Product photography for shirts: a catalog-ready method

Product photography for shirts works best when every image is assigned a claim. Use a factual front or ghost-mannequin frame to establish the product, detail views to show construction, and on-model or styled images to add fit and context. The job is not to make one shirt look vaguely polished. It is to build an image set that keeps the collar, sleeves, placket or print, fabric, color, and hem consistent from one SKU to the next.

That distinction protects both creative quality and product truth. A flat lay can show print placement clearly but says little about fit. A model image can show styling and proportion on one body but should not replace a clean product view. Choose the evidence job first, then choose the display method.

Choose the image claim before the display method

Start with three image jobs: fact, form, and context. The fact frame identifies the exact shirt and variant. The form frame shows how the shirt holds volume or sits on a body. The context frame places it inside an outfit, activity, or campaign. One photograph may help with two jobs, but no photograph should be asked to prove what it cannot show.

For a graphic T-shirt, the factual frame may be a square-on flat lay because print size and placement are the buying details. For a structured button-up, a ghost mannequin can make the collar, placket, cuffs, and body shape easier to read. When shoulder fit, sleeve length, or styling matters, add an on-model frame. Google Merchant Center currently recommends a clear main image and supports additional images for other angles, details, product use, and lifestyle context.

Decision rule: assign each image one primary claim, then state what that frame cannot prove.

This claim-first order prevents a common catalog failure: attractive images that leave basic questions unanswered. It also clarifies the brief. Instead of requesting “a premium shirt photo,” specify “front product-truth frame with the collar, print, sleeve shape, and hem fully visible” or “4:5 context frame that preserves the approved color and print while showing an untucked silhouette.”

Prepare the collar and center line before the silhouette

Styling should begin at the shirt's most recognizable construction points. Confirm the SKU, colorway, size sample, and product condition. Follow the care label when steaming or pressing. Then set the neckline or collar, align the shoulder seams, establish the center line, and only afterward shape sleeves, side seams, and the hem.

For a crew-neck T-shirt, keep the neck rib even and visible. Do not stretch the torso into a perfect rectangle if the real cut narrows or widens. For a button-up, set the collar curve and placket first, decide which buttons should be closed, and keep the cuff treatment consistent across the range. Profoto's shirt-specific flat-lay tutorial uses shoulder and hem alignment, small amounts of tissue for volume, hidden pins or folds, and reflected fill to control shape without letting the garment collapse.

Make every intervention reversible and record the ones that affect the outline. Tissue can support a shoulder or sleeve. Hidden clips can keep a side seam from drifting. Neither should create a waist, shoulder width, or sleeve shape the product does not have. When a sample defect is temporary, note it for retouching. When it is part of the sellable item, do not quietly erase it.

Approve the shirt through five product-truth zones

Use the same five-zone review for the product frame, every detail, and every derivative. This is a Look Atlas editorial framework, not an external standard. Its purpose is to make approval language observable rather than subjective.

Use the same five zones on the factual source and every derivative.

The table is intentionally short enough to use beside a contact sheet. A reviewer should be able to name the zone that failed. “The second button has shifted off the placket” creates a repair. “It feels wrong” creates a meeting.

Lock geometry before tuning the light

Consistency begins with a fixed relationship between camera, garment, and crop. For a flat lay, keep the sensor parallel to the surface and place the shirt near the optical center. For a mannequin or model, set the camera height against the shirt's chest rather than improvising from frame to frame. Mark the camera position, crop boundary, garment center, and support position before the first production batch.

A tripod is useful even when the shutter speed does not require one because it preserves framing. Shopify's current product-photography guidance explicitly recommends keeping the camera and tripod in the same position for consistency. If you use a phone, stay with the main camera, avoid digital zoom, disable beauty or portrait effects, and lock focus and exposure when the device allows it.

Start with one broad diffused source and a white or black card opposite it. Move the source across the fabric enough to reveal weave, ribbing, seams, and folds. Use white fill when a dark shirt's shadow side loses detail. Use a black card outside the frame when a white shirt loses its edge against a pale background. Shopify documents both white cards for brighter shadows and black cards for deeper edge definition on white products.

Do not standardize an aperture or light angle before looking at the material. A heavy overshirt, sheer blouse, glossy performance knit, black jersey, and white Oxford cloth do not fail in the same way. Standardize the base set, then create named exceptions for material risk.

Make color part of the capture record

Conceptual illustration: a black-shirt detail review under directional light. It is not a measured lighting proof or a customer product.

Color is not settled by making the background look white. Photograph a neutral or color reference under the final light whenever the source, modifier, room, or camera changes. X-Rite's ColorChecker guidance ties camera profiles to a specific camera, lens, and lighting condition, while Adobe's Lightroom tutorial shows white-balance presets and Temp/Tint controls for overall color adjustment. That makes the reference a record of the set, not a decorative prop.

Keep the unedited capture, the reference frame, and the approved master together. Compare the shirt with the physical sample under a controlled viewing light before approving hue, brightness, texture, or transparency. A customer's screen can still vary, but the production team should not add its own undocumented color drift.

For black shirts, preserve separation among the neckline, shoulder seam, body, and background instead of lifting everything into gray. For white shirts, inspect channel clipping and keep weave and edge detail visible. For fine stripes or small prints, inspect at full size for moiré, broken lines, and false color before resizing.

Shoot the evidence set in a fixed order

The minimum useful set depends on the shirt, but the sequence can stay stable: front identity, back, neckline or collar, sleeve or cuff, hem, fabric or print detail, and one fit or context view when the product needs it. Add a pocket, closure, inside finish, or care-detail frame only when that feature affects the buying decision.

Google's current image guidance separates the main product image from additional images and recommends variant-specific imagery, extra angles, close-ups, and clothing shown on people. Use that as a distribution constraint, not as permission to repeat nearly identical frames. Every image should answer a different question.

Review coverage before dismantling the set. Place adjacent colorways together and check crop, scale, collar height, sleeve angle, hem position, exposure, and background. Then inspect construction and texture at 100 percent. A missing back view is a capture failure. A shifted print is a product-truth failure. A shadow that changes every third SKU is a set-control failure.

Treat generated on-model imagery as context, not fit evidence

Once the factual shirt views pass, they can support additional campaign and on-model work. Look Atlas AI fashion photography begins with product imagery and lets a team direct model and visual styling. The safe handoff is the approved source set plus the five product-truth zones, not a single compressed thumbnail and a request to “make it editorial.”

The example below pairs a Look Atlas-owned folded black T-shirt source with a generated on-model context. It is a narrow demonstration, not a fit test or catalog-wide result.

First-party source and generated context: compare the visible crew neck and plain black front. The folded source hides the sleeves, hem and side seams; neither fit nor exact color is established.

The visible crew neck, black appearance, and unprinted front can be compared. The folded source hides the sleeves, full hem, side seams, fabric behavior, and real-body fit. The output proves none of those things, and different lighting prevents a color-accuracy claim.

Check every generated frame against the approved source at full size. Reject a changed neckline, sleeve length, print scale, button count, pocket, hem, color, sheen, or silhouette. Keep factual packshots as the primary product record. For a broader view of when flat lay, ghost mannequin, on-model, and lifestyle formats belong together, use the clothing photography for ecommerce guide.

If a generated image goes to Merchant Center, verify the current channel rules before export. Google's present guidance requires generated images to retain the appropriate IPTC digital-source metadata. Preserve provenance through resizing and delivery rather than stripping it from the final file.

Inspect difficult shirts at full size and in the catalog grid

A shirt can pass alone and fail beside its neighbors. Review the full-size file for seams, ribbing, buttons, print edges, dust, clipping, moiré, and retouch artifacts. Then view it at the size and crop customers will actually see. The second view reveals jumps in scale, background, shoulder position, and color temperature that a close inspection can miss.

Black, white, patterned, sheer, reflective, and heavily textured shirts deserve named exception recipes. Record the light and fill adjustment, the reference frame, and the reason for the change. Do not move the camera merely because one fabric is hard to expose. Preserve geometry first, then alter the smallest variable that solves the material problem.

Use the same two-view approval after retouching. At 100 percent, confirm the shirt still has believable texture and construction. In the catalog grid, confirm that the item reads as part of the same system. Retouching may remove temporary dust or a setup mark. It should not rebuild a collar, straighten a deliberately curved hem, or manufacture a colorway from another shirt.

Prove the system with a ten-shirt baseline

Before scaling, run ten shirts chosen to expose risk: a black tee, white Oxford, fine stripe, large print, embroidery, polo collar, cuffed button-up, sheer fabric, glossy performance knit, and an oversized silhouette. Photograph the complete evidence set, apply the five-zone review, and record every exception.

The baseline passes when another operator can recreate the framing, light, styling order, color reference, and release checks without relying on the original photographer's memory. If one shirt forces an unrecorded workaround, add the exception before the next batch. That is the difference between a successful shirt photo and a catalog-ready shirt photography method.