Why Does My Crystal Look Different in Different Light?

Why one crystal bracelet reads purple by a window, wine under a warm LED and blue in a phone photo, and how to check its colour before and after delivery.

In one paragraph Light is the reason. The colour you see is the lamp's light left over after the stone filters it, shaped by angle and background, and in a photo reprocessed by camera and screen. Daylight, a warm LED and a phone each change one of those steps, so one amethyst strand can read violet, wine or grey.

An amethyst strand that looked deep violet on a seller's page can arrive looking greyer under a kitchen LED, turn reddish beside a bedside lamp, and come out almost blue in a phone photo taken at a window. Nothing in the quartz has changed. Each view came from a different light source and a different angle, and the photo also carries the phone's own colour correction.

Four variables do the work: the spectrum of the light, how completely that light renders each hue, the angle and background the bead is seen against, and the processing chain from sensor to screen. Which of the four matters most depends on where a stone keeps its colour: in its body, in its inclusions, in an optical effect such as sheen, or at its surface.

The same amethyst bracelet shown under daylight, warm white LED and in a phone photo
One amethyst under daylight, under a warm LED and as a phone photo.
BE.
Bolivian Amethyst 10mm Strand Bracelet
SHOP NOW

What changes between daylight, a warm LED and a phone photo?

The spectrum of the light changes most, and a phone photo adds its own processing on top. A stone only subtracts. It absorbs some wavelengths and passes or reflects the rest, so the colour you see is whatever the lamp supplied minus whatever the stone removed.

Colour temperature says how warm the light is

Colour temperature, written in kelvin, describes the balance between blue and red in a white light. The CIE reference for average noon daylight, illuminant D65, sits at about 6504 K and is rich in blue. North-sky daylight, illuminant D75, is bluer still at about 7504 K. A tungsten filament, illuminant A, sits at about 2856 K and is heavy in red and yellow. LED packaging uses the same scale, in nominal steps from 1800 to 6500 K defined by ANSI C78.377. Soft and warm white lamps sit at 2700 or 3000 K. Cool white and "daylight" lamps sit at 5000 to 6500 K.

Under a 2700 K lamp, which is short of blue, an amethyst leans towards wine. By a north window, long on blue, the same bead leans blue-violet.

Colour rendering says what the light leaves out

Two lamps can share a colour temperature and still show a stone differently. Colour temperature only summarises the average. The colour rendering index (CRI) scores how faithfully a source shows a set of test colours against a reference such as daylight or a filament. The scale tops out at 100. Incandescent lamps score at or near 100, because a hot filament behaves almost exactly like the reference. For interior lighting, a CRI of 80 is the usual minimum and 90 or above counts as excellent. LEDs are made at both levels.

The general index, Ra, averages only eight test colours of light to moderate strength. It predicts saturated reds badly, so a separate score, R9, is quoted for strong red. The newer IES method, TM-30, uses 99 colour samples and reports both fidelity (Rf) and gamut (Rg), so it can show when a lamp mutes or exaggerates a hue that Ra never tested.

Most white LEDs make their light by passing a chip's blue or violet output through phosphors. That recipe tends to leave a gap in the cyan region between the blue and green peaks. A 2023 study in Frontiers in Chemistry measured the gap in one test lamp at 480 to 520 nm. Filling it with a cyan phosphor raised the lamp's R9 from 11.5 to 49.5. Under a lamp with a low R9, garnet looks flatter and hematoid browner. Under a lamp with a cyan gap, aquamarine, whose blue-green sits in that band, can look greyer.

Angle and background decide how light reaches the bead

A transparent bead has two ways of showing colour. Light can pass through it, which the trade calls transmitted light, or light can bounce off its surface and its inner walls, which is reflected light. A bead on a white card is fed from behind by bounced light, so its body colour reads bright and full. GIA's photography guidance makes the same point: a white background bounces light back at the subject and helps the camera settle on a correct colour balance.

A bead on a wrist gets almost no light from behind. Skin absorbs and warms whatever it returns. The bead now shows mostly reflected light from its front surface and a shorter path through its body, so under the same lamp it reads darker and warmer on the arm than on a white card.

The camera makes decisions of its own

A phone edits the scene before it saves it. Automatic white balance estimates the light from the frame itself. One common method, the grey-world algorithm, assumes the scene averages out to grey, so a frame filled with purple beads biases that estimate and the purple is pulled towards grey. Move the same strand onto a busy table and the correction changes again. Multi-exposure HDR merges several frames and compresses the brightest and darkest tones. The deep centre of a dark bead goes flat. The phone's default picture style then applies its own saturation and contrast before the file is saved.

The screen adds one more layer. Uncalibrated displays differ in white point and saturation. Two people on two devices see two different purples. Mixed light is harder still. White balance corrects for one light colour at a time, so a strand lit by a window and a warm lamp together comes out too blue where the window reaches it or too orange where the lamp does.

Light source Colour temperature Violets and purples Reds and oranges Blues Sheen and inclusions
North-sky daylight (CIE D75) about 7500 K Lean blue-violet; depth reads full Slightly cooler, less glow Strongest and cleanest Diffuse light spreads sheen; fewer sharp flashes
Noon daylight (CIE D65) about 6500 K Reference reading Reference reading Reference reading Reference reading
"Daylight" LED 5000 to 6500 K Close to daylight if CRI is 90 or more Depends on R9; low R9 flattens garnet Good on high-CRI lamps; a cyan gap greys aquamarine Point-source LEDs give sharp flashes off rutile and sheen
Neutral white office LED or fluorescent 3500 to 4000 K Greyer, slightly flat Muted, browner Acceptable; peaky fluorescent spectra vary Overhead panels wash sheen out
Warm white LED 2700 to 3000 K Shift towards wine or plum Deeper, warmer, sometimes darker Weakest; aquamarine goes grey-green Gold sheen and gold rutile look richer
Incandescent or halogen (CIE A) about 2856 K Reddish purple Fullest red; the reference for colour-change tests Weak Single filament gives the crispest sheen line

Which stones shift most, and why?

Transparent body-coloured stones shift most with the lamp's spectrum. Included stones shift with the direction of light. Stones with a sheen or glow depend on angle and on a single light source. Opaque dark stones depend on what their surface reflects. The page on how crystal colours form sets out the five physical routes behind mineral colour: absorption by metal ions, electron transfer between neighbouring ions, defects created by natural radiation, inclusions and films, and physical optics. Which route a stone uses predicts how it behaves when the lamp changes.

Amethyst, citrine and garnet carry colour through their whole volume, so every wavelength the lamp supplies is filtered on its way through. Amethyst's purple comes from iron impurities activated by natural irradiation, and purple is a mix of blue-violet and red light. A warm LED is short of blue. The red half takes over, and the stone reads wine. GIA notes that an amethyst whose colour is too dark can look black under dim lighting, and a dark strand under a dim 2700 K lamp can do the same. Citrine's yellow to orange sits in the middle of the spectrum, where every white lamp gives plenty of light, so its hue moves less. Garnet's red depends on the lamp's red tail, which is why R9 matters more for garnet than for amethyst or citrine. Which ions produce which colours is set out on the page about crystal colour chemistry.

Hematoid quartz and rutilated quartz take their colour from what is inside: iron oxide clouds, or rutile needles, held in a clear or milky host. Their host barely filters the light. What changes is how the inclusions catch it. A single lamp from one side lights the needles as bright lines and leaves the far side dark. A diffuse overhead panel lights them evenly and they go soft. Rotate a hematoid bead under one lamp and the red clouds brighten and fade as the light direction moves through them.

Moonstone and gold sheen obsidian are a third case. Their glow is an optical effect that appears only at certain angles between the light, the bead and the eye. The mechanisms are set out on the page about sheen, adularescence and Tyndall scattering, and the stone itself on moonstone for beginners. One point source, such as a single lamp or the sun, gives a crisp band of light that moves as the bead turns. Many sources, or a big soft panel, spread the effect into a haze or hide it. A phone photo taken under a ceiling of downlights can show no sheen at all on a bead that flashes freely under one lamp.

Black obsidian and other dark stones pass very little light at bead thickness. Their appearance is surface reflection: the sharpness of the polish, the shape of the highlight and the colour of whatever the surface mirrors. Under a warm lamp the highlight looks warm and the body looks brown-black. Under a cool window the highlight looks white and the body looks blue-black. The highlight's colour comes from the lamp; its shape and sharpness come from the polish.

Material class Examples What changes with the light What to check
Transparent, body-coloured Amethyst, citrine, garnet, aquamarine Hue and depth track the lamp's spectrum and CRI; darkest stones go black under warm dim light Depth of colour by a north window, then under the lamp it will be worn beside
Inclusion-coloured Hematoid quartz, rutilated quartz, phantom quartz Contrast and brightness of the inclusions track light direction; body barely changes Rotate under one lamp; count how many beads show the inclusion at wearing distance
Phenomenal Moonstone, gold sheen obsidian The effect appears only at one angle and under one source; soft or mixed light hides it One lamp, bead turned slowly; ignore photos taken under multiple ceiling lights
Opaque, dark Black obsidian, dark garnet, smoky quartz at depth Highlight colour and body tint follow the room; polish quality sets the highlight Highlight shape and evenness of polish, on skin, under the usual lamp

Is this the same as a colour-change stone?

No. A colour-change stone has an absorption pattern balanced so finely that two lamps push it into two different hues. GIA describes fine alexandrite as green to bluish green in daylight and red to purplish red under incandescent light. GIA has also documented a pyrope-spessartine garnet with a high grossular component that reads yellow-green under daylight-equivalent lighting and orange under incandescent light, measured under CIE illuminants D65 and A. The authors describe the change as obvious. Both are recognised colour-change materials. A typical red garnet strand only darkens or brightens with the lamp and stays red.

An ordinary amethyst going wine-coloured under a warm LED is a rendering effect. The stone's absorption has stayed put. The lamp's spectrum has moved, and the stone stays purple. Colour science names a related effect that matters for strands: illuminant metameric failure, where two samples match under one light and separate under another. Two beads with slightly different absorption can match by a window and part company under a warm lamp, so a strand that looks even in daylight can look less even in the evening.

BE.
Hematoid Quartz Strand Bracelet
SHOP NOW

How can a strand's colour be judged before delivery?

Start from the lighting the seller states, and compare every photo against it. Check two things: whether the listing says how its photographs were lit, and whether the seller can answer when asked. Then ask for two photos of the actual piece, one under neutral light and one on the wrist. Photographs add their own variation on top of the differences between beads, a point covered on the page about why crystal bracelet strands vary.

BE. records its readings under stated conditions. The Crystal 4T™ page sets out that Tone, one of the four readings, is taken under daylight-balanced light at 5500 K over a white reference card, with the unaided eye at about 40 cm, and that Texture is read at 10× magnification. A buyer can come close to the Tone condition with a daylight LED rated 5000 or 5700 K, the nominal steps either side of 5500 K, and a sheet of white card. Each strand in the BE. collection shows its Tone word on the product page. The wider checks on finish, drilling and construction are listed under 6 things to check before buying crystal jewellery, and amethyst's colour grades in the amethyst complete guide.

How can the colour be checked after delivery?

BE.
Moonstone Strand Bracelet
SHOP NOW

Look at the strand under three lights, then photograph it once under controlled conditions.

  1. Stand by a north-facing window in daytime, out of direct sun. This is the closest a home gets to the daylight-balanced light that Crystal 4T™ readings use. Note the hue and how deep the colour looks at arm's length.
  2. Move to the room where the bracelet will be worn, under its usual lamp, with the strand on the wrist. This is the reading that matters for wearing, and it will be darker and warmer than the window view.
  3. Hold the strand under one lamp alone, with other lights off, and turn it slowly. Stones with inclusions or sheen show their needles, clouds or glow most clearly this way.
  4. Photograph it on a neutral grey card under the daylight LED, with the white balance fixed in the camera's manual or pro mode. Automatic balance re-guesses every frame. Turn HDR and any beauty or vivid mode off.
  5. Compare that photograph with the seller's images and with the seller's stated lighting. A shift in the direction the table above predicts is the lamp. A shift in the other direction, or a change in the pattern of inclusions, is worth a question to the seller.
  6. Keep the grey-card photo as the reference for any later comparison.

Frequently asked questions

Q1. Why does my crystal look different in different light?

Because a stone only filters the light it is given. Daylight is rich in blue, a warm LED is rich in red and short of blue, and a phone camera re-balances the frame on its own. Each source hands the stone a different starting spectrum, so the colour that comes out differs even though the stone has not changed.

Q2. Why does my amethyst bracelet look grey or brown under LED lights?

Amethyst's purple is a mix of blue-violet and red light, and warm white LEDs at 2700 to 3000 K supply little blue. The red half takes over and the beads read wine or brownish. In dim light a dark strand can look grey-black. By a north window or under a daylight LED with a CRI of 90 or more, the same beads read violet again.

Q3. Which light shows a crystal's true colour?

Daylight-balanced light over a white background is the usual reference. Crystal 4T™ readings use 5500 K over a white reference card, and indirect daylight from a north-facing window is the closest home equivalent. For daily wear, the colour under the lamp in the room where the bracelet is worn matters more.

Q4. Why do my phone photos of a bracelet look more saturated than the real thing?

The phone's default picture style applies its own saturation and contrast, HDR compresses the tonal range, and automatic white balance shifts hue depending on what else is in the frame. Fix the white balance in manual or pro mode, switch HDR and vivid modes off, and place a neutral grey card in the shot to get a photograph that can be compared with anything else.

Q5. Can I tell whether a bracelet is dyed by how it changes under different lights?

No. Dyed and natural stones both filter the lamp's light, so both shift when the lamp changes. The visible sign of dye is where the colour sits. Dye enters through cracks, so it gathers along them, with paler stone between. That is a reason to look for a treatment statement on the listing or the Stone Origin Record. Identification of the material itself needs gemmological testing.

Q6. What lighting does BE. use when it records a stone's Tone?

The Crystal 4T™ page states the conditions: daylight-balanced light at 5500 K over a white reference card, with the unaided eye at about 40 cm for the Transparency and Tone readings, and 10× magnification for Texture. A buyer can come close to the Tone condition at home with a daylight LED rated 5000 or 5700 K and a sheet of white card.

References
  1. Wikipedia: Standard illuminant (CIE illuminants A, D50, D55, D65, D75 and F series, with correlated colour temperatures)
  2. Wikipedia: Color temperature (table of common sources, 2700 K to 6500 K)
  3. Designing Light: ANSI C78.377 adds two new CCTs (nominal steps 1800 to 6500 K)
  4. Wikipedia: Color rendering index (Ra, R9 and typical lamp ratings)
  5. Ansell Lighting: Light source colour rendition, ANSI/IES TM-30 (99 colour samples, Rf and Rg)
  6. Wikipedia: Metamerism (color)
  7. Wikipedia: Color balance (automatic white balance and grey cards)
  8. Apple Inc. patent application US20130093917A1: Alleviating dominant color failure in automatic white balance (grey-world assumption)
  9. Wikipedia: Multi-exposure HDR capture
  10. U.S. DOE Solid-State Lighting programme: LED Basics (phosphor conversion, CCT, CRI and TM-30)
  11. Frontiers in Chemistry (2023): The developments of cyan emitting phosphors to fulfill the cyan emission gap of white-LEDs
  12. GIA: Alexandrite Quality Factors
  13. GIA Gems & Gemology (Spring 2017): Pyrope-Spessartine Color-Change Garnet with a High Grossular Component
  14. GIA: Amethyst Quality Factors
  15. Wikipedia: Amethyst (iron colour centre formed by irradiation)
  16. GIA: An Introduction to Gem Treatments (dyeing of porous or fractured gems)
  17. GIA: How to Photograph Gems & Jewelry
  18. Nassau, K. (1978). The origins of color in minerals. American Mineralogist 63, pages 219 to 229
  19. Hunt, R. W. G. and Pointer, M. R. (2011). Measuring Colour, 4th edition. Wiley