Ghost mannequin photography: how to shoot it, what it costs, and when to put the garment on a model

A ghost mannequin is a mannequin with removable pieces used to photograph clothing: the garment is shot on it, the inside of the collar is shot separately, and the mannequin is edited out so the garment looks worn by an invisible body [1]. Ghost mannequin photography costs from $75 an image at a studio [2], or from about $0.50 an image if you shoot it yourself and send out only the editing [4][5]. When the product page also needs the garment on a person, Uwear puts a ghost-mannequin photo on a saved model, front, side and back, at 3 credits a photo (1K, default image model), with a credit at $0.10 [14].

The look is also called invisible mannequin, hollow man or the 3D mannequin effect [1]. It is a product-page staple with one catch: Amazon's 2021 clothing imaging guide asks for adult apparel on a model in the main image, and lists ghost mannequins among what that image must not contain [8]. The photos on this page are real Uwear runs. Uwear has no dedicated ghost-mannequin tool; it puts garments on models.

What is a ghost mannequin?

A matte white form with removable panels for the neck, chest, arms and hips [1]. The panels come off once the garment is on, so the camera can see inside the collar and the mannequin is easy to edit out against white. Retouchers rebuild what the mannequin hid with one of three joints: the neck joint for tees, shirts and knitwear, the sleeve joint for jackets and coats, and the bottom joint for dresses with an open hem [1]. A ghost mannequin costs about $300 to $1,300 [6][7].

How do you do ghost mannequin photography?

What you need

• A ghost mannequin, matte white so it does not reflect and edits out cleanly on white [1][15].
• A camera on a tripod. Photofixal suggests a 50mm or 85mm lens to avoid wide-angle distortion [1]; Photoroom's guide shoots on a phone [15].
• Two softboxes at 45 degrees, plus a light on the backdrop 1 to 1.5 stops brighter than the garment so the white goes pure white in camera [1].
• The camera at chest height for tops and hip height for bottoms. From higher or lower, the garment looks tapered [1][15].
• Manual settings that stay put for the whole session: f/8 to f/11, ISO 100 to 200, a fixed white balance set with a grey card [1].
• A steamer, pins, clips and tissue paper to shape the garment [1].

How to edit it in Photoshop

1. Open the outside and inside shots as layers in one file and auto-align them.
2. Cut the garment out with the Pen tool and mask out the mannequin.
3. Place the inside shot behind the garment, warp it to the curve of the neckline, and mask it so only what shows through the neck opening is left.
4. Paint a soft shadow inside the neckline on a Multiply layer, and put the brand label back.
5. Remove lint and pin marks, match the inside exposure to the outside with Curves, and export.

Those are the steps in Photofixal's workflow [1]. Apps such as Photoroom do the same with an eraser and an inserted cutout of the inside collar [15].

Can you do ghost mannequin photography without a mannequin?

Yes. Photoroom's guide shoots the garments on a person and erases the body in editing [15]. WearView's guide lists three routes with no form at all: a flat lay lightly stuffed with tissue paper and shot from above, a hanger with clear thread pulling the hem and cuffs into shape, and rolled towels or foam inside the garment for structure. Each still needs a separate photo of the inside of the collar [16].

The top at the head of this page started as a phone photo on a hanger. Uwear made the clean packshot and the photo on a model from it, without a mannequin or a composite.

What are the most common ghost mannequin mistakes?

• A flat neckline with no inside collar showing: it reads as a cutout, not a hollow garment [16].
• Light from one side on the garment and from the other on the pasted-in collar [16].
• Wrinkles left in. Deep folds cast shadows that look like damage, so steam first [16].
• Jagged cutout edges around the shoulders [16].
• A different scale and framing on every product. Tape the tripod, light and mannequin positions on the floor so each product is shot the same way [1][16].

How much does ghost mannequin photography cost?

Doing it in house also takes a camera, lights, a steamer and the time to dress, undress and shoot each garment three or four times.

Ghost mannequin vs on model: which should you use?

Most brands use more than one: on a model for the hero, ghost mannequin for the product page detail, flat lay for accessories [1]. The Amazon row comes from its Spring 2021 clothing imaging guide [8]. It is labelled for information only, so check the current rules in Seller Central before a shoot.

What does an AI ghost mannequin tool do?

It generates the ghost-mannequin image from a photo of the garment on a hanger, a model or a mannequin. Photoroom [10], WearView [11], The New Black [12] and Photta [13] offer one. The New Black lists where it struggles: sheer or lace fabric, complex open backs, and photos cropped so tightly that the hem or the shoulders are cut off [12]. Whatever the tool, compare the result with the garment: buttons, seams and the neckline are where generated images drift.

The difference is what you end up with. A ghost-mannequin tool gives you one thing, the ghost-mannequin image. Uwear puts the garment on a saved model, and the same run gives the packshot, the detail and the back view for listing photos. If the ghost-mannequin image is all you need, one of the tools above is the direct route. If the product page needs the garment on a person too, that is the gap the table shows.

How do you turn ghost mannequin photos into model photos?

Use the ghost-mannequin photo as the product photo. On Uwear, it goes on a saved model and is shot from the front, the side and the back; the shorts below were a hollow-man shot. The ghost mannequin to model page has the full set and the steps.

The set price is an estimate at published rates, not the recorded cost of the photos on this page; rates per image model are on the models page [14]. The estimate is shown before you approve a run.