations. Differences in luminance, color range, scale, surface, resolution, illumination, and viewing distance can all affect the experience an image produces. The human visual system remains the common point of contact across these media, but it does not respond to each medium under identical conditions. Artists must therefore consider both broadly shared features of vision and the particular limitations of the medium in which a representation will be encountered.
By strategically managing elements such as contrast, edges, color relationships, spatial cues, and composition, artists can compensate for these limitations and increase the likelihood that the viewer’s A2 will align with their prior perceptual experiences of A.
In Conclusion…
A thoughtful consideration of the A1 Problem can help artists recognize, understand, and navigate the perceptual mediation inherent in representational art. Creating highly realistic imagery is not simply a matter of reproducing a subject’s measurable properties as faithfully as possible. It depends upon how effectively a representation elicits a perceptual response that aligns with the viewer’s accumulated experience of the subject.
By exploring fields such as visual perception, cognitive science, and psychology, representational artists can refine their methods more deliberately. Rather than attempting to transfer the artist’s perceptual experience directly into an image, they can construct the representation (R) in ways that increase the likelihood that the viewer’s response (A2) will correspond with that viewer’s prior perceptual experiences of the subject (A).
This understanding moves representation beyond mere imitation. It frames representational art as the deliberate shaping of visual information—an approach that acknowledges the non-veridical nature of vision and uses an understanding of perceptual processes to create imagery that is convincing, engaging, and perceptually resonant.
As always, I hope this article proves useful to those following the challenging path of realistic representation. I wish you all great success!!!