Color Temperature Management
Color temperature is the difference between a carpet that reads “wool” vs “polyester” on a customer’s screen, between a brand red that reads as the agreed brand red vs as orange. The color fidelity discipline isn’t one decision — it’s a chain of small disciplines across fixtures, cameras, capture, and review. This manual covers what color temperature is, why drift happens, and what to do about it.
Part of the Lighting Techniques reference library.
§1 — What color temperature is
Color temperature is a numerical description of a light source’s color, measured in Kelvin (K). The scale is derived from the physics of an idealized black-body radiator — at different temperatures it emits different colors of light.
Common reference points:
| Color temperature | Source | Visual character |
|---|---|---|
| 1800 K | Candle flame | Very warm, deep orange |
| 2700–3000 K | Tungsten / incandescent | Warm yellow-white |
| 3200 K | Studio tungsten / film standard | Warm white |
| 4000 K | Neutral white / fluorescent | Slightly warm white |
| 4500–5000 K | Cool fluorescent / sunrise | Neutral white |
| 5500 K | Daylight (mid-day) | Neutral, slightly cool |
| 5600 K | Strobe / electronic flash standard | Neutral, slightly cool |
| 6500 K | Overcast daylight / D65 monitor white | Cool white |
| 7500 K | Shade in sunlight | Cool blue-white |
| 10000+ K | Clear blue sky | Very cool, blue |
Lower K = warmer (more red / orange); higher K = cooler (more blue). PhotoRobot studio work typically operates around 5500–5600 K (daylight / strobe standard) for general capture, sometimes 3200 K for warm-tone product photography or specific brand aesthetics.
1.1 — Why color temperature matters
The human eye + brain compensate for color temperature in real time — a white shirt looks white under tungsten lamp + under daylight, even though objectively the light hitting the shirt is very different color. The camera doesn’t compensate — it records the actual photons.
When the camera records the scene + the result is viewed on a screen calibrated to a different color temperature, the perceived colors shift. Customers viewing on different devices, in different lighting conditions, perceive different color from the same image. For color-critical work (catalog with brand colors, QC sample documentation, automotive paint reference), this is a problem.
Color temperature management is the discipline of controlling the color of light at capture + maintaining color fidelity through post-production + delivery.
§2 — Sources of color temperature drift
In a typical studio session, color temperature can drift between:
2.1 — Different fixtures
Even fixtures rated for the same nominal color temperature can differ — manufacturing tolerances, aging, batch variation. Two “5600 K” strobes may actually be 5500 K + 5650 K. Across many fixtures in a setup, the discrepancies compound.
2.2 — Different cameras
Cameras’ color rendering varies — different sensor chemistry, different color profiles, different white-balance algorithms. Same scene captured with two cameras can produce slightly different color even with identical white-balance settings.
For multi-camera rigs (Carpet Studio’s 7-camera, Catwalk’s multi-fixed-camera), per-camera color calibration is essential — without it, captures from different cameras of the same subject will mismatch.
2.3 — Fixture warm-up + aging
- Tungsten / halogen fixtures shift slightly cooler as they warm up + slightly warmer as the bulb ages
- Fluorescent fixtures need 5–10 minutes warm-up to stabilize color
- LED fixtures are generally most stable but can drift over years of use
- Strobes are stable per flash but can shift color over flash-to-flash + during long sessions as power packs heat up
2.4 — Mixed-source lighting
If a session uses fixtures of different types (tungsten + LED + ambient daylight), each contributes light of different color. The mixed result is hard to white-balance against because no single white-balance setting matches all sources.
2.5 — Ambient bleed
Window light, room lighting, computer screens — all contribute ambient light that may differ in color from the intentional fixtures. Even small ambient contribution can shift overall color subtly.
2.6 — Mains voltage fluctuation
Tungsten + halogen fixtures’ color output varies with input voltage — fluctuations during the session (other equipment cycling on / off, mains supply variation) produce color drift.
§3 — White-balance discipline
White balance is the camera setting that tells the camera “this color temperature is what I’m calibrating against — render colors relative to this.” Setting white balance correctly is the foundational color discipline.
3.1 — Methods to set white balance
Auto white balance (AWB). Camera detects the scene’s overall color + adjusts. Convenient but unreliable for production capture — varies frame-to-frame, doesn’t match across cameras, can be fooled by dominant colors in scene. Don’t use AWB for production capture.
Preset white balance. Camera offers presets for common conditions (daylight, tungsten, fluorescent, flash). Pick the preset matching your dominant light source. More reliable than AWB but coarse.
Manual K setting. Camera lets you specify exact Kelvin value. Match fixture’s nominal color temperature. Reliable when fixtures match nominal spec.
Custom white balance via reference card. Camera captures a white or gray reference card under the actual session lighting + uses that capture to calibrate. This is the canonical production discipline — adapts to actual session conditions, accounts for fixture drift + mixed-source, repeatable across cameras (if all cameras calibrate against the same reference card under same lighting).
3.2 — Custom WB procedure (canonical)
- Place reference card (white or 18% gray card from photographic supplier, or X-Rite ColorChecker) on the platform at the capture position
- Configure each camera in CAPP to capture a frame of the reference card
- Set white balance on each camera from that capture (camera menu → custom white balance from captured image)
- Verify by capturing the reference card again post-WB-set + confirming it renders as neutral
- Optionally save WB setting as a named preset in the camera if it’ll be reused
Repeat at session start; re-check periodically (especially long sessions where fixtures may drift).
3.3 — Multi-camera WB consistency
For multi-camera rigs, all cameras white-balance against the same reference card under same lighting. This ensures all cameras render color consistently. Per-camera drift still happens (per §2.2) but the WB calibration normalizes the starting point.
If different cameras have noticeably different color rendering after consistent WB calibration, post-production color normalization can correct further — but get capture-time WB as consistent as possible first.
§4 — Mixed-source compensation
When session lighting comes from multiple source types (e.g., tungsten key + LED fill, or strobe + ambient window light), white balance has no single correct value. Options:
4.1 — Eliminate one source
Simplest: shut down or filter out the secondary source so all light is from one type.
- Close window blinds to eliminate daylight contribution
- Turn off room overhead lighting during capture
- Use only fixtures of one type (all tungsten, all LED, all strobe)
For most production capture, single-source is preferable to mixed-source.
4.2 — Gel one source to match the other
Colored filter gels placed on fixtures shift their color to match another source. Common gels:
- CTO (Color Temperature Orange) — warms a cool fixture (5500 K → 3200 K with full CTO)
- CTB (Color Temperature Blue) — cools a warm fixture (3200 K → 5500 K with full CTB)
- Partial CTO / CTB — partial shift for fine matching
With both sources matched in color, white balance can be set normally.
4.3 — Accept the mixed-source character
For some creative or documentary capture, the mixed-source character is part of the desired aesthetic (e.g., a product photographed with a deliberately warm key + cool back for visual interest). White balance to one dominant source + accept the other source’s color contribution as creative element.
For production catalog capture, mixed-source acceptance is not the canonical approach — it sacrifices color fidelity for ambiguous gain.
§5 — Reference card discipline
The reference card is the color discipline’s foundation. Three card types are used:
5.1 — White card
Plain white card. Sets white balance — tells camera what “neutral” should look like under current lighting.
Inexpensive; available from any photographic supplier. Card surface should be matte (not glossy) to avoid specular reflection during WB capture.
5.2 — 18% gray card
Standard mid-gray reference (named for 18% reflectance). Sets both white balance + exposure baseline.
Slightly more useful than white card because it also acts as exposure reference (most camera light meters are calibrated to 18% gray). Common in product photography.
5.3 — Color reference target (X-Rite ColorChecker, Datacolor Spyder Checkr, etc.)
A target with multiple known color patches. Used for both white balance + full color calibration (post-production tools can derive a color correction profile from the target capture).
Essential for color-critical work — catalog with brand colors, automotive paint sample documentation, fabric production proofing. Each patch on the target has known color values; post-production tool reads the captured patches + computes the correction to bring them to known values.
5.4 — Reference card session protocol
- Capture reference frames at session start — at minimum white card, ideally color reference target, captured by every active camera under session lighting
- Store reference frames with session captures — typically in a dedicated
references/subfolder - Re-capture reference frames when:
- Lighting changes (fixture moved, repositioned, replaced)
- Long session interval (more than ~2 hours)
- Suspected color drift (visual indication in captures)
- Post-production uses reference frames to derive WB + color corrections, applied across the session’s production captures
The 60 seconds spent on reference frames at session start prevents post-production guessing later.
§6 — Color-critical session protocols
For sessions where color fidelity must meet contractual or regulatory standards (automotive paint, brand catalog, fabric production proofing):
6.1 — Pre-session
- Warm fixtures. All fixtures on + warmed up 15–30 minutes before capture
- Stable mains. Ensure no other high-draw equipment cycling on/off during session
- Eliminate ambient. Close blinds; turn off room lights; clear other light sources
- Reference targets ready. Color reference target on hand + clean
6.2 — Session start
- Custom white balance on every active camera against reference target under session lighting
- Capture full reference frame set — white card + gray card + color target — by every camera
- Verify by re-capturing reference target post-WB + confirming neutral rendering
6.3 — Mid-session monitoring
- Periodic reference re-capture — every 30–60 minutes or per significant scene change
- Visual check on captures — periodic glance at thumbnails for unexpected color shift
- Strobe / fixture status — verify no fixture failure or output degradation
6.4 — Post-session
- Reference frames travel with captures — post-production tools read them for WB + color correction
- Note any session events affecting color (fixture issue, voltage fluctuation, etc.) in session log for post
- Customer review proofs against reference standards — if customer reports color discrepancy, reference frames are the audit trail
§7 — Common problems + recovery
7.1 — Captures look color-cast (everything too warm / too cool)
Symptom. All captures from session have a unified color cast (e.g., everything slightly orange or slightly blue).
Likely cause. White balance set incorrectly at session start — possibly default WB used instead of custom, or reference card was not neutral, or session lighting changed without WB recalibration.
Recovery. Post-production WB correction can shift the overall color. Use reference frames if captured (best); otherwise correct by eye against a known-color element in the scene; otherwise document as session issue + reshoot if customer rejects.
7.2 — Different cameras render same subject in different colors
Symptom. Same sample captured by Cameras 1 + 2 + 3 looks like three different colors despite identical scene.
Likely cause. Per-camera WB not consistent — each camera calibrated independently or some cameras using default WB.
Recovery. Ensure all cameras WB-calibrate against same reference card under same lighting at session start. Post-production can apply per-camera color correction if reference frames are available.
7.3 — Strobe-lit captures inconsistent in color across session
Symptom. Strobe captures from early session have different color than late session.
Likely cause. Strobe color drift over long session — capacitor heating, fixture aging, voltage fluctuation.
Recovery. Re-calibrate WB mid-session; in post, normalize color using mid-session reference captures.
7.4 — Tungsten fixture shifted warmer through session
Symptom. Tungsten-lit captures progressively warmer (more orange) through session.
Likely cause. Tungsten bulb aging effect compounded with extended session use.
Recovery. Acceptable drift may be small; for color-critical sessions, schedule bulb replacement before sessions or switch to more color-stable fixtures (LED).
7.5 — Color reference target captures look wrong
Symptom. Captured color target patches don’t match expected color values (when checked against target specification sheet).
Likely cause. WB not set correctly; mixed-source lighting; faulty fixture; target itself damaged or aged.
Recovery. Verify target is clean + undamaged; re-calibrate WB carefully; if persistent, replace target (targets do fade over years).
7.6 — Customer reports brand color wrong
Symptom. Customer’s review shows brand-defined color (e.g., Pantone reference) rendered incorrectly on captures.
Likely cause. Capture-side color fidelity insufficient; or display device color profile differs from customer’s reference profile; or printing pipeline shifting color further.
Recovery. Establish color audit trail with reference frames + color calibration target. Compare customer’s expected color against capture-side reference. Where color matches reference-side but not customer-side, investigate downstream (display, print, customer device). Document findings; iterate with customer.
§8 — PhotoRobot-specific application
8.1 — Carpet Studio
Multi-camera (7) rig with mix of strobe + cyclorama-reflected light. Per-camera WB calibration at session start is mandatory; mid-session re-calibration for sessions >2 hours.
Color-critical sessions (carpet brand catalog, automotive textile QC) follow the §6 protocol.
8.2 — Catwalk
Continuous lighting (multiple Fresnel / panel fixtures). Pre-warm fixtures; WB calibrate every camera against reference under session lighting; periodic re-check.
Fashion video is somewhat more forgiving than catalog stills for minor color drift (motion + viewer attention to overall feel rather than specific color). But cosmetic + jewelry segments within fashion can be color-critical.
8.3 — Standard product turntables
Single-camera or limited-camera rig with strobes. WB calibration at session start; per-batch re-check for high-throughput catalog sessions.
8.4 — Multi-vertical studios
Studios serving multiple verticals (carpet + automotive + fashion + electronics) may have different color discipline per vertical. Standardize on a single discipline level (the strictest customer’s requirement) + adapt only when sessions allow less strictness.
§9 — Decision checklist
§10 — Further reading
- Continuous vs Strobe Lighting — color stability differs between continuous + strobe
- Studio Lighting Setups for PhotoRobot Sessions — three-point foundation + PhotoRobot multi-light reality; color calibration applies to any lighting structure
- Raking Light Technique — texture-revealing technique; color discipline applies same
- Fiber-Specific Lighting Considerations — material rendering interacts with color fidelity
- Lighting Manuals reference library — return to library overview
For PhotoRobot-specific capture device manuals, see photorobot.com/manuals.