Quick context
For product photography, choose a format that preserves the required quality and works with the commerce platform's current delivery pipeline.
The most useful way to approach ecommerce image formats is to separate four questions: What does the destination require? What property of the image is actually wrong? What is the smallest change that fixes it? How will you verify the result? That prevents unnecessary conversion, repeated recompression, and accidental loss of metadata or quality.
What current first-party documentation says
- Shopify Help Center: Shopify documents supported product-image formats and current product-media image requirements.
- Shopify Help Center: Shopify says storefront images are optimized by its CDN, recommends uploading high-quality source images, documents format delivery behavior, upload limits, and color-profile handling.
- MDN Web Docs: MDN documents practical characteristics of JPEG, PNG, WebP, AVIF, transparency, and web image format trade-offs.
Those details are time-sensitive. This draft records them as verified on October 3, 2026 and should be rechecked before publication if the platform changes.
Choose the right method
- Keep the original format. Best when when the destination already accepts it. The main trade-off is may not solve a compatibility requirement. Avoid conversion without a reason
- Convert format. Best when compatibility, transparency, or delivery requirements. The main trade-off is can change file size or lose features. Conversion cannot restore missing source detail
- Test the destination. Best when confirming real browser/app behavior. The main trade-off is requires an extra verification step. Prefer measured output over format folklore
The right choice is the one that changes the property the destination actually cares about. File size, pixel dimensions, format, transparency, metadata, and provenance are different things.
Step-by-step workflow
- Keep the original. Work from a copy so you can retry without stacking losses or deleting useful metadata.
- Read the destination requirement. Record any maximum file size, required format, pixel dimensions, aspect ratio, transparency requirement, or privacy constraint.
- Inspect the source. Check its current file size, width and height, format, and any metadata or provenance relevant to the task.
- Change one thing first. If the problem is only file size, start with compression. If the destination needs fewer pixels, resize. If it requires another format, convert. If privacy is the concern, inspect and remove metadata separately.
- Use Recordly Image as the task-specific tool. Review the output instead of assuming an export succeeded just because a download was produced.
- Verify the exported copy. Reopen it and compare the property that matters: size, dimensions, transparency, metadata, color, edge quality, or provenance.
- Test the real destination. Upload or preview the final copy where it will actually be used.
What to inspect closely
- Text and sharp edges: screenshots, scans, logos, and UI captures can reveal compression artifacts faster than photographs.
- Faces and fine detail: inspect eyes, hair, fabric, product texture, and small printed text.
- Transparency: place transparent exports over both light and dark backgrounds to expose halos.
- Metadata and privacy: a visual edit does not guarantee metadata removal.
- Platform processing: some destinations resize, compress, convert, strip profiles, or change presentation after upload.
- Authenticity claims: a model score or missing metadata should never be rewritten as proof.
Common mistakes
- Starting with a generic “high quality” preset instead of the destination requirement.
- Changing format, dimensions, and compression at the same time, then not knowing which change caused a problem.
- Recompressing an already-small file just because a target number exists.
- Assuming a filename extension guarantees what is actually encoded inside the file.
- Deleting the original before testing the exported result.
- Copying old platform limits from a third-party article without checking current first-party documentation.
Image brief for the editor
Use the structured image briefs attached to this CMS entry. Prefer current Recordly UI screenshots, original annotated diagrams, or before/after examples made from owned or licensed images. Do not use a fake benchmark, fabricated detector score, or invented platform screenshot.
Final verification checklist
- The image opens correctly.
- The final format is accepted by the destination.
- The real byte size and pixel dimensions meet the actual requirement.
- Important text, faces, products, or fine edges remain readable.
- Transparency behaves as expected.
- Metadata or provenance has been intentionally preserved or removed.
- The original source still exists.
- Any time-sensitive platform fact has a current first-party source.
When to use another Recordly tool
If the remaining problem is different from the one this guide solves, switch operations rather than forcing the same tool. For example, compression and resizing are related but not interchangeable; metadata cleanup is separate from pixel editing; and provenance checking is separate from AI-image detection.
The related guide best image format for websites covers the adjacent concept in more depth.