Awaiting Update on Large PDF to TIFF Conversion Performance

Aspose.Total Product Family
, ,
1

Continuing the discussion from PDF to TIFF Conversion Performance Issue for Large Excel-Derived PDF (14,303 Pages):

Hi Aspose Team,

Could you please provide an update on this issue?

We posted this question about five days ago and haven’t received any feedback yet. This issue is currently blocking our production optimization, as we are trying to reduce the processing time for very large PDF-to-TIFF conversions.

Could you please help us understand:

  • Whether the observed conversion time is expected for very large PDFs (10,000+ pages)?
  • Whether there are any recommended settings or best practices to improve conversion performance?
  • Whether the performance depends on the complexity of the PDF content, or if there are any known limitations when processing large Excel-derived PDFs?
  • Whether there is any possibility of optimizing or parallelizing the PDF-to-TIFF conversion process?

If additional information is required, we are happy to provide the sample Excel file, generated PDF, TIFF output, or a complete test application.

We would appreciate any update or guidance, as this issue is affecting our production workflow.

Thank you.

2

@ragu2736
Thank you for drawing attention to original issue.
I wrote a set of recommendations in attached post.
I’ll duplicate answer here:

One recommendation I can give is to use TiffSettings.Depth = ColorDepth.Format1bpp with CompressionType.CCITT4; this is considered as cheaper to encode and write than the default LZW path.

I tried to check what could also affect perfomance, these seems like

  • Set TiffSettings.SkipBlankPages = true. Excel-derived PDFs contain many empty pages, and the device has a built-in blank-page detector that skips appending them.
  • Lower TIFF resolution— pass a smaller Aspose.Pdf.Devices.Resolution to TiffDevice (e.g. 150 DPI instead of 300+).
    but as I see you’ve already done that

I’ll check the question about parallel processing and add to my answer shortly

UPD. I investigated a bit a parallel scenario. Unfortunately at the moment the most of parallel processing available is to process each document individually. In other words, it won’t work for situation with one file as inner operations are processed sequentially