Grayscale PDF
Convert a colour document to grey, keeping the text searchable.
This tool runs entirely in your browser. Your file is never uploaded, and you can verify that in your browser's network tab. Verify it yourself: open your browser's network tab and watch. You will see one small request asking whether you have tasks left - a tool name and a hash, never the file.
What this does
There is no way to recolour a PDF's vector content inside a browser - the colour lives in every content stream, in whichever of five colour spaces the producer chose. So this tool says plainly how it works: each page is re-rendered in grey and the document is rebuilt from the resulting images. The pages stop being vector as a result. The words do not stop being words: the text is read out first and drawn back invisibly over each image, so the document is still selectable and searchable.
The usual reason is printing. Colour pages cost several times what grey ones do on most office and copy-shop tariffs, and many print accounting systems bill by what the file contains rather than by what the driver was set to. It is also a straightforward way to shrink a colour scan.
How it works
- Drop the PDF onto this page. Files up to 300 MB are accepted.
- Set the resolution. 150 dpi is the default and reads well on screen; raise it towards 300 dpi for a document that will be printed and examined closely.
- Leave Keep text searchable on unless you have a reason not to - it costs a little size and keeps the document usable.
- Adjust the JPEG quality under advanced options if the default of 82 per cent is softer or larger than you want.
- Press Convert to grayscale. Progress shows which page is being rendered, and the result reports the resolution used.
The conversion uses Rec. 709 luma rather than the average of the red, green and blue channels. The eye is roughly three times more sensitive to green than to red and ten times more than to blue, so averaging makes a saturated red and a saturated blue come out as the same middle grey when one should be clearly darker. Weighting the channels the way vision does keeps coloured text, chart series and highlighted rows distinguishable after the colour is gone.
Keeping the text searchable is what separates this from printing to an image. Before any page is replaced, PDF.js extracts every text run with its position and size; those runs are then drawn over the finished image in an invisible rendering mode. You can select a paragraph, search for a phrase and copy a table out of the result, even though what you are looking at is a picture.
The file will not always get smaller. A colour scan usually shrinks substantially, because three channels become one and the JPEG has less to encode. A text document made of vector glyphs frequently grows, sometimes several times over, because a page of crisp type is cheaper to store as instructions than as a picture of itself. If your goal is size rather than colour, Compress PDF is the tool to reach for.
Because the pages are rebuilt from images, what sits on top of them does not survive: links stop being clickable, form fields stop being fillable, and comments, highlights and stamps are gone rather than drawn in. If you need the marks kept as marks, flatten the document first and convert afterwards.
Rendering happens on your device with PDF.js and a canvas, and the rebuild with pdf-lib - both in Web Workers, neither over a network. This is the heaviest tool in the category on a large file, and it is still local: a 300-page colour report is rendered page by page in your own browser.
What this tool cannot do
- The output is images, not vector pages. Zooming in far enough shows the pixels, and the type is no longer infinitely sharp.
- A text-heavy document can come out larger than the original, because a page of type stores more cheaply as instructions than as a picture.
- Links, form fields and annotations are not carried over, since each page is replaced by an image of itself.
Questions people ask
- Why convert the file instead of ticking grayscale in the print dialog?
- Because the print dialog only affects that one job on that one machine. A copy shop, a shared office printer or a colleague opening the file will each make their own decision, and many print accounting systems charge a page as colour if the file contains colour, whatever the driver was set to. Converting the document settles it once.
- How much quality do I lose?
- The colour is gone permanently, and the pages become images at the resolution you choose. At 150 dpi the result reads perfectly on screen and prints acceptably; at 300 dpi it is print quality and the file is roughly four times larger. The text layer is preserved either way.
- Will the file be smaller?
- A colour scan usually shrinks a long way. A text document often grows, because rendering type into a JPEG costs more than the instructions that drew it. Both outcomes are reported as before and after sizes, so you can see which one you got and keep the original if it is not what you wanted.
- Can I still search and select the text afterwards?
- Yes, as long as Keep text searchable is on. The text is extracted before the pages are replaced and drawn back over the images invisibly, so selection, search and copy all still work. If the original was a scan with no text layer to begin with, there is nothing to preserve - run OCR PDF first.
- Is the document uploaded for rendering?
- No. Every page is rendered by PDF.js onto a canvas in your own browser and reassembled by pdf-lib in a Web Worker. Nothing is transmitted, which also means the speed depends on your device rather than on a queue somewhere.