To check the result after using the convert PDF to grayscale tool, open the in-browser preview at 100% zoom, confirm the page count matches the source, scan each flattened page for contrast and readability, and only then download the raster PDF. The tool builds this review step directly into the workflow: every bounded page is rendered to a canvas, recomputed through the explicit per-pixel luminance formula 0.2126·R + 0.7152·G + 0.0722·B, written into a PNG, and assembled into a new PDF. The preview that appears before download is the same raster PDF you would otherwise have to open in a separate viewer, so the check happens without leaving the page. Reviewing at 100% zoom matters because the page dimensions are preserved while the visual detail is rendered at a fixed 1.5 scale and placed back on a page of the original size, so small text and thin lines are best judged at native resolution. Skipping the preview is the most common way people end up downloading a file where yellow or light blue content has dropped below readable contrast.
How the Tool Builds the Preview You Check Against
When you select a local PDF and trigger a conversion, the Grayscale PDF Converter loads the file with PDF.js, renders each page on an opaque canvas, and walks the pixels once. The grayscale value is computed per pixel using the Rec. 709 luminance weights — 0.2126 for red, 0.7152 for green, 0.0722 for blue — and the same integer is assigned to all three channels. That numeric output is encoded as a PNG page image, placed at the original page dimensions in a new pdf-lib document, and then the new document is reopened purely so the page count can be verified. The preview you see is therefore not a separate render: it is the actual output PDF rendered back through the same browser pipeline. That round-trip is what makes the check trustworthy — if a page is missing, faded, or out of order in the preview, the same problem will be in the downloaded file.
Check the Grayscale PDF Step by Step
The tool's three-step flow already sequences the check for you, but the inspection itself is a short routine you can repeat on every conversion.
- Pick the source PDF on your device and keep it under the 25 MB cap so the preflight check accepts it.
- Choose Convert to grayscale and wait while every bounded page is rendered locally.
- Open the preview and zoom to 100% on the first page; this is the size at which the 1.5-scale raster is mapped back onto the original page dimensions.
- Walk page by page, looking for missing pages, color areas that turned too light, or text that lost contrast.
- Confirm the page count in the preview matches the source document.
- Download the raster PDF only when every page meets your readability threshold.
What to Look For on Each Converted Page
The grayscale transform is faithful for the dark end of the spectrum and forgiving for near-black content, but it flattens mid-tone color highlights into a narrow gray band. The most useful checks on a converted page fall into a few repeatable categories.
- Yellow callouts, light blue links, and pale green highlights — these are the colors most likely to disappear against a white background.
- Thin rules and small body text — anti-aliasing softness becomes visible at 100% zoom because the visual detail is raster, not vector.
- Page numbers, headers, footers, and watermarks — confirm they survive the round-trip in the correct positions.
- Page order — if the source contained rotated landscape pages, confirm the rotation is preserved on the new page canvas.
- File size — a noticeably larger output is normal because PNG-encoded pages usually take more bytes than the source PDF's fonts, vectors, and image compression.
Verify the Grayscale Math on One Pixel
Because the formula is explicit and per-pixel, you can sanity-check the result with a single sample color. Take a pixel with red 200, green 100, blue 50 and apply the weights step by step.
- 0.2126 × 200 = 42.52
- 0.7152 × 100 = 71.52
- 0.0722 × 50 = 3.61
- Sum = 42.52 + 71.52 + 3.61 = 117.65
- Rounded to the nearest 8-bit value = 118
- Resulting pixel = RGB(118, 118, 118)
The same arithmetic runs across every pixel in the document, so the page you see in the preview is the page you will get after download. The browser pixel data used for this step is read through CanvasRenderingContext2D.getImageData, which is why the tool avoids relying on the CSS canvas filter property that Safari does not handle reliably as a default path.
What the Output Does Not Preserve
Knowing what the rasterization drops is part of how you check the result. The grayscale rebuild is visual only; the interactive and structural layers of the source PDF are not carried over.
| Feature in the source | Preserved in the grayscale PDF | How to check it |
|---|---|---|
| Selectable text | No | Open the downloaded file in a PDF reader and try to highlight a sentence; nothing should select. |
| Live links | No | Click a known link in the output and confirm nothing navigates. |
| Form fields | No | Look for input boxes or checkboxes in the preview; they will appear as flat artwork. |
| Annotations and comments | No | Open the annotations panel in your PDF viewer; the array should be empty. |
| Vector paths and curves | No | Zoom past 100% on a curve and watch for raster softening rather than smooth vector edges. |
| Original image compression | No | Compare file sizes; PNG-encoded pages are usually larger than the source PDF. |
| Page dimensions | Yes | Open page properties and confirm width and height match the source. |
| Visual layout of text and images | Yes | Eyeball each page in the preview for alignment and pagination. |
This is covered in more depth in Does Grayscale Conversion Keep Text and Links Selectable?, which walks through exactly which layers the raster rebuild drops.
When the Grayscale Result Is Not What You Need
The check you do in the preview is also the moment to decide whether the raster output is acceptable for your use case. The tool is the right pick for a quick visual monochrome copy, a print preview for a black-and-white printer, a reference document, or a grayscale copy for an upload requirement. It is the wrong pick when you need to keep selectable text, preserve accessibility tags, sign the file, meet PDF/A conformance, retain vector graphics, or control compression. In those cases the preview is a useful diagnostic — it tells you immediately that you need a different workflow rather than a different file, and you have not wasted an upload or a paid editor seat on a wrong path.
Limits and Failure Modes to Watch For
The preflight check that runs before any output canvas is allocated is the same gate the preview depends on, so knowing the limits is part of knowing what to look for when the result is missing pages or appears truncated. One file up to 25 MB is accepted, the job is capped at 40 pages, each rendered page may not exceed 12,000 pixels on either side or 40 megapixels, and the document as a whole may not exceed 100 megapixels. Password-protected, signed, malformed, or unusually complex PDFs may fail at the loading stage rather than at the grayscale step, which will show up as no preview rather than a poor one. Starting another conversion cancels the previous job, so the file you see in the preview is always the one tied to the conversion currently in progress. Keep your source file beside you while you inspect the output so that, if the preview does not match what you expected, you can compare the two side by side before deciding whether to re-run the job or move on.