JPG to PDF

Controls and Method

Your Photos Are Already JPEG. This Encodes Them Again.

PDF stores page images as JPEG, and so does your camera roll. That means the photos go through the JPEG encoder a second time on the way into the document, and JPEG is not idempotent: re-encoding decoded pixels does not reproduce the original file, it adds a fresh layer of loss on top of the loss already baked in.

For a photograph of a scene this is genuinely hard to see at quality 85. For a photograph of a printed page — which is what most people are actually converting — the second pass lands on top of edges the camera already softened, and text can visibly degrade. If your JPGs are photos of documents, use 95 and accept the larger file.

The most common quality problem, though, is not the encoder at all: it is the source photo. A 4000-pixel-wide phone photo of an A4 page shot at an angle in poor light will make a bad PDF page at any quality setting. Cropping to the page edges and straightening or resizing first does more for legibility than the slider ever will.

Picking a Page Size for Photos

Photos rarely match paper proportions. A 4:3 phone photo on an A4 page leaves bands of white above and below; the same photo on a fit-to-image page has none. Which one is right depends entirely on whether the PDF will be printed or read on a screen.

Mixed batches are where the two modes diverge sharply. On A4 or Letter, each page is measured independently and switches to landscape whenever that photo is wider than it is tall, so a set of portrait and landscape shots produces a document whose sheets alternate orientation. Printed, that is correct and every photo fills its page; read on screen, the reader has to turn their head every second page. In fit-to-image mode nothing alternates, because each page simply takes its own photo’s shape — but a batch of differently shaped photos then yields a PDF with no consistent page size at all, which some viewers and print shops handle poorly.

If a mixed set has to print cleanly and read comfortably, the reliable fix is upstream rather than here: crop the odd ones out to a shared aspect ratio before converting, so every page lands the same way up.

Rotation, and Why a Photo Can Arrive Sideways

Phone and camera sensors read out in one fixed direction regardless of how you were holding the device. What makes a photo appear upright is not the pixels but an EXIF orientation tag written alongside them, telling every viewer how far to rotate the image before displaying it. This is why the same file can look correct in your photo library and sideways in an older application: one honors the tag and the other does not.

The browser applies that tag when it decodes each photo here, so the page you get matches what you saw in your photo viewer rather than the raw sensor readout. The tag itself does not travel into the document. A PDF page is a placed image with fixed geometry, so the rotation is baked into the page at layout time and there is nothing left for a later viewer to interpret differently. That is the desirable outcome: it is the last point at which the orientation can be got wrong.

The failure case worth knowing is a photo whose tag was already lost before it reached this tool — most often because it passed through something that stripped EXIF without rotating the pixels first. Such a file genuinely is sideways by then, and it will be baked in sideways. Rotate it in your photo app before converting. The same ordering applies if you strip the metadata yourself: the orientation tag goes with the rest of the EXIF block, so rotate first, then clean, then build the PDF.

A Photo of Text Is Not Text

This tool performs no OCR. A photographed contract becomes a picture of a contract: nobody can search it, select a clause out of it, or have a screen reader read it aloud, and neither can any automated system the document is submitted to. That is often fine — a receipt only needs to be looked at — but it is worth knowing before you send a job application this way.

It also means the PDF cannot be edited as a document afterwards: the underlying words are pixels, not text. If the original exists as a Word file or a real PDF, converting that instead will always give a better result than photographing it.

Getting the Order Right

Pages follow the order the file picker returns, which on most systems means alphabetical by filename rather than the order you clicked. Phone photos named IMG_0009 through IMG_0012 sort correctly; files named page1, page2, page10 will put page 10 second. Renaming with zero padding before you start is the simplest fix, and rearranging the pages afterwards is the fallback.

Practical Uses

• Turn a set of photographed receipts into one attachment for an expense claim.
• Send phone photos of a signed form as a single document rather than four images.
• Assemble photographed pages of a book or record for reference.
• Package a small photo portfolio into one file a recipient can page through.
• Combine ID or passport photos into one document an application form will accept.