The quality of a scanned PDF is decided before any software touches it — in the thirty seconds you spend photographing the page. A well-shot page cleans up into a crisp, readable document. A poorly shot one — shadowed, skewed, too close — can only be improved so far, and no enhancement setting can invent detail the camera never captured. If you want a clean PDF from your phone, the shooting habits matter more than the tool you use afterwards.
Scan to PDF takes photographed or scanned pages, cleans them with a document enhancement, lets you rotate individual pages, and binds up to ten pages into one PDF, all in your browser. It is a capable finishing step. But it works with the pixels you give it, and this post is about giving it good pixels: light, angle, distance, and order, practiced until they become a routine.
Light first: shadows are the enemy
The most common flaw in a photographed page is shadow — usually your own shadow, or the phone's shadow, falling across the paper. Enhancement can lighten a grey, shadowed photo considerably, but a hard shadow across a paragraph will still leave that paragraph dimmer and harder to read than the rest, and in bad cases it can swallow words entirely.
Shoot in bright, even light. Daylight near a window, with the page flat and the light coming from the side rather than from behind you, is ideal. Avoid a single overhead light directly above the page, which casts the phone's shadow straight onto the text. If you must shoot indoors under a lamp, move the lamp to the side so its light rakes across the page, and hold the phone outside the light path. Two light sources, one on each side, are better than one bright source, because they cancel each other's shadows.
Also watch for glare. Glossy paper, laminated pages, and plastic sleeves reflect a bright patch that hides whatever is underneath it. Tilt the page slightly, or move yourself slightly, until the glare patch moves off the text. A page shot at a very slight angle with no glare is far more useful than a perfectly square page with a whiteout across the middle.
Angle and distance: square, steady, and complete
Hold the phone parallel to the page, directly above its centre, so the page appears as a rectangle rather than a trapezoid. You do not need a tripod or a stand — resting your elbows on the table, or bracing the phone hand against a book, steadies the shot enough for most purposes. The key errors to avoid are shooting from an angle, which distorts the page shape and makes lines converge, and shooting while moving, which blurs fine print.
Distance is a trade-off. Too close, and the page edges fall outside the frame, cutting off margins and sometimes text. Too far, and the text becomes small in the frame, losing detail that enhancement cannot restore. Aim to fill the frame with the page plus a small margin all round — enough margin that the page edges are clearly visible, not so much that the page is a small rectangle in a large photo. For a standard letter or A4 page, that usually means holding the phone roughly an arm's length above the page, but check the preview rather than trusting a fixed distance: different phones have different lenses.
Shoot every page the same way. Consistent distance and angle make the finished PDF look like a document rather than a scrapbook, and they make enhancement behave consistently from page to page. If one page is shot close and the next from far away, the text sizes will jump around in the final PDF, and the reader will notice even if they cannot name the cause.
Order, completeness, and a quick review
Photograph pages in order, and do not shuffle them afterwards if you can avoid it. Page order matters: a scan is a document, and a document read out of order is confusing even when every page is legible. If you must photograph out of order — because a page is stuck, or a section is elsewhere — note the correct order as you go, so assembly is a matter of following notes rather than reconstructing a sequence from memory.
Check each shot immediately, before moving on. Zoom into the preview and confirm the whole page is in frame, the text is sharp, and no shadow or glare hides anything. Re-shooting a page takes five seconds when the document is still in front of you. Discovering a blurred page after the document has been filed away costs a retrieval, a re-shoot, and a reassembly. This immediate check is the single highest-value habit in phone scanning, and it is free.
Worked example: a fictional ten-page file
Imagine a fictional tenant, Grace Liu, photographing a fictional ten-page tenancy file for a fictional landlord at example.org, using only her phone. The file includes a lease, an inventory, and several letters. Grace works through it methodically.
She places the file on a plain table near a window, in the morning light, with the pages flat — she weights curled pages with a book at the edge, outside the text area, until they lie flat. She photographs page one, checks the preview, and finds her shadow across the lower half. She moves the lamp to the side, re-shoots, and the shadow is gone. She continues in order, checking each preview, and re-shoots page six because a fold in the paper threw a line of shadow across a paragraph. The whole shoot takes about eight minutes, including the re-shoots.
Grace then opens Scan to PDF, loads the ten photographs in order, rotates page four — which she photographed sideways to fit a wide table — and applies the document enhancement. The result is a single PDF she can send to the landlord. Because the source photographs were well lit and square, the enhancement has good material to work with, and the finished pages are evenly readable. Had she shot in dim light with shadows, the same enhancement would have produced a greyer, less even result, and no setting would have fully compensated.
Remember two facts about the result. First, image quality and page order matter: they are set at shooting time, and they carry through to the finished PDF. Second, a scan is an image, not searchable text, unless text recognition is also run on it afterwards. If Grace later needs to search the tenancy file, OCR PDF can add a searchable text layer to the scanned PDF as a separate step — but the scan itself, on its own, is pictures of pages, and should be filed and described accordingly.
If the photographed pages need trimming or resizing before binding, Image Optimizer handles image preparation, and JPG to PDF is an alternative route for combining images into a PDF. The guide to resizing and cropping images covers the sizing decisions that often come up alongside a scan.
Honest limitations
Scan to PDF assembles up to ten pages, so a longer document must be bound in parts or trimmed first — it is not a bulk archival tool for a whole filing cabinet in one pass. Enhancement improves a grey or shadowed photograph, but it cannot restore text hidden by a hard shadow, glare, or blur, and it cannot correct a page photographed out of order or with missing edges; those faults must be fixed by re-shooting. A scan is an image, not searchable text, unless text recognition is also run, so do not promise a recipient a searchable document when you have only sent them pictures of pages. Finally, per-page rotation corrects orientation, but it does not straighten a skewed page or flatten a curled one — flatten the paper and shoot square, rather than relying on software to undo a bad shot.
FAQs
How do I turn photos of documents into a PDF?
Photograph the pages in order, in bright even light, with the phone parallel to each page and the whole page in frame. Then load the photographs into Scan to PDF, rotate any sideways pages, apply the document enhancement, and download the bound PDF.
How do I clean up a dark or shadowed photo of a page?
Start by re-shooting with better light — light from the side, not from behind you, and no phone shadow on the page. Enhancement can then lighten the result considerably, but it cannot fully rescue text hidden under a hard shadow or glare.
Can I rotate individual scanned pages before combining them?
Yes. Scan to PDF offers per-page rotation, so a page photographed sideways — for a wide table, for example — can be turned upright before the pages are bound together.
How many scanned pages can I put in one PDF?
Up to ten pages in one pass with Scan to PDF. For a longer document, bind it in parts, or split the source material into sections of ten pages each and combine the resulting PDFs afterwards.
Related reading: Make the scan searchable afterwards.
human review pending client designation