Blurry text, especially when outputting to PDF

[Reasonable Listener]

No matter what settings I use (at least the settings that I am aware of), when I am outputting a file, the resulting files are blurry and hard to read.

I'm uploading examples so you can see what I mean. In this case I typed up some nonsense in Word, printed it out and scanned it as a PDF. Then I opened the PDF in GIMP and made no changes to the file. I outputted the files as both a PDF (soft and blurry), and a PNG (okay, but prints out soft and blurry). I have even tried printing directly from GIMP and it didn't improve anything.

If it helps to know this, the original document was scanned at 400x400 black and white.

How do I fix this problem? I tried inputting the original PDF into GIMP at various DPIs and it didn't change anything. Am I just using the wrong software for this sort of thing?

[Blighty]

When you import a pdf into Gimp you need to set the dpi in the import dialogue. See attached.

But Gimp converts text to bitmap, so you will not be able to edit the text. Gimp is not the right tool for editing pdfs that are predominantly text.

[rich2005]

I can not reproduce that blurring. I can see from your description, Black & White scan (not greyscale) which gives no antialiasing. Opening up in Gimp @ 400 ppi as your description then re-exporting as a new PDF.  All settings in and out are default.

The new PDF just the same as the original. The PDF properties shown come from a little utility.

   

Now looking at your exported from Gimp PDF all looks equal, apart from the blurriness, there is that slightly smaller file size, might/might not be relevant.

   

Can you pin down in your workflow where some blurring might come in.

[Reasonable Listener]

(03-25-2019, 04:07 PM)Blighty Wrote: When you import a pdf into Gimp you need to set the dpi in the import dialogue. See attached.

But Gimp converts text to bitmap, so you will not be able to edit the text. Gimp is not the right tool for editing pdfs that are predominantly text.

What software do you suggest for editing PDFs instead of GIMP?

[rich2005]

Your example is an embedded graphic file. Gimp is as good as anything.

If you are starting from scratch use an application such as LibreOffice where  text +font information are embedded in the file.

Edit: A quick page of txt from LibreOffice. Notice the file size

[Reasonable Listener]

(03-25-2019, 04:50 PM)rich2005 Wrote: Your example is an embedded graphic file. Gimp is as good as anything.

If you are starting from scratch use an application such as LibreOffice where  text +font information are embedded in the file.

Edit: A quick page of txt from LibreOffice. Notice the file size

But if I'm starting from a text-heavy PDF, what should I use? Or should I scan it as a TIFF or something uncompressed like that and edit from there?

[rich2005]

(03-25-2019, 05:02 PM)Reasonable Listener Wrote: But if I'm starting from a text-heavy PDF, what should I use? Or should I scan it as a TIFF or something uncompressed like that and edit from there?

If you have a PDF that has embedded text and you want to edit the text then LibreOffice is a good bet. The only issue you will have is if the font used in the PDF is not available on your computer. You will get a font replacement usually sans.

Otherwise if you have to scan it, try greyscale instead of B&W next time, unless of course it contains colour graphics. Tiff, png any format except jpeg.

[Blighty]

(03-25-2019, 05:02 PM)Reasonable Listener Wrote: But if I'm starting from a text-heavy PDF, what should I use? Or should I scan it as a TIFF or something uncompressed like that and edit from there?

Best is to use the software that created the pdf in the first place.

Otherwise use LibreOffice to open the pdf.

If you scan the text will be converted to a bitmap image and will no longer be editable as text, no matter what format you use.

[Reasonable Listener]

(03-25-2019, 05:13 PM)rich2005 Wrote: If you have a PDF that has embedded text and you want to edit the text then LibreOffice is a good bet. The only issue you will have is if the font used in the PDF is not available on your computer. You will get a font replacement usually sans.

Otherwise if you have to scan it, try greyscale instead of B&W next time, unless of course it contains colour graphics.  Tiff, png any format except jpeg.

Well there's some writing/drawing on the text, it's just mostly text, so that's why I have to scan it. Will LibreOffice be able to read the non-text as is, or will it try to re-configure it?

[rich2005]

(03-25-2019, 05:24 PM)Reasonable Listener Wrote: My workflow was, import the scanned PDF into GIMP and then output it as a PNG and then a PDF.

It is a bit of a mystery. The image is held in memory by Gimp, so importing the pdf then exporting again without edits will yield the same image. Wonder where the blurring comes in?

Quote:What PDF property utility are you using?

It is a PDF viewer 'Evince' but only because I wanted to compare the outputs from Gimp. Sometimes older versions of Cairo can give different results, both the same in this case.