How to Turn a PDF Black and White for Printing

How to convert a PDF to grayscale for printing

A print shop's billing system calls a page colour if it finds a single coloured pixel on it. A blue hyperlink, a red table header, a logo in the corner, and the sheet bills at the colour rate, which is normally five to ten times the mono one. Hence the task: strip the colour so thoroughly that the shop's automated check genuinely cannot find any.

The tool does that reliably, by a method worth knowing about in advance.

What actually happens to a page

Going grey here is not a filter laid over the content. Every page is rendered into an image, the image is reduced to a single brightness channel, and that image becomes the new page. The finished file is a stack of pictures carrying the original sheet dimensions.

There is a real strength in that. The result carries the DeviceGray colour space, that is, one-channel grey rather than grey painted inside RGB. DeviceGray is exactly what the shop's automation looks for, and it will have nothing to object to.

The price of that reliability is the text layer. It does not carry over. You cannot search the finished file, you cannot select or copy from it, and an intake form that checks for a text layer will reject the document.

When it is a win and when it is a loss

Which way the file size moves depends on what was inside. Measured on two typical documents:

The logic is straightforward. The scan already held images, and replacing three colour channels with one brightness channel is an honest saving. The text document held fonts and vector drawing commands taking a few kilobytes, and a full-page raster arrives in their place.

Hence the rule. Scans, photographed documents, decks with coloured blocks: converting pays. A contract, an invoice or a statement laid out in Word and saved to PDF: it does not. Black text on white already prints at the mono rate, and the conversion only damages the file.

The awkward case is a text document with a coloured logo on the title page. There it is cheaper to cut the document with split-pdf, convert the single title page and put it back together, than to rasterise all forty sheets.

How to convert the document

1. Open pdf-to-grayscale and upload the document. 2. Leave the standard profile if the source is properly exposed. 3. Touch the contrast threshold only when there is a visible problem, described below. 4. Run the job and download the result. 5. Open the file and check the pages holding signatures, stamps and small type: that is where losses show up first.

Three profiles and what they really change

The profile controls whether the brightness range is stretched after the conversion to grey.

The names mislead: deep is the more aggressive of the two, although preserve contrast sounds like it would be. The difference between them only shows where there is something to stretch. On a clean document that already holds true black and true white, all three profiles give practically the same result.

The contrast threshold: 50 is neutral

The threshold takes a number from 0 to 100 and defaults to 50. That is not half the effect, it is the neutral position: the image is left alone. Higher values raise contrast, lower ones flatten it, and the pivot sits at mid-grey.

Measured on one and the same text document, the brightness range of the finished page from its darkest pixel to its lightest:

The bottom of the scale is worth remembering. Zero does not give you a black and white image and does nothing useful at all: the whole page is squeezed into a narrow band around mid-grey. If you want it harder, move up, not down.

The top of the scale earns its keep in exactly one scenario: a dull scan where the letters are grey and the background is dirty. Values around 70 to 80 pull them apart again. On a normal document those same 80 will eat the midtones out of stamps and signatures.

What carries over into the result

The document properties survive: title, author and keywords move across from the source file. Bookmarks carry over too and still point at their pages, because the page count does not change.

Sheet dimensions are preserved to a fraction of a point: A4 stays A4 rather than drifting into some other format, which is what naive reassembly from images usually does.

What does not carry over, besides the text layer, is everything interactive: form fields, links, annotations. They were objects on a page, and the page became a picture.

Limits, and the order of work with other tools

A document longer than a thousand pages will not be accepted: rasterising every page is expensive, and the ceiling here is hard. A document locked with an open password will not pass either and has to be decrypted first.

If the file still needs slimming after conversion, compress-pdf will manage: the pages are pictures now, and pictures are exactly the material compression works best on.

If you need the text layer back and the grey version already exists, recognition through ocr-pdf restores search. The layer will be a recognised one rather than the original, so typos are possible, but the document becomes searchable again.

If the file heads for an archive afterwards, run pdf-to-pdfa last: the grey conversion rebuilds the document and would strip a conformance marker applied earlier.