Scanned PDF Too Dark or Washed Out? Fix It Before You Send It
Published 2026-11-13 · NoFilePDF
You can handle this directly in your browser with our free pdf to images tool — nothing gets uploaded.
You scanned a document, opened it, and it looks wrong. Either the background is a muddy grey that makes the text hard to read, or the page is so washed out that the letters fade into the paper. Recipients comment on it, better yet they ask you to resend. The file is not broken; the scanner captured the page at the wrong brightness and contrast, and those values were baked into the image. The fix is to correct the image, not to rescan blindly.
This guide explains what actually goes wrong in a scan, how to choose between colour, grey and black-and-white output, and how to correct a file you already have before you send it.
Why scans go dark or washed out
A scanner is a camera with a lamp. Its automatic exposure looks at the whole page and tries to average it. A form with a lot of black text on white paper can average to dark, so the scanner lightens everything and the thin strokes disappear. A dark or coloured background, or a shadow from a folded corner, can pull the average the other way, and the scanner overexposes the page. A dirty platen glass makes it worse, because the dirt is read as content and shifts the exposure.
Auto exposure fighting the page. The most common cause by far, especially with forms and table-heavy pages.
Low resolution. At 150 dpi or less, grey antialiasing around letters looks like smoke and the page reads as a grey blur.
Colour mode where nobody needs colour. A colour scan of a black text page keeps every temperature shift and every yellowed paper stain.
Compression. Heavy JPEG compression smears letters and makes the page look soft and dirty.
Choose the right output mode
Most documents should be scanned in grey, not colour. Grey keeps pencil marks and stamps readable while ignoring the paper tint. Use black and white only for clean text where you actively want the background gone; it produces the smallest files but turns faint signatures into noise. Use colour only when colour carries information: photos, highlighted text where the highlight matters, product labels, ID cards with colour elements.
Resolution matters more than most people think. 300 dpi is the sweet spot for a readable, reasonably sized file. 200 dpi is acceptable for plain text if you must keep the file small. Below that, letters start to break apart regardless of how good the exposure is.
Fix a scan you already have
Increase contrast slightly and set the white point. The background should become white, not light grey. In an image editor, use levels rather than brightness: pull the white point down until the paper is white, then adjust the black point until the text is solid.
Do not over-brighten. Raising brightness washes out thin strokes and signatures. Contrast plus a white point fixes the background; brightness alone does not.
Convert to grey if colour adds nothing. Removing colour also removes the yellow cast of older paper and shrinks the file.
Re-save with sensible compression. A moderate quality setting keeps text crisp; the maximum compression setting is what produces the smeared, dirty look.
Before you send it
Open the corrected file at 100 percent zoom and read three lines of the smallest text. If you can read them without squinting, a recipient on a phone can too. Then check the file size: a four page grey scan at 300 dpi should be comfortably under two megabytes. If it is far larger, the scanner saved raw colour images, and converting to grey will solve both the look and the size problem.
Troubleshooting table
Problem
Cause
Fix
Grey background, faint text
Auto exposure averaged to dark
Raise contrast and set white point
Letters look smoky
Resolution too low
Rescan at 300 dpi
Page looks soft overall
Heavy JPEG compression
Re-save at moderate quality
Yellow paper tint
Colour scan of tinted paper
Convert to grey
Signatures disappear
Black-and-white threshold too aggressive
Use grey, not bilevel
Set up the scanner so you never fix it again
Avoiding the problem at scan time is far cheaper than repairing it afterwards. All three major scanning paths offer a way to stop the software from guessing the exposure.
On a phone, the built-in document scanner in iOS Files and on Android from Google Drive or the camera app both let you adjust the crop corners and, in some versions, the filter. Choose the document filter rather than the photo filter: the photo filter keeps the paper tint and the shadows, the document filter flattens the background to white and leaves text dark. Hold the phone parallel to the page and use even light from a window or a lamp, not from a torch held at an angle that creates a hot spot.
On a flatbed scanner, switch the driver from Auto to Manual and set the brightness so that a white sheet reads white rather than grey. Clean the glass; a smudge under the lid is read as content and drags the average exposure in the wrong direction. Keep the lid closed, because a light leak on one side is read as a gradient and the software compensates by darkening the whole page.
For a feeder scanner, check the paper guides. A page fed slightly skewed creates a shadow band at one edge, and the automatic contrast rule reacts to that band. Ten seconds of alignment saves a whole correction pass later.
Recognising a scan that cannot be fixed
Some damage is irreversible. If the source was scanned at 100 dpi, the letters were never captured at enough detail for a correction to sharpen them; raising contrast only makes blocky letters darker, not clearer. If the page was photographed at an angle and the perspective was not corrected, text at one end is compressed and the other end is stretched, which no contrast tool will undo. If the file was compressed to a fraction of its size, the blocking artefacts around letters are permanent.
In those cases a rescan from the paper original is the only real answer. When there is no paper original, the realistic goal changes: make the text as legible as possible for a human reader and state in the email that the source was poor. A file that is honest about its condition is more useful than one that has been over-filtered into illegibility.
Avoiding the problem at scan time is far cheaper than repairing it afterwards. All three major scanning paths offer a way to stop the software from guessing the exposure.
On a phone, the built-in document scanner in iOS Files and on Android from Google Drive or the camera app both let you adjust the crop corners and, in some versions, the filter. Choose the document filter rather than the photo filter: the photo filter keeps the paper tint and the shadows, the document filter flattens the background to white and leaves text dark. Hold the phone parallel to the page and use even light from a window or a lamp, not from a torch held at an angle that creates a hot spot.
On a flatbed scanner, switch the driver from Auto to Manual and set the brightness so that a white sheet reads white rather than grey. Clean the glass; a smudge under the lid is read as content and drags the average exposure in the wrong direction. Keep the lid closed, because a light leak on one side is read as a gradient and the software compensates by darkening the whole page.
For a feeder scanner, check the paper guides. A page fed slightly skewed creates a shadow band at one edge, and the automatic contrast rule reacts to that band. Ten seconds of alignment saves a whole correction pass later.
Recognising a scan that cannot be fixed
Some damage is irreversible. If the source was scanned at 100 dpi, the letters were never captured at enough detail for a correction to sharpen them; raising contrast only makes blocky letters darker, not clearer. If the page was photographed at an angle and the perspective was not corrected, text at one end is compressed and the other end is stretched, which no contrast tool will undo. If the file was compressed to a fraction of its size, the blocking artefacts around letters are permanent.
In those cases a rescan from the paper original is the only real answer. When there is no paper original, the realistic goal changes: make the text as legible as possible for a human reader and state in the email that the source was poor. A file that is honest about its condition is more useful than one that has been over-filtered into illegibility.
FAQ
Can I fix a bad scan without rescanning?
Often yes. Contrast and white point correction fixes most grey, washed out pages. If resolution was too low at scan time, no correction can recreate the lost detail.
Should I scan in black and white to get a small file?
Only for clean text. Bilevel scans destroy signatures, photos and faint pencil marks, and produce harsher results than grey at only slightly smaller sizes.
Why does the scan look worse than the original?
Because the scanner decided the exposure for you. Manual mode, or a fixed exposure in a scanning app, keeps the page looking like the paper.
Do I need special software?
No. Any viewer that offers contrast or levels works, and browser tools can crop and compress the result. The free tools at pdf.dingsheng.cyou run locally without uploading the file.
Will compression make the text unreadable?
Not at a moderate setting. Text stays crisp until compression is pushed to the maximum, which is exactly where the smeared look appears.