Selecting packshots

The single biggest lever for output quality is the packshot you feed in. This page covers what to pick, how to mark it in PackScene, and how to run multiple products in one batch.

What counts as a good packshot

PackScene works from product images. The cleanest input gives the cleanest output. Look for:

• High resolution. At least 1000x1000 pixels. PackScene will work with smaller, but textures, embroidery, and weave detail get lost.
• Full garment in frame. Sleeves, hems, collars, hoods, straps all visible. PackScene infers fit from the cut you show it.
• Even lighting, no strong shadows. Studio packshots photographed flat or on a ghost mannequin work best.
• Clean background. A pure white or pale grey background gives the most predictable model output. Cluttered backgrounds confuse the model.
• Wrinkle-free. Steamed and laid flat or hung properly. Folds get baked into the AI output as if they were design intent.

Good inputs that work very well: editorial flat lays, ghost-mannequin shots, dressform shots with a plain backdrop.

Inputs that struggle: angled shots, crops that hide hems or sleeves, dim phone photos, packshots with a hand or arm visible.

Marking the front packshot

On the Generate page, each product appears as a row with its existing Shopify image strip. To pick the front:

1. Hover over the image you want.
2. Click it once.
3. The image gets a green border and a FRONT badge.

Click the same image again to unselect, or click Undo under the product to clear all selections for that row.

Marking the back packshot

To pick a back, click a different image in the same product's strip after you've already picked a front. It gets a BACK badge. Back is optional - if you don't pick one, PackScene only generates a front model image.

When you do supply a back, PackScene generates the back after the front, using the front as a reference so the model identity stays consistent. This means the order of generation matters - back depends on front - and it's automatic.

Multi-product selection

You can mark front and back on as many products as you like in one go. Every selected product becomes one generation job in the run. Credit math:

• 1 product, front only, full body = 1 credit
• 1 product, front + back, full body = 2 credits
• 1 product, front only, full body + upper body = 2 credits
• 5 products, front only, full body = 5 credits
• 5 products, front + back, full body = 10 credits

Multi-product runs queue jobs in parallel. A run of ten products finishes in roughly the same wall-clock time as a run of one (limited by your plan's concurrency).

Choosing the right framing per product

Framing is selected once per run and applies to every product in the run. To get the right framing without over-spending credits:

• Upper body is best for tops, hoodies, jackets, bras, shirts, sweaters.
• Lower body is best for pants, skirts, shorts, jeans, swimwear bottoms.
• Full body is best for dresses, jumpsuits, coats, swimwear sets, anything that drapes top to bottom.

If a collection has mixed product types (tops + pants), run them in two separate runs with the matching framing each time. Mixing framings within a collection creates a messy storefront.

Setting product position

The Image position dropdown above the product list controls where PackScene-generated images get slotted into the Shopify product media. Two options:

• Before existing images - the AI photo becomes position 1. Your packshot moves to position 2. This is what most apparel stores want, because the model shot is the hero on the PDP.
• After existing images - the AI photo appends to the end. Use this when you want the packshot as the hero and the model image further down.

You can always reorder media later in Shopify admin if you change your mind, or use Easy Image Organizer to bulk-reorder across many products at once.

What's next

• Generating images - the actual generation step, plus credit costs.
• Best practices - tips beyond the basics.