DEV Community

Muhammad Omer Mirza
Muhammad Omer Mirza

Posted on

PNG to PDF without losing transparency: what most converters get wrong

You export a logo as a transparent PNG, drop it into a "PNG to PDF" converter, and the result has an ugly black box around it — or the transparency silently becomes white and ruins a design meant to overlay something. This is the single most common bug in PNG→PDF conversion, and it comes from one fact people forget:

A PDF page is not transparent. When you print or flatten it, "transparent" has to become some colour.

So the real question isn't "does it keep transparency?" — it's "what should the transparent pixels become, and who decides?" Good converters make that a choice. Let's build one client-side.

The library: pdf-lib

pdf-lib creates and edits PDFs in pure JavaScript — in the browser, no server. It can embed PNGs with their alpha channel:

<script src="https://cdnjs.cloudflare.com/ajax/libs/pdf-lib/1.17.1/pdf-lib.min.js"></script>
Enter fullscreen mode Exit fullscreen mode
const doc = await PDFLib.PDFDocument.create();
const png = await doc.embedPng(pngBytes); // alpha preserved
const page = doc.addPage([png.width, png.height]);
page.drawImage(png, { x: 0, y: 0, width: png.width, height: png.height });
const pdfBytes = await doc.save();
Enter fullscreen mode Exit fullscreen mode

That "alpha preserved" is the trap. If you embed a transparent PNG directly, the transparent areas show whatever ends up behind them — which, on a white-printed page, is fine; on a coloured workflow, is not; and in some viewers, renders as black. You want to decide on purpose.

The fix: offer three modes

Give the user (or yourself) three options:

  1. Flatten onto white — the safe default for anything you'll view or print.
  2. Flatten onto black — for light artwork meant for dark backgrounds.
  3. Keep transparency — for PDFs that will be layered over another design.

"Flatten" just means: paint a background colour, draw the PNG on top, re-encode. A <canvas> does this in a few lines:

// Returns PNG bytes, flattened onto `bg` ('white' | 'black' | 'keep')
async function preparePng(file, bg) {
  if (bg === "keep") {
    return new Uint8Array(await file.arrayBuffer()); // embed alpha as-is
  }
  const img = await loadImage(file);              // see helper below
  const canvas = document.createElement("canvas");
  canvas.width = img.naturalWidth;
  canvas.height = img.naturalHeight;
  const ctx = canvas.getContext("2d");

  ctx.fillStyle = bg === "black" ? "#000" : "#fff";
  ctx.fillRect(0, 0, canvas.width, canvas.height); // background first
  ctx.drawImage(img, 0, 0);                         // PNG on top

  const blob = await new Promise((r) => canvas.toBlob(r, "image/png"));
  return new Uint8Array(await blob.arrayBuffer());
}

function loadImage(file) {
  return new Promise((resolve, reject) => {
    const url = URL.createObjectURL(file);
    const img = new Image();
    img.onload = () => { URL.revokeObjectURL(url); resolve(img); };
    img.onerror = () => { URL.revokeObjectURL(url); reject(new Error("bad png")); };
    img.src = url;
  });
}
Enter fullscreen mode Exit fullscreen mode

Order matters: fill the background, then draw the image. Do it the other way and you paint over your own picture.

Putting it together (multiple PNGs → one PDF)

async function pngsToPdf(files, { bg = "white" } = {}) {
  const doc = await PDFLib.PDFDocument.create();
  for (const file of files) {
    const bytes = await preparePng(file, bg);
    const png = await doc.embedPng(bytes);
    const page = doc.addPage([png.width, png.height]); // fit page to image
    page.drawImage(png, { x: 0, y: 0, width: png.width, height: png.height });
  }
  return doc.save(); // Uint8Array — download or zip it, still no server
}
Enter fullscreen mode Exit fullscreen mode

Why not just convert PNG→JPG first?

Because PNG is lossless and often holds screenshots, diagrams, and text — exactly the content JPG compression smears. Embed the PNG and you keep crisp edges at full quality. Flattening onto a solid colour doesn't recompress the visible pixels; it only resolves the alpha.

Quick reference

Mode Transparent pixels become Use when
White white normal viewing/printing (default)
Black black light logos on dark layouts
Keep whatever's behind in the viewer PDF will overlay another design

The whole thing runs in the browser, so nothing uploads and there's no backend to pay for.


I shipped this as a free tool — PDFNest's PNG to PDF converter has exactly this white/black/keep transparency control, runs client-side, and adds no watermark. There's a sibling Image to PDF tool for mixed JPG/PNG/WebP. Questions about the pdf-lib/canvas details — drop them below.

Top comments (0)