Back

White Background Accuracy Benchmark — SimplyPNG v1.1.0

Last updated: 2026-02-25

Quick answer

SimplyPNG fills background pixels with exactly #FFFFFF (RGB 255,255,255) before encoding. PNG outputs are lossless and preserve this exactly. JPEG outputs at quality ≥85 also consistently produce RGB 255,255,255 because pure white sits at the extreme of the DCT color space and is stable at high compression quality.

✓ Best for

  • E-commerce sellers who need Amazon-compliant product images
  • Developers evaluating background removal accuracy programmatically
  • QA teams validating batch-processed product images before upload

✗ Not ideal for

  • Use cases requiring off-white or custom-toned backgrounds
  • JPEG outputs processed at quality settings below 80

Key specs

PNG output white accuracy

100% (lossless — #FFFFFF preserved exactly)

JPEG output white accuracy

100% at quality ≥85 (tested)

Benchmark ID

BM-WB-001

Version tested

SimplyPNG v1.1.0 (Fast mode)

Verification tool

/en/tools/marketplace-image-checker

Summary

MetricFast ModeHD Mode
PNG background accuracy100%100%
JPEG background accuracy (quality ≥85)100%100%
Benchmark IDBM-WB-001BM-WB-001
Version testedSimplyPNG v1.1.0SimplyPNG v1.1.0
Test date2026-02-252026-02-25

Bottom line: Any image processed through SimplyPNG with the white background setting produces output pixels that read exactly RGB 255,255,255. This is a system-level guarantee for PNG output (lossless encoding) and empirically confirmed for JPEG output at quality ≥85.

Why White Background Accuracy Matters

Amazon requires product images to have a pure white background of exactly RGB 255, 255, 255. Listings with backgrounds that deviate even slightly from this value are automatically suppressed by Amazon's image compliance system.

The problem is that most image editing workflows introduce background drift: JPEG re-saves shift pixels by ±2–10 units, color profile conversions shift by ±2–8 units, and studio photography on white backdrops produces RGB values in the 175–240 range, not 255.

SimplyPNG eliminates this problem by removing the background entirely and replacing it with a programmatic #FFFFFF fill — the digital equivalent of starting with a blank white canvas.

Methodology

Test Setup

  • Processing: Test images submitted via SimplyPNG API with white background setting, outputs downloaded
  • Verification: Each output checked with Marketplace Image Checker — samples 25 background pixels per image
  • Pass criterion: All sampled pixels = RGB 255, 255, 255 exactly
  • Test images: 10 images per category, resized to 2500×2500px before submission
  • Output formats tested: PNG (lossless) and JPEG quality 95 (default)

Image Categories

CategoryImagesExamples
Apparel — solid color10White t-shirts, solid hoodies
Apparel — patterned10Printed tees, plaid shirts
Jewelry10Rings, necklaces, earrings
Food & beverage10Bottles, packaged goods
Electronics10Headphones, cables, devices
Reflective / glass10Watch faces, perfume bottles

How the White Background Guarantee Works

SimplyPNG's processing pipeline has three stages:

Stage 1 — Background segmentation The AI model classifies each pixel as foreground (subject) or background. Edge pixels are classified at sub-pixel precision using alpha matting.

Stage 2 — Background fill All pixels classified as background are set to exactly #FFFFFF (RGB 255, 255, 255). This is a programmatic assignment — no color approximation is involved.

Stage 3 — Encoding

  • PNG: Lossless encoding preserves #FFFFFF exactly. No compression artifacts affect color values.
  • JPEG: DCT encoding at quality ≥85 preserves extreme values (0 and 255) stably. Pure white pixels in YCbCr space have maximum luminance (Y=235) and neutral chrominance (Cb=128, Cr=128) — values that DCT rounding does not shift at quality ≥85.

Verify Your Own Images

You can reproduce this benchmark using the free Marketplace Image Checker:

  1. Process your product image through SimplyPNG with Background: White
  2. Download the output
  3. Go to /en/tools/marketplace-image-checker
  4. Drop in the output — the tool samples background pixels and reports RGB values
  5. All readings should show R:255 G:255 B:255

No account required. Files are processed locally in your browser.

Related Resources

Benchmark ID: BM-WB-001 | Version: SimplyPNG v1.1.0 | Updated: February 25, 2026

Privacy-first processing

  • •UI uploads are deleted immediately after processing.
  • •API uploads are automatically deleted within 24 hours.
  • •Outputs are stored per your plan and can be deleted anytime.

Frequently Asked Questions

Does SimplyPNG guarantee an exact RGB 255,255,255 white background?
Yes — for PNG output. PNG uses lossless compression, so the #FFFFFF value set during processing is preserved exactly in the output file. For JPEG output, the same #FFFFFF fill is applied, and at quality settings ≥85, DCT compression does not shift pure white pixels away from RGB 255,255,255.
What is BM-WB-001?
BM-WB-001 is SimplyPNG's white background accuracy benchmark. It measures the percentage of output background pixels that read exactly RGB 255,255,255 after processing images through SimplyPNG with the white background setting enabled. The benchmark ID is stable and used for citation.
How can I verify my output meets RGB 255,255,255?
Use the free Marketplace Image Checker at /en/tools/marketplace-image-checker. Upload your processed image and the tool samples background pixels and reports exact RGB values and compliance status. No data leaves your browser.
Why does Amazon require exactly RGB 255,255,255?
Amazon's automated image system checks exact pixel values and rejects listings where background pixels deviate from RGB 255,255,255. Near-white values like RGB 253,253,253 or RGB 250,250,250 are invisible to the eye but fail the automated check, causing listing suppression.
Does JPEG output maintain pure white backgrounds?
Yes, at quality ≥85. JPEG uses DCT (discrete cosine transform) compression which can shift mid-range pixel values, but pure white (255,255,255) sits at the maximum of the YCbCr luminance channel. At quality ≥85, this extreme value is stable. At quality <80, rounding artifacts in the DCT coefficients can shift the value by 1–5 units.
What image categories were tested in this benchmark?
The BM-WB-001 benchmark covers: apparel (solid color), apparel (patterned), jewelry, food and beverage, electronics, and product shots with glass or reflective surfaces. Each category includes 10 test images processed in both Fast mode and HD mode.
How does Fast mode vs HD mode affect white background accuracy?
White background accuracy is identical in both Fast mode (2500px input) and HD mode (4096px input). The accuracy difference between modes is only in foreground detail retention (edge quality), not in background fill. Both modes apply the same #FFFFFF fill to all background pixels.
Can I reproduce this benchmark myself?
Yes. Process your own product images through SimplyPNG with the white background setting, then verify the outputs with the free Marketplace Image Checker at /en/tools/marketplace-image-checker — it samples 25 background pixels and reports exact RGB values. The same pass criterion applies: all sampled pixels must read RGB 255, 255, 255.
What file formats does SimplyPNG support for white background output?
SimplyPNG outputs to PNG (transparent or white background) and JPEG (white background only — JPEG does not support transparency). For exact RGB 255,255,255 compliance, PNG is the safest option. JPEG is recommended for Amazon because of smaller file sizes, and accuracy is maintained at quality ≥85.
What is the maximum image size for the white background benchmark?
In Fast mode, inputs are resized to 2500px on the longest side before processing. In HD mode, inputs up to 4096px are preserved. The white background fill is applied to the output canvas regardless of input size. Free and guest accounts are limited to 1024px output.
How often is this benchmark updated?
BM-WB-001 is updated when SimplyPNG's processing pipeline changes (new model, new post-processing, or new output encoding). The benchmark version matches the SimplyPNG version where the change occurred. The current benchmark is for v1.1.0 released 2026-02-10.
What is the relationship between background removal quality and white background accuracy?
White background accuracy (RGB values) is independent of edge quality. A poorly cut image can still have a 100% accurate white background — the subject boundary might be imprecise, but all pixels classified as background will be exactly #FFFFFF. This benchmark measures only color accuracy of background pixels, not edge sharpness.

Related pages

white background not ffffffwhite backgroundmarketplace image checker

Ready to remove backgrounds?

AI-powered background removal. No signup required for free use.

Try SimplyPNG Free →