Edit Reflective Products in 16-Bit, Deliver Them in 8-Bit

Use 16-bit editing when a product photo needs substantial colour, exposure or gradient work. Export an 8-bit delivery copy when the image is finished. That split gives the editor more tonal values while adjustments are still being calculated, without burdening a product listing with a large working file.

The difference matters most around polished metal, glossy packaging, soft shadows and plain studio backgrounds. Those areas contain gradual tonal transitions. Push an 8-bit file hard enough and the transition can break into visible steps called banding or posterisation. For a lightly corrected phone JPEG, however, an 8-bit workflow may be entirely adequate. More bits are useful headroom, not an automatic quality upgrade.

What 8-bit and 16-bit actually change

Bit depth describes how many tonal values a file can store in each colour channel. Adobe’s current Photoshop bit-depth guide lists 256 values per channel for 8-bit images and 65,536 for 16-bit images. The practical benefit appears during editing: more intermediate values reduce the risk that repeated corrections leave gaps between tones.

That does not mean a shopper will see 65,536 separate steps on a normal screen. The extra precision mainly protects calculations while you adjust the photograph. Once the image looks right, a carefully made 8-bit output can display smoothly and load efficiently.

There is a cost. A 16-bit RGB document carries roughly twice the pixel data of an 8-bit RGB document before layers and compression are considered. Large catalogues consume more storage, require more memory and can feel slower to retouch. Some specialised filters may also require an 8-bit copy.

Run the comparison on one real product photo

This is a workflow comparison, not a benchmark of a particular camera or editing app. You can reproduce it with any editor that supports both modes. Start with a RAW file or a genuine 16-bit TIFF so the higher-bit version contains real source information.

1. Open the same source twice. Keep one document at 16 bits per channel. Convert the second to 8 bits per channel before making any adjustments. In Photoshop, the control is under Image, Mode.

2. Apply the same non-destructive adjustments to both files. Use identical Curves, Levels, white balance and colour corrections. If the product sits on a gradient background, keep that gradient identical too.

3. Make the test demanding but relevant. Lift a dark metal area, recover bright reflections, correct a colour cast and smooth the transition from the product shadow into the background. Avoid random filters that would never appear in your actual workflow.

4. Inspect both files at 100 per cent. Look for visible bands in the background, abrupt changes across polished metal, broken fall-off in soft shadows and rough colour transitions around gemstones or enamel.

5. Convert a copy of the 16-bit result to 8-bit, place both final files in sRGB and export them at the same dimensions and JPEG quality. Compare the delivery files, not only the working documents.

Where 16-bit earns its larger file size

Polished jewellery is a strong case because metal is described by gradients rather than one flat colour. A silver bangle may move from white reflection to mid-grey and back to a dark edge across a few millimetres. Strong Curves or colour work can make those transitions brittle in an 8-bit working file.

The same is true when you combine separate exposures. A texture-pass workflow for embossed product detail or a restrained exposure bracket for chrome, black and white products asks the editor to blend smooth tonal information. Keeping the master at 16-bit gives those masks and adjustments more precision.

Gradient backdrops are another obvious stress point. Adobe’s guidance on avoiding gradient banding recommends increasing colour depth to 16-bit where possible. Subtle noise or dithering can help a final 8-bit export, but it should not replace a higher-precision master when heavy editing is planned.

Repeated colour corrections also benefit. Jewellery sellers may need to neutralise a warm reflection, protect an enamel hue and keep the surrounding metal believable. The edits still require a colour reference and visual checks. Bit depth prevents some tonal damage, but it cannot tell you whether the final gold or gemstone colour is accurate.

When 8-bit is the sensible choice

Stay in 8-bit when the source is already a finished JPEG, the exposure and white balance are close, and the edit is limited to a crop, a small brightness correction and light sharpening.

An 8-bit workflow also makes sense for quick listing graphics that use flat colours, text and simple cutouts rather than delicate photographic gradients. If the image is already approved and only needs resizing for another marketplace, converting it to 16-bit adds complexity without recovering information that is missing from the source.

This last point matters: changing an 8-bit JPEG into a 16-bit document does not recreate the tonal values discarded earlier. It may give later calculations more room, but it cannot rebuild blown highlights, clipped shadows, erased texture or banding that is already baked into the file.

Do not confuse bit depth with colour space

Bit depth and colour space solve different problems. Bit depth controls the number of tonal values available for editing. A colour space such as sRGB or Display P3 defines the range and interpretation of colours. The sRGB versus Display P3 product-photo workflow explains why a high-quality master can still shift colour when it is exported or displayed incorrectly.

A practical product master can therefore be 16-bit and use an appropriate working colour space, while the listing copy is converted to sRGB and 8-bit. Check the final file after conversion because colour-profile changes, bit-depth reduction and JPEG compression can each affect smooth areas.

How AI changes the decision

If an AI editor returns only an 8-bit rendered image, converting that output to 16-bit later will not restore missing source information. Keep the pre-AI master at its original quality, treat the generated result as a derivative and avoid stacking aggressive colour edits after generation.

For product fidelity, bit depth is secondary to geometry and material accuracy. A perfectly smooth 16-bit gradient is still wrong if AI changes a clasp, label, stone count or surface finish. Compare the output with the real SKU before judging polish.

When the AI step changes only the scene around a preserved product cutout, keep the higher-bit master and masks separately. That makes it easier to rebuild the background or shadow without repeatedly processing the product pixels.

The working rule

Edit in 16-bit when you start from RAW or a true high-bit source and expect strong tonal, colour, masking or compositing work. Keep that layered master. When the image is approved, create an 8-bit sRGB copy and export it at a sensible quality. Adobe’s JPEG documentation confirms that Photoshop converts 16-bit documents to 8-bit when saving as JPEG.

Use 8-bit from the start when the source is already a good JPEG and the corrections are minor. The best workflow is not the one with the largest file. It is the smallest workflow that preserves smooth tones, accurate colour and the real product through the edits you actually need.