PDF to PNG renders every page of a document into a lossless PNG image inside your browser, at a resolution you choose between 96 and 600 DPI. The file is parsed locally by a self-hosted copy of pdf.js, so nothing is uploaded and no copy of your document sits on a server waiting for a retention timer to expire. You pick the resolution, the tool states the exact pixel dimensions you will get, and the pages come back as PNG files or as a single ZIP.
What DocPivot PDF to PNG Does
The tool converts each page of a PDF into a PNG image at a resolution you control, and it can also pull out the photographs stored inside the PDF instead of rasterizing the pages around them. Those are two different jobs sharing one tool name, and most converters only do the first. Rendering produces a picture of the page including its text, rules, and margins. Extraction reaches into the file structure and returns the embedded images at their native resolution.
Everything runs on your device. DocPivot parses the file with pdf.js version 6.2.108, vendored and self-hosted along with its worker, character maps, and standard font files. There is no content delivery network dependency and no API call, so the tool keeps working after the page has loaded even if your connection drops, and there is no upload step to wait through on a large document.
Page edits happen before conversion, not after. You can rotate a page, delete one, reorder the set, or keep only a chosen range, and the renderer reads that state at export time. A page you rotated comes out rotated, and a page you deleted does not come out at all. If your goal is only to reorder or drop pages and keep the document as a PDF, Organize PDF and Remove Pages from PDF do that without conversion, and Rotate PDF writes a rotation back into the file itself.
How DocPivot PDF to PNG Works
Conversion runs as a five stage pipeline, and every stage executes locally.
- Open locally. Page dimensions are captured at load time so the panel can state real output sizes before anything is rendered. Encrypted files are detected and refused with a specific message rather than a generic failure.
- Edit first, convert second. Rotate, delete, reorder, and keep-only-these operate on a page list rather than on the file. Rendering happens at export time, so your edits are simply the state the renderer reads.
- Render onto an opaque canvas. Scale is DPI divided by 72. The canvas is created without an alpha channel and filled white before drawing, so a PDF that paints no background produces white paper instead of transparency. Page rotation composes with the page's own rotation flag.
- Encode with the in-house PNG writer. Pixels are written to PNG by a purpose-built encoder rather than by the browser's canvas export, which produces a noticeably smaller file for identical image data.
- Hand back one download. Filenames are zero padded, so report-03.png sorts correctly next to report-11.png. Multiple pages are packed into a ZIP that stores rather than deflates, because PNG data is already compressed.
The result panel reports page count and total size, then points toward the next step. Common follow-ups are Image Compressor for further size reduction and Image Resizer for fitting output to a specific pixel target.
Choosing the Right DPI for PDF to PNG
Resolution is offered as four fixed steps rather than a free-text field, because those four cover almost every real use and each has a predictable output size. The figures below were measured on one page of a real 22 page document at US Letter dimensions, so treat them as a guide to the ratios rather than a promise about your own file.
| Setting | DPI | Output at US Letter | PNG size per page |
|---|---|---|---|
| Screen | 96 | 816 by 1056 px | 166 KB |
| Good (default) | 150 | 1275 by 1650 px | 220 KB |
| 300 | 2550 by 3300 px | 513 KB | |
| Maximum | 600 | 5100 by 6600 px | 1,368 KB |
The whole 22 page document at 150 DPI produced 4.7 MB of PNG files in 1.17 seconds. Render time stays close to flat across the four settings because pdf.js draws vector instructions rather than resampling a bitmap. What grows with resolution is encoding time and output size, not drawing time.
Choose 96 DPI for thumbnails and screen-only viewing, 150 DPI as the default for slides and documentation, and 300 DPI when the image will be printed or zoomed into. Reserve 600 DPI for floor plans, blueprints, and archival scans, where thin lines and small annotations start disappearing at 300. DocPivot states the exact pixel dimensions for your document's real page box first, so an oversized sheet does not surprise you with a file ten times larger than expected.
Two Modes Behind One Tool Name
People looking to get images out of a PDF are asking one of two questions, and DocPivot answers both. Every Page renders the pages themselves. Just the Pictures extracts the photographs embedded inside the file.
| Mode | What it produces | When it is the right choice |
|---|---|---|
| Every Page | One image per page, with text, rules, and margins included | Archiving, printing, slide decks, sharing a page as a picture |
| Just the Pictures | The photographs stored inside the PDF, at their own native resolution | Product catalogs, scanned photo albums, reports containing charts |
Extraction walks the PDF operator list, collects image objects, and deduplicates them by name, so a logo appearing in a running header across 40 pages is saved once instead of 40 times. Anything smaller than 32 pixels on either edge is skipped, which filters out bullet glyphs and rule fragments. Extracted images are decoded and re-encoded rather than copied byte for byte, because lifting original bytes only works cleanly for plain JPEG streams with no mask, no CMYK color, and no unusual color space.
If you need the words rather than the pictures, rendering is the wrong tool. PDF to Text pulls the text layer directly, and OCR PDF recovers text from a scan that has no text layer to pull.
Why the PNG Files Come Out Smaller
The browser's built-in canvas export uses a compression setting tuned for speed, leaving roughly a third of the achievable savings on the table. The in-house encoder writes the PNG structure directly and compresses with the browser's native deflate stream at a better setting. On the reference page, output dropped from 338,686 bytes to 225,024 bytes, a reduction of about 34 percent for pixel-identical output. The same margin held across all three pages sampled from that single document, with bit-exact results on repeated runs.
Row filtering was tested here and made things worse. Sub filtering produced 290,241 bytes, Paeth produced 298,393 bytes, and adaptive filtering produced 256,297 bytes, all larger than the unfiltered 225,024 bytes. The reason is specific to rendered pages: a page is mostly long runs of identical white pixels that the compressor already matches as whole repeated sequences, while filtering injects fresh entropy at every glyph edge. Filtering helps photographs and hurts pages of text.
There is more compression available and DocPivot deliberately leaves it unused. An optimizer such as oxipng would take another 51 percent off the reference page, down to 109,879 bytes, but costs about 702 milliseconds per page against roughly 20 milliseconds for the current path. On a 22 page document that is the difference between an instant download and a fifteen second wait. If small files matter more than lossless fidelity, PDF to WebP gets there faster and further.
Key Features
- Edit before converting. Rotate, delete, reorder, and keep-only-these apply in the same operation.
- Purpose-built PNG encoder. About 34 percent smaller than the browser's own canvas export on the pages tested, for roughly 20 milliseconds of extra work per page.
- Opaque white background. A PDF that paints no background produces white paper rather than a transparent image that looks broken in some viewers.
- Sortable filenames and store-only ZIP. Page numbers are zero padded so files sort in document order, and output is packed without a wasteful second compression pass.
- Offline capable, with no caps. Conversion keeps working without a network connection, and there is no page limit, quota, signup, or watermark.
When PNG Is the Right Choice
PNG is the correct output when the image must be lossless and must not accumulate compression artifacts. Text edges, thin table rules, line art, screenshots, and charts all stay crisp in PNG and visibly degrade in JPEG at the same file size. A PNG can also be edited and re-saved repeatedly without further quality loss.
PNG is the wrong choice when file size dominates. A photograph-heavy page saved as PNG can be several times the size of the same page as JPEG with no visible difference, because lossless compression has almost nothing to remove from continuous-tone photographic data. For those pages, PDF to JPG is the better starting point.
Practical Use Cases
Slide decks and documentation. Export the pages you need at 150 DPI, drop the ones you do not, and paste the images into a presentation or a knowledge base article. Because deletion happens before rendering, you never generate the pages you intend to throw away.
Print-ready page proofs. Render at 300 DPI when a page is going to a printer, and at 600 DPI for technical drawings where hairlines and small callouts have to survive. The panel states the pixel dimensions so you can confirm the output meets a printer's requirement first.
Extracting product photography. Run Just the Pictures over a supplier catalog to recover product images at their embedded resolution rather than with catalog text baked around them. Deduplication means the header logo does not arrive once per page.
Archiving a subset of a long report. Use keep-only-these to isolate a range, then render it. To keep those pages as a PDF instead, Extract Pages from PDF produces a document, and Compress PDF suits a goal of a lighter file rather than a set of images.
Honest Limitations
Rendering is not extraction, and this is the most common source of disappointment with any PDF to image tool. The text in a rendered PNG is a picture of text, so it cannot be selected, searched, copied, or read by a screen reader. If any of that matters, convert to text or run OCR instead.
The ZIP is assembled whole in memory. A 22 page document at 600 DPI is comfortable on a laptop, a hundred pages is not, and a phone runs out of memory well before a desktop does. The panel says so plainly rather than silently dropping your resolution to something it can survive.
Very large pages at high DPI are scaled down to stay inside the 16,384 pixel canvas limit that browsers enforce. On oversized paper such as A0 plans, the requested DPI is therefore a target rather than a guarantee, and the panel shows the dimensions you will actually get.
Password-protected PDFs are refused rather than cracked. If you hold the password and want to remove it from your own document first, Unlock PDF handles that as a separate step.
The size and timing figures on this page come from a single 22 page test document. They are measurements taken in the loop the code actually runs in, not a broad benchmark across many file types, so your own results will vary with page content.
How This Compares to Server-Based Converters
The general-purpose conversion services are capable tools with a different architecture and business model. The practical differences show up in three places: where the file goes, what the free tier allows, and how much control you get over resolution.
Free tier caps, verified on September 1, 2026: CloudConvert starts free at 10 conversions per day, Convertio allows 10 conversion credits per 24 hours with a 100 MB file size ceiling, and FreeConvert provides 20 conversion minutes per day. These figures change, so check the provider before relying on them. A browser-based tool has no equivalent limit because there is no server cost per conversion to meter.
On resolution, most free converters apply one fixed setting or offer vague quality labels without stating what they produce. Naming the DPI and showing the resulting pixel dimensions saves a re-run. On privacy, the honest framing is that client-side processing has become common rather than rare, so it is no longer much of a differentiator by itself. What remains uncommon is the combination DocPivot ships: four stated resolutions up to 600 DPI, page editing folded into the same operation, image extraction with deduplication, and an encoder tuned for the content it is encoding.
Frequently Asked Questions
Yes, it is free with no signup, no daily quota, and no watermark on the output. There is no page limit either, though very large documents at 600 DPI are constrained by your device's available memory rather than by a policy.
No, the file never leaves your device. Parsing, rendering, and PNG encoding all happen in your browser using a self-hosted copy of pdf.js, and there is no upload step and no server-side storage to worry about.
Use 150 DPI for general purposes, 300 DPI for printing, and 96 DPI for screen-only thumbnails. Reserve 600 DPI for floor plans, blueprints, and archival scans where thin lines and small annotations are lost at lower resolutions.
Because rendering produces a picture of the page rather than a copy of its text layer. Every PDF to image conversion works this way; if you need selectable text, use a PDF to text tool or run OCR on the document instead.
Yes, you can delete pages or use keep-only-these to isolate a range before converting. Because edits are applied before rendering, pages you exclude are never generated in the first place, which also makes the export faster.
No, encrypted PDFs are detected at load time and refused with a specific error message. Remove the password from your own document first, then convert the unlocked file.
Because the tool uses its own PNG encoder instead of the browser's canvas export, which is tuned for speed rather than size. On the pages tested, output was about 34 percent smaller for pixel-identical images, at a cost of roughly 20 milliseconds per page.
No, pages render onto an opaque white canvas by design. A PDF that paints no background of its own would otherwise produce a transparent image that appears broken or unreadable in some viewers and messaging apps.
There is no fixed page cap, but the ZIP is built entirely in memory, so device memory is the real constraint. A 22 page document at 600 DPI is comfortable on a laptop, a hundred pages is not, and phones reach their limit sooner.
Every Page renders each page as an image including its text and layout, while Just the Pictures extracts only the photographs embedded inside the PDF at their own resolution. Choose extraction for catalogs and scanned albums, and rendering for archiving, printing, or sharing a page as an image.
Choose PNG for text, charts, line art, and screenshots, where lossless compression keeps edges sharp. Choose JPG for photograph-heavy pages, where PNG files can be several times larger with no visible quality advantage. If you already rendered to PNG and want smaller files, PNG to JPG converts them without returning to the source document.
Yes, once the page has loaded, conversion continues to work offline. The pdf.js library and its supporting font and character map files are self-hosted rather than pulled from a content delivery network at conversion time.
