[Bug] SubprocessOutputError about ocrmypdf

[jbarlow83]

ocrmypdf -v1 --max-image-mpixels 1000 --tesseract-downsample-large-images --output-type pdf test.pdf ocr.pdf will work in recent versions (v16+). In some future release that argument will be enabled by default since there's no reason not to.

A further complication in this file it that uses a lot of stencil masks, some with odd proportions, and ocrmypdf is very cautious about degrading image quality so it picks a very large output image size as a way of accurately accounting for these stencil masks. This is a tricky issue since there are legitimately use cases of very high resolution files, when working with scanned film for example.

You can add --tesseract-downsample-above 3000 for example, and OCR will complete more quickly.

[user1823]

You can add --tesseract-downsample-above 3000 for example, and OCR will complete more quickly.

Thanks for the suggestion.