Scan a Document to PDF Offline (Android)

To scan a paper document to PDF with a phone, photograph each page in even, indirect light with the page filling the frame and the camera held parallel to the paper, then let the tool write the pages into one PDF. The two things that actually ruin a phone scan are a shadow — usually your own head or hand — and perspective skew from holding the phone at an angle. Neither is fixed by a better camera, and both are fixed by where you stand.

On Android, Convertlix's Cam to PDF scans pages with your camera and writes them into a PDF on the handset, so the pages never leave the device — no upload, no server-side copy, nothing to retain. The honest scope: Android 8.0 or later, sign-in required (there is no guest mode), 25 MB per file on the free tier against a server-set monthly allowance of conversions shown in the app, output in the phone's Downloads folder, and the camera is the app's only runtime permission — it asks for no storage access at all. Everything below applies whichever scanner you use.

Why people search for "offline" on this specific task

Scanning is the one file job where the content is almost always the sensitive part. Nobody photographs a page they don't care about: it's a payslip, a tenancy agreement, a passport page, a medical letter, a signed contract. So "offline" isn't a technical preference here, it's the actual requirement.

And it is a fair thing to check, because the default in this category runs the other way. As of September 2026, several mainstream scanner apps sign you into an account, store scans in that account's cloud storage by default, and declare documents as collected in their Play Data safety section. That is not a scandal — it is how a product that offers sync across your devices has to work, and their own listings say so. It just means the answer to "where did my payslip go?" is a sentence about a server, and for some documents that's the wrong sentence.

The checkable version of the question: turn the network off and try to scan. A tool that processes on the device carries on working. A tool that needs a server tells you so.

Light and geometry beat megapixels

The most common wrong instinct is to reach for more resolution. You already have too much.

An A4 page scanned at 300 DPI — the density most portals and archives ask for — is 2,480 × 3,508 pixels, about 8.7 megapixels per page. Any phone from the last eight years shoots more than that. Resolution stopped being the constraint a long time ago; what limits a phone scan is the two things a flatbed controls for you:

• Even illumination. A flatbed drags a light bar the width of the page, so every square centimetre gets identical light. Your ceiling lamp does not. One bright patch and one dim corner is what makes a phone scan look like a phone scan, and no amount of contrast adjustment afterwards puts back detail that was blown out or crushed.
• A fixed, parallel focal plane. The glass holds the page flat and exactly perpendicular to the lens. Handheld, you are guessing, and tilting the phone even 10–15° makes the far edge of the page measurably shorter than the near edge — the page comes out as a trapezoid rather than a rectangle.

Two practical consequences. First, open shade or two soft light sources beat one bright lamp: a single overhead bulb puts a specular hotspot on anything glossy and puts your own shadow across the page when you lean in. Second, get square to the page before you worry about anything else — a tilted capture is the defect hardest to fix later, and if you have one already, Straighten a Crooked Photo of a Document covers the difference between a rotation and a perspective correction, which are not the same repair.

What a phone scan costs you in bytes

This is where a scan-to-PDF job usually goes wrong for people, and it is worth knowing before you start rather than at the upload form.

A scanned PDF has no text in it. Every page is a photograph, so a ten-page colour document typically lands somewhere in the 8–15 MB range, and a portal with a 2 MB cap will bounce it. That is not a fault in the scanner; it is what storing ten 8.7-megapixel images costs. Why Is My Scanned PDF So Large? explains the arithmetic in full, and if you already have a cap to hit, Compress a PDF Under 2 MB on Android (Free) is the next step rather than a rescan.

The one thing you can do at capture time: don't scan in colour if the colour carries no information. A greyscale scan of a typed contract loses nothing a reviewer needs and is routinely 40–60% smaller. Do keep colour for anything with a stamp, seal, hologram or coloured signature that someone has to verify.

How to scan a document to PDF on your phone

1. Lay the page flat on a matte, contrasting surface. A dark table under white paper gives the edge detection something to find. Smooth out folds first — a crease throws a shadow line that reads as a mark on the page.
2. Fix the light before you fix the framing. Face a window, or use two soft sources from opposite sides. Then check for your own shadow by watching the preview as you lean in, not by looking at the paper.
3. Hold the phone parallel to the page and fill the frame. Centre the page, get the whole thing in with a small margin, and keep the phone flat rather than angled. Parallel is worth more than steady.
4. Shoot every page the same way. Same distance, same light, same orientation. A bundle where page 4 is brighter and page 7 is rotated looks careless to whoever reads it, and mixed exposure defeats any later compression pass.
5. Write the pages into one PDF, in page order. In Convertlix, open Cam to PDF and capture the pages in sequence; the tool assembles them into a single document on the phone. If the order comes out wrong, that is a page-manager job rather than a rescan — see Remove, Reorder or Rotate PDF Pages (Android).
6. Check the output size, then read the smallest text on every page. Read the file size in Downloads before you go near an upload form, and zoom to the smallest footnote on the worst-lit page. That is the page a reviewer will complain about.

When a phone scan isn't good enough

Be honest about these before you spend an afternoon on it:

• Glossy, laminated or plastic pages. ID cards, certificates in plastic sleeves and magazine stock all throw a specular highlight straight back at the lens. Take the sleeve off, or use a flatbed.
• Bound books. The gutter near the spine curves away from the lens, so it falls out of focus and into shadow at the same time. A 2D correction cannot flatten a physical curve — that is a photocopier or a proper book scanner.
• Certified colour fidelity. Reproducing an exact Pantone or a colour-matched proof needs a calibrated scanner and a colour profile, not a phone.
• Very faint pencil, carbon copies and thermal receipts. Low-contrast marks are exactly what even lighting exists to rescue, and thermal paper also fades in daylight while you work.
• A scan is not text. Cam to PDF produces page images; nothing in that file is selectable or searchable. If you need the words rather than the picture, Copy Text Out of a Photo or Scan (Android) is the tool for that, and it is a separate job.

FAQ

Is a phone scan good enough for a visa or job application?

Usually yes, provided the text is legible at 100% zoom and the page is square and fully in frame. Most portals care about legibility, page order and file size rather than the device you used. Where they specify a scanning density — 200 or 300 DPI is common — a phone shooting a full-frame A4 page comfortably exceeds it, because the page is already being captured at more than 8 megapixels.

Can I scan a document without an internet connection?

Yes, if the tool processes on the device. Convertlix needs one successful sign-in when it is first set up; after that its tools open and run with mobile data and Wi-Fi switched off, including Cam to PDF. An app that shows an upload progress bar or stalls in airplane mode is doing the work on a server instead, and airplane mode is the fastest way to find out which kind you have.

Does Convertlix upload my scans?

No. The pages are captured by the camera, assembled into a PDF on the phone, and never sent anywhere — and the app has no storage permission, so it cannot read files you did not hand it. Its network calls carry identity, licence, quota, configuration and payment data, never a document. Sign in once, switch on airplane mode, and the tool still works.

Why is my scanned PDF so much bigger than a Word document?

Because it is a stack of photographs rather than text. A born-digital eight-page document stores letters as characters and typically weighs 100–400 KB; the same eight pages scanned store roughly 8.7 million pixels each and land in the megabytes. Dropping to greyscale and downsampling to 150 DPI is what closes the gap, at the cost of some fine detail.

Is there a free way to scan a document without installing another app?

Yes, and one of them may already be on your phone. Google Drive has a built-in document scan on Android, and Microsoft Lens and Adobe Scan are both free and capable — as of September 2026 all three are genuinely good at edge detection and cleanup, and Lens and Adobe Scan also do OCR, which Cam to PDF does not. The difference is where the file goes: those products are built around an account and cloud storage, so check their current terms and Data safety entries and decide per document. If you are comfortable at a desk, a flatbed plus ImageMagick or img2pdf gives you more control than any phone app.

---

If the document is on paper in front of you and you would rather it never touched a server, capture the pages on the phone and read the file size in Downloads before you send it anywhere.