Photographing bags and shoes for an online shop, and how it differs from clothing

• The shape does not come from a body. Put clothing on a person and the real shape appears. A bag or a shoe already holds its own shape, so buyers look at other things entirely: the metal hardware, the clasp, the strap and the handle, the heel height, and the shape of the sole.

• Framing is a completely different problem. Standing up, a person's head and feet sit at opposite ends of them. A frame in which the shoe is large enough to read has no room left for a face. That is geometry, not something better prompt writing makes go away.

• The rule that is right for clothing is exactly wrong for shoes. Normally we instruct that a trouser hem should break over the top of the shoe, so the legs do not look pasted on. But when the shoe is the product, that same rule means half of it is under fabric.

In the Product to Model tool, the very first control is called Product type, and it offers eight choices: unspecified, clothing, footwear, bag, hat or eyewear, necklace, earrings, and ring or bracelet. The default is unspecified, which means the system reads the photo itself — that works, but the result is not the same every time. The moment you tell it footwear or bag, three things change at once: the framing the system selects for you, the pose the model takes, and the list of things that must never cover the product.

• Choose footwear and the framing is set to half-length for you. But half-length for footwear does not mean waist-up as it does for clothing — it means the system brings the model down to floor level, sitting, kneeling, crouching or reclining, so that her face and both shoes fit inside the same photograph without cropping her head away.

• If you pick a standing or walking pose card yourself, the system always obeys the pose you chose and crops from the waist down to the floor instead — because standing and seeing both the face and the shoes genuinely cannot happen at once.

• When the shoe is the product, the hem rule flips. The system instructs that the trouser hem must end above the ankle, so nothing drapes over the product.

• The verb changes too. With no product type set, the instruction says wearing or carrying, which leaves room for the model to hold the shoes in her hand rather than wear them. Once you press footwear, only one verb remains: worn on her feet. If the held look is what you wanted, go back to unspecified.

• Choose bag and half-length becomes mid-thigh up, so the whole bag and the hand or shoulder carrying it are both visible. The system instructs the model to carry it forward of her body — hanging from one hand at hip height, hooked over the forearm, or held by the top handle in front — never tucked behind her hip or pressed flat against her back.

• Bags get one more instruction that is sent every time regardless of the pose you choose: the front face of the bag, its closure and its hardware must be turned toward the lens rather than edge-on or away.

• Product to Model: yes, and it is the main route. 2 credits per image at standard quality, 5 at premium. It is the only tool with a Product type control.

• Image Set: yes, and it has the same Product type control. One press gives you 7 images — but the product type you chose is passed only to the ones with a model in them. One image in every set is a ghost mannequin shot, which does not receive it and means nothing for a bag or a shoe.

• Ghost mannequin: not for bags or shoes. The control on screen is called Garment category and offers only tops, bottoms and dresses, while the instructions behind it are written for six kinds of clothing: t-shirt, shirt, jacket, dress, skirt and trousers.

• Try-on: we do not recommend it. It takes two images, a person and a garment, with no text box and no product-type control at all — and our testing records contain no bags or shoes whatsoever. The test sheet we wrote covers clothing types only.

• Background removal, 1 credit per image: yes. It is not tied to a product type. But to be straight with you, we have not tested it systematically on other kinds of product.

• Upscale, 1 credit per image: yes. Product type is irrelevant to it, because it enlarges the image you already have rather than drawing a new one.

• Unspecified and clothing are set to full-length.

• Footwear, bag, hat or eyewear, and necklace are set to half-length.

• Earrings and ring or bracelet are set to close-up.

• All of this is a starting choice, not a lock. The framing control still offers all three options and you can change it whenever you like.

• The reason it is set for you is that the product reads more clearly, not that fewer jobs fail. We wrote that second reason down once and withdrew it the same day, because the figures behind it came from barely more than a dozen images — far too few to conclude anything.

• Product to Model: 2 credits per image at standard quality, 5 at premium.

• A model from the system library: plus 1 credit at standard, nothing extra at premium.

• Scene, pose and styling items counted together: plus 1 credit for 1 to 2 items, plus 2 for 3 to 5.

• Try-on: 2 credits per image at a single quality level. Ghost mannequin: 2 credits per image at standard quality and 4 at premium.

• Background removal: 1 credit per image. Upscale: 1 credit per image.

• Failed jobs are refunded automatically without you having to ask, but an image that generated successfully and simply did not please you is not refunded.

Signing up gives you 10 free credits, once, with no card required. With no extras selected that is 5 Product to Model images at standard quality — enough to answer whether the kind of bag or shoe you sell comes out well. One thing to know before you spend them: your first image on the free plan has no watermark, but every image after it carries our watermark burned into the file itself, not laid over the screen, so it cannot be removed and the image cannot be listed for sale. The free plan is there for judging quality, not for doing real work.