Source: https://restorebunny.com/guides/test-scanner-automatic-corrections/ # Test Scanner Automatic Corrections Before a Family Photo Batch Compare baseline scans with one scanner correction at a time so colour restoration, sharpening and cropping do not quietly change your archive master. Scanning & preservation4 min readUpdated 11 October 2026 ## Try it on your own photo Have your scan ready? Repair a copy, then compare faces and details with the original. [Restore my photo]()1 credit per successful photo ยท Account required[See credit packs]() Quick answer ## The short version Make a baseline scan with optional corrections documented, then change one feature at a time and compare the saved files at matching scale. Keep a minimally altered capture master and separate corrected viewing copies. A feature name such as colour restoration does not prove historical accuracy. ## Treat scanner processing as an editing choice A scanner can apply more than resolution and file-format settings. Automatic exposure, sharpening, colour correction, dust treatment and edge detection can affect the pixels or crop before you open an editor. For the Epson V600 software documented in the source below, thumbnail preview includes automatic operations and several corrections have separate controls. Other scanner models and newer applications can behave differently; use the documentation for the software you actually run. Begin with a settings record, not an assumption that an unchecked-looking image is untouched. Write the scanner, application version, document type, output mode, resolution and correction controls. Some automatic processing may remain part of the normal capture pipeline. Describe your baseline accurately as the least-altered practical capture you selected, rather than claiming it is a raw measurement of the print. ## Build comparisons that answer one question Use three representative sound photographs: a face with fine hair, a low-contrast image and one with visible paper or print pattern. Save a baseline for each. Make separate captures with only one optional correction changed, leaving other settings and crop constant. If the software resets options after preview mode changes, verify the record before the final scan. Compare the actual files at equal displayed size. A larger output can look sharper simply because it is shown differently. Check eyes, hair boundaries, handwriting, bright fabric and dark clothing. Return to the whole image too: a correction that is helpful close up may produce an excessive colour shift or a flattened appearance at normal size. ## Worked example: sharpening helps a sign but harms hair A fictional pilot has three baseline scans and three sharpened scans. On the sign, edge contrast improves readability. On the portrait, a bright fringe appears beside dark hair. On the textured print, the paper pattern becomes more conspicuous. The evidence supports separate decisions for those pictures, not a rule that the same sharpening box should be enabled for the full collection. Keep the baselines and choose a restrained viewing derivative for the sign if it serves the task. Do not bake the strongest result into every master. Likewise, an automatic colour correction that makes grass greener does not verify the grass's original colour. A historically important colour needs corroborating evidence beyond the scanner's appearance preference. ## Run a controlled correction pilot 1. Record the baseline settings and capture filenames. Inspect each full border and the saved pixel dimensions so crop changes cannot hide inside a colour comparison. 2. Enable one optional correction, save a separate output and write exactly what changed. Avoid changing resolution, crop and several filters in the same trial. 3. Compare vulnerable features against both the baseline and the safely accessible original. Use matching zoom and keep brightness differences from being mistaken for extra detail. 4. Record keep, optional viewing copy or reject for each setting and image. Describe the observed effect, not an unsupported percentage quality improvement. 5. Choose batch defaults conservatively, retain masters and add exception rules. Repeat the pilot when the material or scanner application changes substantially. ## Retain enough history to undo the decision A selected correction may be useful for family sharing while being inappropriate for a research reference. Store a corrected export beside, rather than over, the baseline. Keep the setup record with the batch so the next editor knows whether a halo or colour shift began in scanning or later restoration. Do not assume a preview represents every final operation. The Epson documentation specifically distinguishes some final-scan effects from preview behaviour. Reopen representative final files before accepting the batch. If a feature has film-specific restrictions, follow the relevant model documentation rather than borrowing a print workflow. This test establishes your selected files' suitability for a stated task; it is not a universal scanner benchmark or a claim of authentic missing detail. ## FAQs Should all automatic corrections be disabled? Choose and document a baseline, then test optional processing. Some pipeline operations may remain; avoid calling it raw unless that is supported. Does colour restoration recover the original colours? The feature can change appearance. Its name is not evidence of historical colour accuracy. Can I judge only the preview? Inspect saved final scans. Some documented effects appear only in the final output, and crops or settings may change. ## Sources Preservation and technical guidance reviewed for this article. - Epson: [Epson Scan detailed setting instructions for Perfection V600]() - Adobe: [Sharpening overview]() ## Related - [Choose the underlying capture settings]() - [Diagnose export colour differences]() - [Calculate scan pixels from the print size]() ## Ready to restore your photos? Upload an image, choose your settings, and preview results with a before/after slider. [Restore my photo]()[View Pricing]() RestoreBunny Team Photo workflow editors Guides follow our [editorial policy]() for sourcing, product claims, AI limitations, updates, and corrections.