Patterns in photographs: what survives, what shimmers, and what steals the frame
Patterns fail photographs in two unrelated ways. A fine regular repeat interferes with the pixel grid and shimmers, which is a sampling problem and depends entirely on how large the picture is displayed. A bold pattern competes for attention, which has nothing to do with size. Scale and contrast decide, not the pattern's name, which is why a tonal paisley is safer in a headshot than a high-contrast pinstripe.
In short
- Two unrelated failures: fine regular repeats shimmer, bold repeats compete for attention. Work out which one you face before you change clothes.
- Shimmer is a sampling problem, so it depends on display size. The same jacket can be flawless at full width and broken in a feed thumbnail.
- A tonal paisley is safer than a high-contrast pinstripe. Scale and contrast decide, not what the pattern is called.
- Texture — cable knit, twill, oxford basketweave — gives a photograph depth with none of the pattern risk, and almost nobody reaches for it.
- Nothing with legible text, a logo, or a face on it. People read and recognise involuntarily, and the eye goes there first every time.
Two ways a pattern fails, and they point in opposite directions
“Avoid patterns” is the most repeated wardrobe instruction in professional photography, and it is one real phenomenon over-generalised until it stopped being useful. A pattern damages a photograph in exactly two ways, with different causes and different fixes, and any rule treating them as one will talk you out of clothes that would have been perfectly fine.
The first is shimmer. A regular repeat too fine for the available pixels beats against the sampling grid and produces bands, colour fringes and a crawling texture that were never in the cloth. That is not a matter of taste but a consequence of how any grid of samples records a repeating signal, and the physics is cited at the foot of this page.
The second is competition. A bold pattern with an irregular repeat — a paisley, a large floral, a hand-blocked print — never shimmers, because an irregular repeat gives the grid nothing regular to beat against. It fails, when it fails, for an unrelated reason: it becomes a second thing to look at in a photograph meant to have one.
Hence the conclusion this page exists for, which surprises most people who hear it. A tonal paisley is safer in a headshot than a high-contrast pinstripe. The pinstripe is small, regular and high in contrast, the precise recipe for shimmer; the paisley is large, irregular and, if it is tonal, quiet. Scale and contrast decide, not the name.
The spectrum, ordered by how big the repeat is
Sort every fabric you own by the size of its repeat rather than by what the pattern is called, and it lands in one of seven bands, each with its own behaviour. Two garments sharing a name — a hairline stripe and a chalk stripe are both stripes — can sit two bands apart and behave nothing alike.
Read the table as a diagnosis rather than a permission list. Only the last row is a flat no, and it is a no for a reason that has nothing to do with either failure mode above.
- The two middle bands are where nearly every real decision lives: micro patterns shimmer, and small classics survive a portrait but fall apart in a feed.
- A chalk stripe beats a pinstripe for exactly one reason: the repeat is wider and the edges are soft, so there is less regular high-contrast detail for the grid to fight with.
| Band | Examples | What happens | Verdict |
|---|---|---|---|
| Solid | plain weave, jersey | No repeat: nothing to interfere with, nothing to look at | Always safe, never memorable |
| Texture | cable knit, waffle, twill line, oxford basketweave, linen slub, corduroy wale, flannel nap | Too small to resolve, so it reads as fabric — depth without pattern risk | Underused in headshots |
| Micro / “neat” | pinstripe, hairline stripe, birdseye, nailhead, sharkskin, end-on-end, fil-à-fil, micro-gingham | The shimmer zone. A tonal birdseye survives, a white-on-navy hairline stripe does not | Risky — check contrast, not the name |
| Small classics | chalk stripe, glen or Prince of Wales check, houndstooth, gingham, tattersall, bengal stripe | Resolves at portrait size, breaks up in a feed thumbnail | Depends on display size |
| Architectural | windowpane, tartan, madras, awning stripe, buffalo check | No shimmer at any size. Dominates | Fine if the frame is quiet |
| Expressive / organic | paisley, floral, botanical, ikat, shibori, block print, batik, abstract watercolour | Irregular repeat: no interference at any size, only attention | Contrast alone decides. Tonal is safe |
| Never | logos, slogans, legible text, lanyards, sequins, metallic, novelty print, anything with a face on it | Text and faces are read involuntarily, before anything else | No |









Display size is the real variable
Shimmer is a sampling problem, so it is not a property of the jacket but of the jacket at a particular size. A suit that looks superb at full width on a laptop can be a mess of bands and speckle in a directory listing, because the pixels covering each stripe fell below what it takes to record the stripe as a stripe.
It is why the advice you get is so inconsistent: someone judging a large file on a calibrated screen and someone glancing at a small circle in a candidate list are looking at two different photographs. Drag the slider below: the fine regular swatches dissolve first, and the irregular ones never at all.
So the useful question is not “does this pattern work” but “does it work at the size this photo will be seen”. The sizes a LinkedIn photo is actually displayed at are far smaller than the file you upload, and that is the size your fabric has to survive.

400 px wide
48 px wide
Scale ruler
Ten fabrics, all shown at the same size, and you pick the size. Drag toward the small end and watch which ones hold together. The fine regular repeats stop being resolved somewhere on the way down; the wide and the irregular ones just get smaller.
160 px, about a profile photo on a web page
Pinstripe
Micro · High contrast
shimmering, 1.6 px repeat
The case the usual advice is built on.
Hairline stripe
Micro · High contrast
shimmering, 1.2 px repeat
The finest repeat here, and the highest contrast. It goes first.
Birdseye
Micro · Tonal
blurring, 1.2 px repeat
Finer than the pinstripe and it survives anyway. Contrast is why.
Micro-gingham
Micro · Two-tone
softening, 2.4 px repeat
Houndstooth
Small classic · High contrast
clean, 6.4 px repeat
Glen check
Small classic · Muted
clean, 9.6 px repeat
Also sold as Prince of Wales when it carries the overcheck.
Windowpane
Architectural · High contrast
clean, 19 px repeat
Tartan
Architectural · Deep tones
clean, 22 px repeat
The fine overstripe goes first. The wide bands hold at every size.
Paisley
Expressive · Tonal
steady, 26 px repeat
Irregular, so there is nothing for the pixel grid to fight with. It only gets smaller.
Oxford basketweave
Texture · Tonal
cloth, 2.4 px repeat
Reads as cloth rather than as pattern at every size on this ruler.
A repeat needs roughly two pixels of width to survive being shown small. The figure under each swatch is that repeat at the current size, so you can watch a fabric cross the line. What happens on the other side is decided by contrast, not by the pattern’s name: the hairline stripe breaks into something that looks like a fault in the picture, while the birdseye — which is finer — settles quietly into a plain tone. The paisley never breaks at all, because an irregular motif has nothing to interfere with. This ruler only tests whether a pattern survives being shown small; whether it competes with your face for attention is a separate question, and the wide checks are the ones that do.
Texture is the option almost nobody uses
The texture band is badly underused. A cable knit, a waffle weave, an oxford basketweave, the wale of a fine corduroy, the nap on flannel, the slub in a linen — all have a repeat small enough that no camera at portrait distance resolves it as a pattern. What survives is surface: the garment reads as cloth with depth rather than a flat panel of colour.
The usual alternative is a plain solid, which does its job by disappearing completely. Texture disappears too — nobody looks at it instead of your face — but it leaves the photograph something to hold onto. It is the only band that buys interest without spending attention.
The three examples below are from our own catalogue, and the point of the comparison is the fabric rather than the styling. The full set at proper size shows the difference better than any description does.

A crew-neck knit is the texture band doing its job: the stitch is far too small to resolve as a pattern, so it reads as depth and nothing more.

The control case: a solid suiting cloth has no repeat to interfere with anything, which makes it the safest fabric in the catalogue and the least interesting.

The fabric contributes almost nothing here and the contrast between garment and background does all the separating — the pattern question is sidestepped entirely.
Created with PoseLab from selfies. Not stock photography.
Traditional and cultural dress: the same two questions, and no others
Every other guide handles this in a single sentence — wear a plain navy top — which tells a large share of its readers that their clothes are a problem, and is not even photographically true. The two failure modes above are properties of cloth, light and pixels, and they apply identically to a suit, a sari and a hanbok. The only thing worth saying is which band a fabric sits in and what that band predicts.
What follows is a diagnosis rather than a suggestion. You already know what you want to wear, and we are not in a position to have a view about that. This is only an account of what the camera does with the cloth, band by band, so you can predict the result instead of being surprised by it.
- Ankara and kente: large, regular, high in contrast. Architectural band. No shimmer at any size, and the first thing the eye lands on — which is sometimes the intention.
- Batik, shibori, ikat and block print: irregular repeats, so no interference however small the picture gets. Contrast alone decides — a tonal indigo behaves almost like a solid, a sharp two-colour print like a large pattern.
- Sari and salwar kameez: a head-and-shoulders crop only shows the cloth near your face, so the drape below may never appear. A fine metallic zari border at the shoulder is in the shimmer zone and also catches light — two effects at once.
- Kurta and guayabera: usually solid or tonally embroidered, which is the texture band and the safest there is. White fabric under a strong light is a brightness question, not a pattern one.
- Hanbok and kimono: large organic motifs on a plain ground. Architectural or expressive depending on the piece; either way they resolve cleanly and dominate, which works when the background behind you is quiet.
- Hijab, turban and other headwear: the weave is the whole question, because it sits closer to your face than anything else. A matte cloth behaves like a solid, a fine regular stripe is the one case worth a second look, and a satin sheen is a highlight rather than a pattern.
- Tartan: architectural. Regular, wide and high in contrast, so it resolves at every size and dominates the frame.
- Embroidered blouses: raised thread reads as texture at portrait size and vanishes at thumbnail size, which is usually the outcome you wanted.
Wearing two patterns at once
A head-and-shoulders crop rarely shows enough of two garments for this to come up, but it does happen — a checked shirt under a windowpane jacket, a printed scarf over a textured knit. What makes it work is scale separation: the two repeats have to be far enough apart that nobody reads them as the same kind of thing arguing with each other.
A windowpane over a fine texture works because one is architectural and the other barely visible. A pinstripe under a bengal stripe does not: two similar repeats in one frame produce a visual buzz and neither wins. If in doubt, wear one pattern and one plain, with the plain nearer your face.
When the patterned jacket is the only jacket you own
This is the common real-world case, and it has decent answers. Unbutton the jacket: a column of plain shirt down the centre of the frame pushes the pattern to the outer edges, where the eye spends less time. Turn a shoulder towards the camera so the patterned lapel is foreshortened and reads as a smaller area of cloth.
The third move is depth of field. With a wide aperture, anything more than a few inches behind the plane of your face softens, and a shoulder set back from the camera softens with it. A pattern slightly out of focus stops resolving as a pattern and starts behaving like texture, which is the band you wanted anyway. Keeping the light off the patterned side lowers its contrast, and contrast is half the problem.
Fabric under scrutiny behaves the same way in a meeting room or on a stage. Our guide to dressing for work covers that wider version, including the parts that have nothing to do with being photographed.
If the headshot is AI-generated, the pattern moves upstream
The wardrobe in an AI headshot comes with the style you pick, so you are not choosing the fabric in the finished picture — you are choosing the fabric in the selfies you upload, where the calculation is different. A busy print near your face gives the process more to disentangle: which edges belong to your jaw and which to a piece of cloth, and every extra edge is another chance to get that slightly wrong.
So the reference selfies want to be plain: solid colour, matte fabric, nothing with text on it, nothing with a repeat fine enough to shimmer in a phone photo taken in poor light. How to take reference selfies covers the rest, and it is the highest-leverage part of the process. Once the uploads are clean, the pattern question belongs to the style you choose rather than to you.
A fine repeat can also arrive already shimmering from the phone that photographed it, and that shimmer is then part of what the process works from rather than a fault anyone can clean up afterwards.
Sources
- Nyquist–Shannon sampling theorem (Wikipedia)(opens in a new tab)
The sampling result behind the claim that a repeat finer than the grid can record cannot be reproduced faithfully.
- Moiré pattern (Wikipedia)(opens in a new tab)
What the interference between a regular fabric repeat and a sensor or screen grid actually looks like, and why fine stripes are the classic case.
- Aliasing (Wikipedia)(opens in a new tab)
Cited for the general effect of sampling a signal too coarsely, which is what happens when a photograph is displayed smaller than the detail in it.
Platform rules and specifications change. Everything quoted here was checked against the live pages on August 23, 2026; if you spot something out of date, tell us and we'll fix it.