
Affinity Photo 2's layered OpenEXR export lets you choose float precision separately for Image, Spatial and Other channels. Do not assume selecting a 32-bit document makes every exported pass full-float, or that a visually acceptable color preview proves a position pass is suitable for the next compositing operation.
Key Takeaways
- Image and spatial channels have separate export precision controls.
- Half-float and full-float are delivery choices, not descriptions of how pleasant the preview looks.
- Test data passes in the operation that consumes them.
List the passes before choosing a preset
The official EXR guide describes 16-bit half-float and 32-bit full-float choices for three channel classes in the export More dialog. It also documents channel affixes and an Include unknown channels option. Read the actual package requirements instead of inferring every channel's purpose from a rendered thumbnail.
A proposed architectural render might include color and XYZ spatial information for downstream treatment. The color plate is judged for its visible image, while a spatial pass must remain useful as data. Treating the second as a normal photograph can conceal a delivery failure.
Specify the receiving requirement
Ask which passes the next editor needs and whether their pipeline specifies precision. Preserve that specification with the export settings. If you do not know the purpose of a channel, do not silently classify it as unnecessary because the image looks unchanged when it is hidden.
Keep the source EXR and layered working file separate from experimental exports. Make one proposed candidate using the agreed precision settings. If a comparison is needed, change the relevant channel class while keeping other export decisions fixed.
Check in the consuming operation
- Export a separate layered EXR with the required channel classes present.
- Reopen it in the intended compositor and inspect the channel list.
- Verify that expected color, spatial and other components are recognized.
- Run a representative downstream operation using the data pass.
- Compare the result with the pipeline's approved reference or source data.
Do not claim a precision setting is sufficient merely because no artifact appears in the RGB view. Equally, do not insist that every pass must always use the largest representation. The choice depends on the recipient's actual operation and requirements; this guide provides no universal error tolerance or benchmark.
Accept a documented package
The handoff passes when the consumer finds the required channels and the selected precision supports the agreed use. Record which class used which float depth. If an unknown component is required, check that its export option and interpretation were verified rather than trusting its presence in Photo 2 alone. Keep the original package available for a corrected export if the recipient discovers a different precision need.
Related tasks
- How Do I Keep Extra ImageMagick Meta Channels Instead of Flattening Them Into RGB?
- Why Does a Transparent Blender Surface Appear in the Render but Not the Depth Pass?
- How Should I Choose PNG, EXR, or AVIF in Save Image Advanced?
Sources and scope
- 32-bit OpenEXR support: EXR More dialog separately chooses 16-bit half or 32-bit full float for Image, Spatial and Other channels; unknown channels may be included.
Directly read official Photo 2 help. Original documentation-based synthesis and proposed checks, not firsthand application testing. Existing and reserved titles screened; near-neighbor decisions are identified explicitly.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: William Liu / Pexels, Pexels License. This is an illustrative source photograph, not an application screenshot or a generated test result.