PDF to Image Online
Turn document pages into pictures you can crop, share or redesign.
Each page becomes a bitmap
PDF to Image uses PDF.js 3.11.174, loaded when you convert. The worker script is the matching pdf.worker.min.js from the same release. The PDF bytes are read in the browser with getDocument. There is no password box. A file that asks for a password will fail here, because nothing is sent to unlock it and nothing tries a guess.
Every page is rendered onto a canvas with alpha turned off and a white background, so a page that was “transparent” in some viewers becomes white. You then pick JPG or PNG. JPG is encoded at quality 0.92. PNG is lossless relative to that canvas, which is already a raster of the page, not the original vectors. Text you could select in the PDF becomes pixels. You cannot get the words back out by choosing PNG.
Detail is a render scale. Standard is 1.5, High is 2 and it is the default, Very high is 3. The scale multiplies the PDF page size in CSS pixels. A large architectural sheet at scale 3 can be slow and can hit a canvas limit in the browser. If the tab stalls, use High or Standard. The files download one page at a time as name-page-1.jpg or .png, staggered so the browser keeps the set.
| Control | Value |
|---|---|
| Engine | PDF.js 3.11.174, local render after the script loads. |
| Format | JPG at quality 0.92, or PNG. |
| Detail | 1.5 standard, 2 high (default), 3 very high. |
| Background | White. Alpha is off. |
| Names | filename-page-N.jpg or .png. |
What you lose on purpose
A vector logo inside the PDF stays sharp only up to the scale you picked. It does not stay a vector. Zooming the PNG past 100% shows the raster. Annotations, form fields, and layers are painted as they render, or skipped if PDF.js does not draw them. This is a preview export, not a prepress rip. For a slide that only needs a picture of page 1, High and JPG are enough. For a figure you will print large, try scale 3 once and watch memory.
Going back the other way with Image to PDF wraps those bitmaps in a new PDF. It does not restore the text layer you just discarded.
Scale 3 on a large page can exhaust the tab. If the render stops, drop back to High. The pictures are a view of the page, not a new editable document, so keep the PDF if you still need the text layer.
Exports that match the scales
- A ten-page report you will paste into a doc as pictures. JPG, scale 2. You get ten files, not a zip. The page number in the filename is the index.
- A one-page poster for a social crop. PNG at scale 2 or 3, then crop. Cropping the PDF itself is not available here.
- A scanned photo album that is already blurry. Scale 3 will not invent detail. It only samples the same scan more finely. Standard is kinder to the laptop.
- A password-protected statement. This page will not open it. There is no unlock field.
Steps
- Drop one PDF. The button stays disabled until a file is selected.
- Choose JPG or PNG, and a detail level. Leave High unless you know you need the larger canvas.
- Start the convert and leave the tab in the foreground if the document is long. Status text counts pages.
- Open
-page-1before you assume the rest are framed correctly. A crop box or a weird rotation in the PDF shows up on that first raster.
Frequently asked questions
Is the PDF uploaded?
No. PDF.js renders it in the browser. The library files may download from a CDN. The document is not posted to ImageSwift.
What do the detail options change?
The render scale: 1.5, 2, or 3. High (2) is the default. Larger scales make bigger images and use more memory.
Why is the background white?
The canvas is created with alpha off and a white background. Transparent regions in the PDF paint as white in the JPG or PNG.
Can I get selectable text?
No. Both JPG and PNG are pictures of the page. The text layer is not copied out.
How are the files named?
The PDF’s base name, then -page-, the page number, and .jpg or .png. Downloads are staggered.
Will a password-protected PDF open?
No. This page does not ask for a password and does not try to bypass one.