Photographing 2D, 3D, and time-based work to submission standard

The two-light "copy lighting" baseline for flat work

The field-standard setup for 2D work, repeated almost verbatim across CaFÉ (CallForEntry), Envato Tuts+, ExpertPhotography and the US National Archives/Sustainable Heritage Network object-documentation guide, is two identical lights placed at roughly 45° to the plane of the work, equidistant from center, at matched output.

As Envato Tuts+ explains, this is called "copy lighting," "because we are lighting to make as exact a copy of the subject as possible." One light's spread cancels the other's shadows, producing an even wash. The durable verification step — not the gear — is what matters: meter or test the four corners and the center and confirm they read the same.

CaFÉ states the rule negatively: "Using only one light source causes uneven lighting and shadows." A larger piece needs the lights moved farther back to cover it evenly, because a softbox's intensity falls off toward its edges (Derek Au notes a one-meter softbox "will not illuminate a one meter long painting evenly"). Both ExpertPhotography and the US National Archives/Sustainable Heritage Network guide endorse north-facing-window light, bright overcast, or open shade as a no-cost, roughly daylight-balanced alternative when shot-to-shot consistency isn't required — while warning that direct sun causes specular hotspots and open shade casts blue.

The standard text on reflection control, Light — Science & Magic (Hunter, Biver, Fuqua), reframes glare not as a lighting problem but a geometry problem. Light reflects off a polished surface at the same angle it strikes it, so there is a predictable "family of angles" from which a light will bounce a direct reflection into the lens; "any light positioned within the family of angles will produce a direct reflection. A light placed anywhere else will not." The practical consequences:

1. Place lights outside that family (the 45° rule is the everyday expression of this)
2. You can shrink the family by moving the camera farther back and using a longer lens.

This is the single most useful mental model for diagnosing hotspots on varnish, glazes, glossy photos, and metallic media — it tells you where to move the light rather than guessing.

Polarization — the professional's glare eliminator

For varnished paintings, glossy surfaces, and reflective media, cross-polarization is the technique reproduction photographers and conservators rely on. As the Cultural Heritage Science Open Source (CHSOS) guide and Light — Science & Magic both explain: put polarizing sheets (gels) over each light and a polarizing filter on the lens, then rotate the lens filter ~90° to the light polarization to extinguish the polarized direct reflection while diffuse reflection (the actual image of the work) passes through. CHSOS flags two durable cautions. First, light loss is steep: cross-polarized capture (linear film on each light plus a lens filter at 90°) costs "upwards of 4 f-stops" through the combined thin-film polarizers, versus roughly 1.5 stops co-polarized.

Second, cross-polarization can oversaturate or alter color, because you are removing reflected light that is "a natural property of the painting" — so it is not always a faithful match to the naked eye. Use a circular polarizer (not linear) if relying on autofocus.

Raking light — a light placed to one side at a very low, oblique angle so it "glances across" the surface — is the conservation-standard way to reveal impasto, brushwork, weave, and relief. The AIC Conservation Wiki specifies the source "should be a narrow slit or a point in order to provide the high contrast and definition to the shadows"; diffuse light gives weak texture. This is explicitly a supplementary image: Sparkl and others advise a frontal soft-lit shot for color plus a separate raking-light shot for texture, never conflating the two. Mark the light position with tape so texture captures are repeatable (Conservation Wiki).

Double (cross) polarization, not just a lens polarizer, is what fully kills "sparkle" reflections — but verify color against a reference because polarization can shift it.

Highest-leverage insightDiffusion (softbox/umbrella) plus equal distances is what makes the corners match; uneven corners are the most common amateur tell on large paintings.

One light source, one color temperature

The foundational rule, stated by Sparkl and the National Archives guide: choose one light-source type and do not mix color temperatures. Mixing window light with warm bulbs produces color shifts that cannot be globally corrected, because — as the Bob Atkins white-balance reference explains — "if you have different parts of the subject illuminated by different light sources… you can't white balance so that all colors will look correct." Daylight-balanced sources around 5000–5500K are a common neutral choice, but there is "no 'correct' color temperature" in the abstract (Derek Au) — the point is consistency and a reference, not a magic number.

The highest-leverage color control, agreed across Derek Au, Sparkl, X-Rite/Calibrite, and cultural-heritage guidance, is to photograph a neutral gray card and/or a color-reference chart (e.g., an X-Rite/Calibrite ColorChecker) under the same light, in at least one frame, before removing it. A gray card "accounts for the overall tint and temperature" (B&H); a color checker adds per-color references. Why not a sheet of paper? X-Rite: "most objects are not actually neutral under all lighting conditions," so paper introduces its own cast. Many professionals leave white balance on Daylight, shoot raw, include the ColorChecker in the first frame, set white balance off a neutral patch, then apply that setting to the rest of the session (Bob Atkins; Derek Au).

When the target is a small part of the frame, lens defects (purple fringing, chromatic aberration) can bleed into the gray patch and make the correction "wildly and obviously off" (Photographic Reference Scale / smallpond). Light the target the same way as the work, and size it large enough to sample cleanly.

Highest-leverage insightA reference target in the frame is the difference between guessing color in post and measuring it — it is the most-cited single practice that separates credible from amateur color, and it is cheap.

Square the camera to the work

Keystone (trapezoidal) distortion occurs when the sensor plane is not parallel to the work's plane — the same effect seen when a projector is tilted (Wikipedia; B&H). The fix is mechanical: center the lens on the work's midpoint, set the camera so its back is parallel to the work, and confirm with levels. ArtStorefronts and ExpertPhotography both insist on matching lens height to the center of the artwork and using a tripod with levels, "don't eyeball it." CaFÉ names the failure the jurors see: shooting at an angle creates "the 'trapezoid effect,' which will not look good to the juror."

Hang or lay the work flat and level first; vertical orientation with the lights to the sides is often easier to balance (ExpertPhotography).

Wide-angle lenses exaggerate keystone and edge distortion and are "more pronounced when using ultra wide angle lenses" (A Few Pictures; Shanghai Optics). The durable rule: back the camera up and use a longer ("normal" to short-telephoto) focal length so the work fills the frame without a wide setting — Sparkl cites roughly 35–70mm-equivalent on APS-C, 50–105mm on full-frame as distortion-friendly ranges. A longer working distance also shrinks the Light — Science & Magic "family of angles," easing glare control. Envato Tuts+ adds a craft bonus: from farther back, only a shallow camera-tilt adjustment is needed to get perpendicular.

A tripod plus self-timer, cable release, or mirror lock-up eliminates shake; aperture around f/8–f/11 gives adequate depth of field and lens sharpness for flat work (Cultivate Grand Rapids; ArtStorefronts). Autofocus can struggle on low-contrast flat art — focus manually or on a high-contrast edge (Artsy Shark).

Highest-leverage insightDistortion correction in software stretches pixels and crops the file; correcting geometry in capture (square + longer lens) preserves resolution and is the professional default. Tilt-shift ("PC") lenses correct optically before capture but are specialist gear (B&H).

Clean, neutral, non-competing backgrounds

The consensus across CaFÉ, GYST, and the National Archives guide: backgrounds should be neutral, patternless, and non-reflective so nothing competes with the work. For 2D, a white/gray wall or, for works on paper, a seamless sweep; for 3D, a seamless paper sweep (taped to a surface and curved up to vertical) eliminates the distracting horizon line. St. Olaf's handbook recommends a white-to-black gradient sweep for small 3D work and warns against "grass, black velvet, cloth" backdrops. A registrar's nuance from Harriete Estel Berman: the background's shadow direction and gradation should also signal which way is up, so the image reads correctly with no ambiguity.

A flat image strips scale, so it must be re-supplied. Durable options, in rough order of formality:

1. Accurate dimensions in the caption/metadata.

2. An installation view showing the work on a wall or in a room, which gives architectural scale cues.

3. For some contexts, a human figure interacting with the work — GYST and the Minot State juried-show guide specifically advise that for installation, interactive, or kinetic work you "photograph a viewer or audience member interacting with the work" both to convey scale and to show behavior.

Caution: a clean isolated object shot is still usually the primary image; scale-cue shots are typically supplementary.

Highest-leverage insightFor sculpture and installation, an establishing/installation view is the most reliable scale cue and is often the difference between a juror grasping the work's physical presence and misreading a monumental piece as a maquette.

When to include details

Detail/macro shots earn their place when texture, surface, material, or technique carries meaning the full shot can't convey — impasto, glaze, weave, joinery, fine marks, or a "secret" element. Harriete Estel Berman: "If… a secret moving element is a significant aspect of a piece, be sure that this is somehow illustrated photographically." Detail shots supplement, never replace, the full establishing image.

Close-up work collapses depth of field, so the entire detail can't hold focus at a wide aperture. Stopping down helps only to a point: past a sensor-dependent threshold, diffraction begins softening the whole image. On 24–30MP full-frame bodies (pixel pitch ~5.5–6.0µm) diffraction is noticeable from about f/11–f/13 and clearly visible at 100% by f/16; on 24–26MP APS-C sensors (~3.8–4.0µm) it shows from about f/8–f/10. The professional solution is focus stacking — shooting a series at a lens's sharp "sweet spot" (often f/5.6–f/8) with the focus point stepped through the subject, then blending in software for front-to-back sharpness (Tamron; ProGrade; Digital Camera World). Requirements: a tripod, manual focus, low ISO, no subject movement, and turning off image stabilization when tripod-mounted.

Use raking light to emphasize texture; use frontal soft light when the detail is about color or fine line. Reflective or glossy details may need the polarization controls described earlier.

Highest-leverage insightFor high-relief or deep 3D detail, focus stacking is what lets you use a lens's sharpest aperture and hold full sharpness — the combination amateurs miss by simply stopping down to f/22 and getting a soft, diffraction-limited file.

Light for form, not flatness

The defining shift described earlier, stated by St. Olaf, CaFÉ, Artwork Archive and the National Archives guide: 3D lighting should not be perfectly even — it should model form through controlled shadow. The common method starts from two lights at ~45°, then deliberately moves one (changing its distance/angle) to create directional light and shadow that reveal volume, often adding a third light as fill while watching for competing shadows. Artwork Archive: place lights "about halfway between the subject and the camera," then adjust to achieve "dynamic — but not overpowering — shadows." For ceramics, a single overhead-ish light that grounds the object with a shadow under its base is a common alternative (CaFÉ).

The National Archives guide notes that for 3D objects it is acceptable to move into directional sunlight, "as it will give you shadows that can better show the depth and form of the work."

Multiple sources (Fiveable AP Art, St. Olaf, CaFÉ, Juniata) converge on a minimum of at least two significantly different views described earlier work, and ideally a considered set covering the principal aspects plus at least one detail. Move around the work (or rotate it on a turntable/lazy Susan for consistent 360° coverage); a camera angle slightly above can reveal a hollow interior or the roundness of a form (Artwork Archive on Lauren Mabry's ceramics). Keep lighting consistent within a series even though it need not be even (St. Olaf). Shadow control and a seamless sweep keep the focus on the object.

A seamless sweep (white, gray, or gradient) prevents a horizon line behind the object; a soft contact shadow under the base "grounds" the object so it doesn't appear to float (CaFÉ). For glossy or metallic 3D pieces, expect to manage reflections with diffusion, tenting, or polarization (the 3D extension of Light — Science & Magic's reflection logic), and consider focus stacking for deep objects. This is a frequent threshold for hiring a pro.

Highest-leverage insightThe amateur error on sculpture is flat, even 2D-style lighting that erases volume; the credible move is controlled shadow that describes form without competing shadows or blown highlights.

The governing principle: documentation can become the work

For performance, installation, and ephemeral/durational work, the record is often all that survives — so documentation planning is part of making the work. GYST: "Many times, documentation is all that is left over from non-permanent work… the documentation will be the only thing that survives, and can actually become the work itself."

Performance theorist Philip Auslander, in "The Performativity of Performance Documentation" (PAJ: A Journal of Performance and Art 84, Vol. 28 No. 3, Sept. 2006, pp. 1–10), distinguishes "documentary" documentation (e.g., Chris Burden's Shoot, 1971) from "theatrical" documentation (e.g., Yves Klein's Leap into the Void, 1960), arguing that theatrical documentation supplies viewers "with the experience of a performance that never took place in the way the photograph or record presents it" — for such work, "the space of the document… becomes the only space in which the performance occurs." The practical takeaway: decide before the event how the work will be recorded, because you cannot reshoot it.

The media-conservation consensus (Electronic Arts Intermix / IMAP; Guggenheim; Matters in Media Art) is to document the whole environment, not just isolated components.

EAI/IMAP: "treat a media installation as a special kind of sculpture. Each part is equal to the whole," and document "space dimensions, light levels, acoustics, and power requirements." The Guggenheim's time-based-media program states that "on-site documentation during the install is an essential tool." Capture: a wide establishing installation view showing the work in the room and the viewer's spatial relationship to it; multiple vantage points a visitor would occupy; and, where relevant, a person interacting. Matters in Media Art's instruction set (the consortium of New Art Trust, MoMA, SFMOMA and Tate): "Photograph the installation and its individual components / Videotape the initial installation / Archive installation plans and specifications for display."

Video documentation — durable capture norms

Durable (non-dating) norms for recording video of time-based work, drawn from the Arts Midwest performance-documentation guide (Valerie Oliveiro, choreographer and former professional photographer) and EAI/IMAP: shoot at the highest available resolution and test settings during the tech rehearsal; always use a tripod, and specifically a fluid head designed for video so pans are smooth (still-photo tripods produce "herky-jerky" motion); avoid in-camera digital zoom (it "produces pixelated images"); level the frame to square; expose and focus manually for consistency; and record horizontally "for archival documentation and grant applications." For audio, capture a clean feed — Arts Midwest recommends pulling sound from the venue soundboard plus a mic near the stage for live acoustics, giving options in the edit.

A separate portable recorder (e.g., the Zoom-style field-recorder genre) yields a backup track and cleaner isolation than camera-top mics, which sound "distantly" placed unless very close (B&H). EAI/IMAP's durable archival principle: distinguish the archival master from service/reference copies, keep geographically redundant copies, and put a reference signal (color bars and tone) plus timecode at the head of the master.

There is a real tension in the sources. The practical/archival camp (Arts Midwest) urges capturing the full run from a fixed wide camera so "you never miss a full run," ideally with a second operated camera for medium/close shots, across at least two performances.

GYST cautions against over-directing the documentation: "do not try to make the portfolio of your performance into an artwork itself… Simply document your work, not your directing abilities." The fine-art-historical camp notes that a single composed still can become the canonical work — as with David Hammons's Bliz-aard Ball Sale (1983), an unannounced one-day street sale of graded snowballs laid out on a rug at Cooper Square in New York, which survives only through Dawoud Bey's series of roughly twenty photographs (Bey's print David Hammons, Bliz-aard Ball Sale I is now in the Whitney Museum collection). Critic Steven Stern (Frieze, Issue 121, 2009) called those photos "probably the most frequently reproduced images in the artist's oeuvre."

The defensible synthesis: capture comprehensive full-duration coverage as the archival baseline, and deliberately make/select strong representative stills for submissions — serving both the record and the application without letting either compromise the other.

Audio is the most-neglected variable in time-based documentation — clean, synchronized sound from a board or dedicated recorder is what makes video read as professional rather than "phone in the back of the room."

Variable/ephemeral work — record intent so it can be re-created

For variable media and works that will be re-installed or migrated, the field uses structured documentation of intent: the Variable Media Questionnaire (Guggenheim/Rhizome lineage) records a work's "appearance, functions, and behaviors… separate from any one technology," and the Guggenheim's Iteration/Identity report model (developed by time-based-media conservator Joanna Phillips and now adopted by institutions including the Met) captures both the work's defining identity and each specific installation. Phillips's model shifts the focus from a single fixed "original" toward the work's enduring identity across installations (she writes of "the absence of an original" and treats the "identity of the artwork" as the working concept).

Bake the pro option into the decision per work, not as an afterthought. The durable thresholds at which professional reproduction photography is worth it:

1. The work is highly reflective, varnished, glass-covered, or metallic (cross-polarization and reflection control are hard to do well).
2. It is large, deep, or dimensionally complex (even illumination or focus stacking get difficult).
3. The stakes are high and one-shot — a grant, MFA review, or gallery submission where the image is the work.
4. Color fidelity is critical to the work's meaning.
5. You simply cannot reliably hit even illumination, neutral color, and square geometry yourself.

The National Archives guide concedes professionals yield "the best possible quality," produced "by an individual who knows how to use the equipment properly," while framing DIY as the sustainable fallback when budgets don't allow it.

Beyond gear: reliable cross-polarized glare control, calibrated and verified color, large-work even lighting, focus-stacked depth on 3D detail, and consistent results across a body of work. Harriete Estel Berman's durable workflow tip: "Plan for all the images submitted to one jury to be photographed at the same time," and maintain a relationship with one photographer over time so the inventory stays consistent in background, lighting, and tone across a presentation. She also reframes the buy-vs-DIY math: professional camera bodies, lights, backgrounds, and the years of experience to use them often make it "wiser to invest in professional photographic services instead of equipment." For complex media installations, EAI explicitly notes institutions hire qualified media technicians for installation and documentation.

The asymmetry is the whole argument: jurors and curators judge only what they see, and assume documentation quality mirrors work quality (Berman; GYST's "Bad documentation = bad work"). A single weak image can eliminate strong work, and the marginal cost of professional documentation is small relative to the opportunity cost of a missed grant, show, or admission. The defensible decision rule: DIY when you can demonstrably control the five variables here and the stakes are moderate; hire a pro when reflectivity, scale, color-criticality, or one-shot stakes exceed your control — and when in doubt on a high-stakes submission, hire.

Highest-leverage insightThe establishing installation view that conveys scale, space, and viewer experience is what separates a credible installation record from a folder of disconnected object shots.

1. Build a repeatable capture setup and verify it every session: Two matched diffused lights at 45° for 2D; meter/test the four corners and center for evenness; tripod with levels; lens square to the work; longer focal length from farther back. Benchmark that changes the call: if corner-to-center exposure or geometry won't come into line, escalate to Recommendation 5.
2. Always put a gray card/ColorChecker in the first frame under the shooting light, shoot raw, and set white balance off a neutral patch. This is the cheapest highest-impact upgrade.
3. Match the shot to the work type: even light for flat color; controlled shadow for 3D form; raking light and focus-stacked macro for texture/detail; establishing installation views and clean horizontal video with separate clean audio for time-based work.
4. For time-based/ephemeral work, plan documentation before the event: fixed wide camera for the full run plus a second camera for detail; capture board audio plus a stage mic; record the installation environment as if it were sculpture; for variable work, log artist intent (Variable Media Questionnaire / iteration-report model) so it can be re-created.
5. Escalate to a professional when the work is reflective/large/dimensionally complex, when color fidelity is central, or when the submission is high-stakes and one-shot — and photograph all images for one jury in a single consistent session.

Highest-leverage insightA repeatable setup verified every session beats a better setup used inconsistently, because the failure you cannot see is the one that reaches the jury.