Trying to make such things:

1st step:

convert img.jpg img.pdf

2nd step:

pdfimages -j img.pdf img1

Comparing source and extracted images in HEX shows difference. How to make such conversion without data loss?


One way is to use pdflatex instead of convert.

You need in a extra file, which is here called image.tex:


Then run pdflatex image.tex to generate image.pdf.


Are you sure that there's an entire JPEG - metadata and picture data, in JFIF/JPEG format within the PDF? If not, then even if the image data is extracted verbatim, pdfimages will have to reconstruct the container and that may not match.

You can get a similar situation with audio files and tags etc. - you can't do sum comparisons if you change the metadata.

IN that situation, you need to calculate hashes for just the data portion, rather than the whole file.

  • 1
    The approach with convert does really return a different picture. Comparing the images with gimp shows slight differences. With pdflatex as described in my answer, you get exactly the same image back - including the original metadata. – jofel Mar 14 '12 at 12:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.