Three eras, one gaze. From the analog click to latency, without losing the soul
There is a simple way to tell a long career without turning it into a commemoration. Admit the truth: an image was never just an image. It has always been a pact. A tacit agreement between the one who looks and what is looked at, between reality and its translation, between the event and its trace.
In my case, that pact changed materially three times. And that is exactly where the journey becomes visible. Not geographical, but visual. Across three eras that describe me better than any résumé: the analog, the digital, the generative.
Calling it “evolution” is convenient, but it is often an excuse to avoid the details. I prefer to call it a change of substance. Because what truly changes is not aesthetics, nor speed, nor resolution. It is the very nature of what we call an image.
The analog: light that costs
Analog photography is philosophically shameless. It stages a world pressing on matter. Light passes through a lens, strikes an emulsion of sensitive crystals, and produces an invisible imprint, a latent image, like a secret written in a microscopic language. Development makes it readable. Fixing saves it from self-destruction. What remains is a negative, an object you can lose, scratch, keep, or pass on. Material proof that something happened in front of a machine at a certain time.
That materiality is more than nostalgia. It is psychology. Analog teaches consequence. Every frame is a decision that costs, not so much in money, but in attention. The lack of immediacy forces you to imagine before you see. And imagining before seeing is a severe school. The error cannot be corrected with a command. It stays there, looking back at you, shaping you.
The darkroom is a metaphor almost too perfect: the image emerges from darkness, but only if you accept that darkness exists and plays a role.
The digital: reward and risk
With digital, the substance changes. Light becomes an electrical signal, then a number. The sensor reads it, a pipeline converts it, and what you get is raw data, an incomplete grammar that needs development, only now it no longer smells of chemistry. It stains the mind instead, because it tempts you to believe it is neutral.
This is the second psychological revolution. Digital gives you an immediate reward. You see right away. And that is exactly where the risk is born: seeing right away can become the habit of looking less.
Digital does not force discipline on you. It offers it as a choice. Used well, it is a formidable amplifier: coherence, control, repeatability, quality. It's used badly; it is a slot machine with a beautiful interface. The point, as always, is not the technology. It is the attention you grant it.
The generative: plausibility without a trace
Up to this point, analog and digital share one decisive fact: the image is born from a causal contact with the world. Even when manipulated or transformed, at its origin, there was light hitting a support. That is why photography has always carried particular weight in the realm of credibility. Not a guarantee of truth. Something else: a regime of trust. The presence of a trace.
In the generative realm, that trace is no longer necessary. An image can come from a sentence, an instruction, or a statistical model that has learned correlations and forms. The typical process starts from noise, reduces it step by step, and decodes it into pixels. What you see looks like a photograph. But its appearance is not its origin.
Here, irony helps because we are facing an elegant paradox. For decades, we thought photography was a way to stop time. Now we produce images that look like time stopped, without that time ever having happened. It is an ontological jump, not an update.
A semiotics kit for the confused
With Saussure, analog, digital, and generative do not merely change style. They change the signifier, the substance of the sign. Chemical grain is a signifier. The pixel is a signifier. Generative synthesis, with its plausible textures and learned coherences, is a signifier. And when the signifier changes, it is never only aesthetics. It changes the relationship between what we believe and what we can demonstrate.
If we widen the lens with Peirce, the matter becomes even clearer. Photography has always had a strong indexical component, not by moral merit, but because there is a physical link between world and support. The generative is born, instead, as icon and symbol. It resembles, and it does so well. It becomes indexical only when anchored to something external: a reference, a chain of provenance, a verifiable context. Otherwise, it is an image that speaks a language without paying the price of having happened.
This is also where Barthes becomes useful without turning him into a saint. Photography, as it has been, as the stubbornness of the referent, is not superstition. It is the description of a perceptual and cultural experience. In the generative realm, that feeling can be simulated. And often it is. So the problem is not creativity. The problem is the pact.
What changes when everything becomes plausible
The philosophical and psychological implications are not a luxury here. They are inevitable.
The human brain is a pattern-seeking machine. It does not like uncertainty. When it sees something plausible, it tends to believe it. The generative produces plausibility on an industrial scale. This means attention becomes an even more political resource, in the sense of the polis, the community. Whoever controls plausibility controls a portion of the common narrative.
Ethics, in this landscape, is not “for” or “against.” It is coherence between the medium and the declaration. If you change the regime of the signifier, you must change the way you present it. Not to play the moralist, but to avoid becoming a magician. Transparency does not mean ruining the magic. It means deciding when magic is legitimate and when it becomes a scam.
Three materials, one posture
The journey through three eras is not a technical whim. It is the biography of a gaze.
In analog, I learned that an image is a material responsibility. In digital, I learned that speed is not a value; it is a temptation. In the generative, I am learning, and the sentence must be said calmly, so that plausibility can replace reality if we do not impose rules of reading.
The risk is not that images become “false.” Images have always been manipulable. The risk is that they become indifferent to the distinction between trace and invention. And that we get used to living inside that indifference, because it is comfortable.
Tools do not define an author, but by the kind of pact he demands with the viewer. I have crossed three materials of the visible: chemistry, number, and latency. What distinguishes me is not that I changed means. It is what I tried, each time, not to change: the posture. To look with attention. To translate with respect. And to use irony as a gentle blade: it cuts illusions, not people.
If the future brings us ever more persuasive images, the real skill will not be producing them. It will be knowing what they are, where they come from, and what pact they ask for.
In the end, it is the old question of literature. Now we do not write the sentence with ink. We write it with light, with numbers, and with latency.
Three reference portraits, three panels, one single image. Nano Banana Prompt
I used the following three reference images for the three-panel single image. So the three images, three heads in one single generation.
Subject: photorealistic editorial triptych in one single image, three vertical panels side-by-side (left, center, right), same male body identity across panels but different age in each panel. Panel 1 uses uploaded Image 1 as face reference (age 20). Panel 2 uses uploaded Image 2 as face reference (age 40). Panel 3 uses uploaded Image 3 as face reference (age 60): controlled smile, quiet author presence, no text.
Global composition; horizontal 16:9 canvas containing the three vertical panels, equal panel widths, perfectly aligned top and bottom edges, same subject scale and base pose in all panels, centered symmetry, consistent eye line across panels, head upright and centered in each panel.
Action per panel;
Left panel (Analog, age 20, Image 1 reference); full frontal, straight-on. I hold an analog rangefinder camera centered at chest level, and a 35mm film roll pinched between thumb and index on the other hand, hands open and clearly readable, precise ritual gesture, objects fully visible and sharply defined.
Center panel (Digital, age 40, Image 2 reference); full frontal, straight-on. I hold a professional mirrorless camera, centered at chest level, with the rear display on and an unreadable histogram, fingers on the joystick and controls, on-set posture, smiling toward the viewer/camera (not looking away), gear clearly visible.
Right panel (Generative, age 60, Image 3 reference); full frontal, straight-on. I “write” in the air with two fingers. In front of me, a molecular particle fog (nano-micro particles with subtle lattice behavior: clusters, chains, gentle turbulence) organizes into an emerging image, a soft photorealistic resemblance of a Chinese woman forming inside the particle field, elegant and conceptual, no gaming glow, no hologram UI.
Camera (lens, framing, angle, distance); 40mm, 16:9, eye-level camera, medium distance, framing from mid-thigh to above the head in each panel, full frontal straight-on in all panels, shoulders parallel to the camera plane, gaze controlled and coherent (Panel 1 and 3 gaze to lens, Panel 2 smile to lens even if eyes slightly acknowledge the display), identical scale and neutral base pose across panels, cinematic depth of field, micro-detail on hands and objects.
Light per panel;
Left panel: darkroom red light ambiance, soft lateral key with minimal fill, velvety shadows, controlled contrast, natural skin texture, no horror look.
Center panel: high-end neutral studio lighting, large frontal softbox as key, subtle fill, clean reflections on metal and glass, realistic natural skin texture (no beauty glow).
Right panel: cool, futuristic, volumetric key revealing molecular particles; thin rim light outlining face and jacket edges; controlled contrast; realistic shadows; natural skin texture (no beauty lighting).
Mood: anthropological, lucid, subtle irony, quiet authority, “evolution without nostalgia.”
Environment per panel.
Left panel: dark workbench with lab-like texture, blurred trays and tools in the background, realistic darkroom atmosphere, no readable labels, no UI, no signage.
Center panel: clean, contemporary set; out-of-focus monitor and neatly managed cables in the background; premium production vibe; no readable UI; no signage; no text.
Right panel: minimal, almost-empty space with clear geometric shapes in the background (translucent planes, sharp prisms, clean wireframe-like volumes), physically plausible, slightly out of focus, localized particle activity only in front of me, no readable UI, no signage, no text.
Outfit per panel;
Left panel: military green reporter jacket over a dark turtleneck or neutral t-shirt, discreet watch, darkroom vibe.
Center panel: refined hip-hop t-shirt (clean cut, minimal graphic, no readable logos), street-elegant but believable on a professional set.
Right panel: modern kung-fu-inspired black Armani-style jacket, matte technical fabric, mandarin collar, minimal asymmetrical frog-closure elements (no logos), high-end street-lux, not costume, not “old man coat.”
Constraints (realism, continuity, anatomy, aspect ratio, do-not-crop); photorealistic, perfect anatomy (credible hands, 5 fingers), sharp details, 16:9, do not crop head, hands, elbows, feet in any panel, no extra limbs, no glitches, no watermarks, no logos, no text anywhere. Panel-specific identity locks:
Panel 1: match reference face identity, match head geometry exactly, do not stylize, no beautify, age-down naturally to 20 only.
Panel 2: match reference face identity, match head geometry exactly, do not stylize, no beautify, age-adjust naturally to 40 only.
Panel 3: match reference face identity, match head geometry exactly, match reference smile exactly, do not stylize, no beautify, age-adjust naturally to 60 only.
Maintain consistent body proportions and posture across all three panels, with only age, outfit, props, lighting, and environment changing.
Negatives; no side profile, no 3/4 view, no head turn, no expression drift (especially Panel 3 smile must match reference), no face drift, no beautification, no smoothing, no plastic skin, no stylization, no extra fingers, no cropped head/hands/feet, no text, no logos, no readable UI, no neon gaming look, no hologram interface.
BONUS: Three panels for each age, a single image. Nano Banana Prompt
This time, I used only text and no reference image, and imagined the character getting older in each image.
Subject: photorealistic editorial triptych in one single image, three vertical panels side-by-side (left, center, right), same man across panels, different age in each panel (20, 40, 60). Controlled smile, quiet author presence, no text. The man is an Asian male with a consistent identity and facial structure across all panels.
Global composition; horizontal 16:9 canvas containing the three vertical panels, equal panel widths, perfectly aligned top and bottom edges, same subject scale and base pose in all panels, centered symmetry, consistent eye line across panels, head upright and centered in each panel.
Identity continuity (no reference images); all three panels must depict the same person as he ages. Lock these traits and keep them identical across panels (only natural aging changes):
- Same skull shape and head geometry, same facial proportions, same jawline, cheekbones, nose shape, lips shape, chin shape, ear shape and placement, same hairline placement and overall hairstyle silhouette, same eye shape and spacing, same eyebrow shape.
- Same height, body proportions, shoulder width, neck length, and hand proportions across panels.
- Same expression baseline and smile style (controlled smile), only subtle age-related changes in skin and muscle tone.
Aging rules; Panel 1 looks 20 (smooth skin, youthful firmness). Panel 2 looks 40 (mild nasolabial lines, slight under-eye texture, subtle forehead lines). Panel 3 looks 60 (clear but natural wrinkles, mild sagging consistent with age, realistic skin texture), without changing the person’s core facial structure. Do not change ethnicity. Do not beautify. Do not stylize.
Action per panel;
Left panel (Analog, age 20); full frontal straight-on. I hold an analog rangefinder camera centered at chest level, and a 35mm film roll pinched between thumb and index on the other hand, hands open and clearly readable, precise ritual gesture, objects fully visible and sharply defined.
Center panel (Digital, age 40); full frontal straight-on. I hold a professional mirrorless camera, centered at chest level, with the rear display on and an unreadable histogram, fingers on the joystick and controls, on-set posture, smiling toward the viewer/camera (not looking away), gear clearly visible.
Right panel (Generative, age 60); full frontal straight-on. I “write” in the air with two fingers. In front of me, a molecular particle fog (nano-micro particles with subtle lattice behavior: clusters, chains, gentle turbulence) organizes into an emerging image, a soft photorealistic resemblance of a Chinese woman forming inside the particle field, elegant and conceptual, no gaming glow, no hologram UI.
Camera (lens, framing, angle, distance); 40mm, 16:9, eye-level camera, medium distance, framing from mid-thigh to above the head in each panel, full frontal straight-on in all panels, shoulders parallel to the camera plane, gaze controlled and coherent (Panel 1 and 3 gaze to lens, Panel 2 smile to lens even if eyes slightly acknowledge the display), identical scale and neutral base pose across panels, cinematic depth of field, micro-detail on hands and objects.
Light per panel;
Left panel: darkroom red light ambiance, soft lateral key with minimal fill, velvety shadows, controlled contrast, natural skin texture, no horror look.
Center panel: high-end neutral studio lighting, large frontal softbox as key, subtle fill, clean reflections on metal and glass, realistic natural skin texture (no beauty glow).
Right panel: cool, futuristic, volumetric key revealing molecular particles; thin rim light outlining face and jacket edges; controlled contrast; realistic shadows; natural skin texture (no beauty lighting).
Mood: anthropological, lucid, subtle irony, quiet authority, “evolution without nostalgia.”
Environment per panel.
Left panel: dark workbench with lab-like texture, blurred trays and tools in the background, realistic darkroom atmosphere, no readable labels, no UI, no signage.
Center panel: clean, contemporary set; out-of-focus monitor; neatly managed cables in the background; premium production vibe; no readable UI; no signage; no text.
Right panel: minimal, almost-empty space with clear geometric shapes in the background (translucent planes, sharp prisms, clean wireframe-like volumes), physically plausible, slightly out of focus, localized particle activity only in front of me, no readable UI, no signage, no text.
Outfit per panel;
Left panel: military green reporter jacket over a dark turtleneck or neutral t-shirt, discreet watch, darkroom vibe.
Center panel: refined hip-hop t-shirt (clean cut, minimal graphic, no readable logos), street-elegant but believable on a professional set.
Right panel: modern kung-fu-inspired black Armani-style jacket, matte technical fabric, mandarin collar, minimal asymmetrical frog-closure elements (no logos), high-end street-lux, not costume, not “old man coat.”
Constraints (realism, continuity, anatomy, aspect ratio, do-not-crop); photorealistic, perfect anatomy (credible hands, 5 fingers), sharp details, 16:9, do not crop head, hands, elbows, feet in any panel, no extra limbs, no glitches, no watermarks, no logos, no text anywhere. Maintain consistent body proportions and posture across all three panels, with only age, outfit, props, lighting, and environment changing—no readable UI. No stylization. No beauty.
Negatives; no side profile, no 3/4 view, no head turn, no expression drift, no face drift between panels, no identity change, no different skull shape, no different jawline, no different nose, no different eye shape, no beautification, no smoothing, no plastic skin, no stylization, no extra fingers, no cropped head/hands/feet, no text, no logos, no readable UI, no neon gaming look, no hologram interface.