When a removal does not go the way you expected

Most problems on BOARD fall into one of a few buckets: the upload does not process, the app misses the object you want, the first result looks incomplete, or you run into a credit wall sooner than expected.

This troubleshooting page is intentionally narrow. It is not a generic software support page; it is a guide to the failure patterns that show up most often in BOARD's current removal workflow. In practice that usually means upload friction, missed detection on tiny or low-contrast targets, a first result that needs one more pass, or confusion about where the starting 5 guest credits went. If you can identify which bucket you are in, the recovery path is usually short.

Written by the BOARD team at Rainn Inc. and reviewed against the live workflow on March 16, 2026. Questions: support@rainn.ai.

If the upload fails

  • Confirm the file is an image and stays within the current size guidance.
  • If the file is HEIC or HEIF, export a JPG or PNG copy and retry.
  • Refresh the page and upload again before assuming the image itself is unusable.

If BOARD misses the object

  • Start with a clearer crop where the object is larger in frame.
  • Use a version of the photo with better lighting or less compression.
  • Try a different image if the target is extremely small or blended into the background.

If the first removal looks off

  • Use another pass before discarding the photo entirely.
  • Check the original-image compare view to see whether the issue is truly noticeable at normal viewing size.
  • Very busy patterns, reflections, and shadows are the most common hard cases.

If you run out of credits

  • Guest users can sign in to keep editing and move into an account-based balance.
  • If you already have an account, use the credit balance in the top bar to open the purchase flow.
  • Save credits for the jobs where BOARD is strongest: clear, specific objects in otherwise usable photos.
Good removal candidates are specific and visible.

BOARD tends to do best on obvious photobombers, labels, date stamps, signs, wires, and clutter. Tiny details hiding inside a patterned background are the most common failure case.

Fast recovery checklist

  1. 1
    Retry the same job once

    Many near-miss removals only need one more pass.

  2. 2
    Switch to a cleaner crop or a sharper source

    Detection improves when the target is larger and more separated from the background.

  3. 3
    Use a simpler example to confirm the workflow

    If a hard image fails, test BOARD on an easier removal so you can separate product fit from file-specific problems.