How to Convert PDF to Images (and Back Again)

Two directions, two different jobs. Going from PDF to images is about extracting something visual out of a document. Going from images to PDF is about turning loose photos into something that will be accepted as a document. Here is how each works and where each one bites.

PDF to image

Say you have a signed contract as a PDF and you need just the signature page as a PNG to drop into a form somewhere.

1. Open PDF to Image.
2. Upload the PDF.
3. Download the result - a zip with one image per page.

Every page becomes its own PNG, rendered at 150 DPI. That resolution is a deliberate middle: high enough that text stays readable and a signature stays crisp on screen and in normal printing, low enough that a 20-page document does not produce a download the size of a video file.

If your document is a single page, you still get a zip containing one image. Consistent output beats guessing whether you will get a file or an archive back, especially if something on your end is scripted against it.

Prefer JPG? PDF to JPG does the same thing with JPG output. The choice matters more than it looks:

• PNG is lossless, so text edges and thin lines stay exactly sharp. Files are bigger. This is what you want for a signature page, a form, a diagram, or anything with fine print.
• JPG is lossy and much smaller, which is fine for a page that is mostly photograph. On dense text it introduces faint ringing around the letters - not usually a problem, but visible if you zoom.

For documents, PNG is the better default. For a page that is essentially a picture, JPG saves a lot of space.

Where this runs into limits

PDFs here are capped at 50 pages and 20 MB, and rendering has a 60-second ceiling. Two things follow from that, and they are the failures people actually hit:

• The zip has fewer images than the PDF has pages. You are over the 50-page cap. Split PDF first, then convert each part.
• Text looks soft when you zoom in. 150 DPI is tuned for screen use. If you need print resolution, keep the source as a PDF rather than rasterising it - once it is an image, that detail does not exist to recover.

If the file will not open at all, it may be password-protected. Run Unlock PDF first, using the password you already have.

Image to PDF

Going the other way is just as direct, and it does more than it used to. Image to PDF takes several images at once and builds one multi-page PDF, not one PDF per image.

Drop in up to 10 images, drag them into the order you want, and the builder gives you a few things worth knowing about before you download:

• Per-image rotation. A photo of a receipt taken sideways gets rotated in the builder, so you are not fixing it afterwards in a separate tool.
• Page numbers. Optional, added automatically across the finished document.
• A password. You can encrypt the output at build time rather than making a PDF and then protecting it in a second pass.

This is the fast path for the everyday version of this job: photographing a stack of receipts for an expense claim, turning whiteboard shots into something you can attach to a ticket, or submitting photos of a document to an organisation that only accepts PDF.

Getting a good result

The conversion embeds your images as they are - it does not enhance them. So the output is only as good as the input, and a few habits help:

• Shoot straight on, in even light. A skewed photo stays skewed in the PDF.
• Crop out the desk and background first. It looks more like a document and the file gets smaller.
• If the images are large phone photos, run Compress Image before building. Ten uncompressed 12-megapixel photos make an unwieldy PDF, and nobody needs that resolution to read a receipt.

Round-tripping, and why to avoid it

A pattern worth naming: converting a PDF to images and then back to a PDF to "fix" something. It works, but it is destructive. The output is a stack of pictures, so the text is no longer text - you cannot select it, search it, or copy from it, and screen readers get nothing. The file is usually much larger too.

If what you actually want is to reorder, remove, or rotate pages, do it directly and keep the text layer intact:

• Organize PDF to reorder, rotate, and delete in one pass
• Rotate PDF for a sideways scan
• Delete PDF Pages or Extract PDF Pages for a subset
• Merge PDF to combine files

Those all move pages around without re-encoding anything. Round-trip through images only when you specifically want a flattened, non-selectable document - which is occasionally the point.

The one-hour rule

Whatever you upload, and whatever comes back, both are deleted from our servers automatically after an hour. No account, no watermark, and nothing kept.