8 Replies Latest reply on Jan 15, 2013 2:32 AM by Srinath Menon-Oracle

    PDF Optimization (web-optimized)


      I am looking for some capabilities of WCC 11g for optimizing the uploaded PDF files. Essentially following is what i am looking for:

      Step1 User upload High res PDF files
      Step2 It goes through the workflow and admin checks the PDF file
      Step3 Once it is approved a Low resolution PDF file needs to be created.

      So system will contain High resolution and Low resolution PDF files after approval.
      I am fine till Step2, but can Step3 be automated in WCC i.e. automatic creation of low resolution PDF files (web-optimized) after approval using OOB WCC capabilities?

      Do we need to setup IBR for this??

        • 1. Re: PDF Optimization (web-optimized)
          Yes, it sounds like a use case for IBR - few years ago we had a similar use case when a PDF was converted to a PDF/A.

          OOTB you may setup "the other PDF" to be created by IBR when the first one is checked in (not necessarily as result of a workflow). It'll be stored as a web-viewable (or alternative) file of the original native file.

          You could also create another content item and link these two content items together - this would, however, require a customization.
          1 person found this helpful
          • 2. Re: PDF Optimization (web-optimized)
            Thanks for reply.

            Can this low resolution PDF file be created after the workflow is approved??
            Are their any configurations in IBR, where PDF optimizations can be specfied?

            I assume we will not require any 3rd party tools like Adobe Distiller for this purpose and it can be achieved using OOB WCC capabilities?

            • 3. Re: PDF Optimization (web-optimized)
              Are their any configurations in IBR, where PDF optimizations can be specfied?
              Actually, this is not a question directly about IBR. If you see this link: http://docs.oracle.com/cd/E23943_01/doc.1111/e10800/c03_pdfi_all.htm#sthref255 IBR is just a server, which processes conversions, but conversions themselves are performed by other technologies. Oracle now holds two of them: OutsideIn, and OpenOffice, but you can also use Microsoft Office or other 3rd party conversion tools.

              The mentioned chapter in the manual also provides information what parameters can be set up at the IBR side - unfortunately, resolution or size is not listed.

              You may have to get accustomed to other technologies - for OutsideIn (OIT) see http://docs.oracle.com/cd/E26550_01/index.htm

              I have checked it real quick and OIT has an option to modify resolution of graphical images...

              ...which leads to a question: what exactly is your use case? As far as I know PDFs can be with the textual layer, or graphical ones (a simple test: double-click a word in a file and when it highlights the word, it has the textual layer). I'm not sure if resolution can be changed for textual PDFs (except for few images, that of course might be contained). Also, what are the reasons for having a "smaller" PDF?
              Can this low resolution PDF file be created after the workflow is approved?
              It certainly can. I am not, however, aware of any method how to do this with a customization.

              Out of my head I can think of two methods:
              - conversions are performed during CHECKINs - you could customize these services not to perform conversions, create your own service that would do it, and call this service explicitly when an item is release from a workflow
              - read also http://docs.oracle.com/cd/E23943_01/doc.1111/e10800/c02_ibr_all.htm#sthref29 This contains an option: http://docs.oracle.com/cd/E23943_01/doc.1111/e10800/c02_ibr_all.htm#CBHBGGCF If you could a) modify the file format at the end of a workflow b) resubmit the item (it is possible, but only manually from the Repository Manager) you could achieve the required without modifying the services.

              Somehow, I don't like either of these methods. If feels there must be something simpler, but I don't know where to find it.
              • 4. Re: PDF Optimization (web-optimized)
                Thanks a lot for reply.

                I will try to analyze it using first option you provided and check the behavior:
                IBR to convert High res pdf to Low res pDF using OOB technologies ( OutsideIn and Open text).

                Only reason why i am looking for this optimization is to create web optimized PDF files. I do not expect to have PDF in specific resolution.
                I expect Web optimized PDF to be lesser in size and will be easily consumable if sent to some web address.

                Do you think if this is possible??

                Thanks for help!
                • 5. Re: PDF Optimization (web-optimized)
                  I expect Web optimized PDF to be lesser in size and will be easily consumable if sent to some web address.
                  Do you think if this is possible??
                  I don't see any reason why it should not be. In fact, even textual PDFs can be optimized - for instance, from our case I know that an approximate PDF/A is about twice that large than the original PDF (PDF/A is a format required for long-term content preservation, and it also contains e.g. true-type fonts so that the content can be rendered properly even on workstations without them). Another good example can be transformation of graphical PDF (scanned text) to a textual one - this is, however, more a task for OCRs than converters (WebCenter Capture from Oracle portfolio).

                  I think you should get sample PDFs so that you can estimate what conversion ratios, etc. you may expect. Unfortunately, all available conversion technologies (incl. Adobe-ones) might corrupt the information rendered in the document. For OIT, fortunately, you don't need to implement the whole IBR story - just download the product from http://www.oracle.com/technetwork/middleware/webcenter/content/downloads/oit-dl-otn-097435.html and it will contain a simple desktop application which may take a file and convert it. It also contains few sample programs if you need to set some parameters that are not available in the app.

                  Personally, I have some doubts if optimization of PDFs provides so good results to justify the effort - how much will your user experience improve if instead of 200KB PDF a 100KB PDF is returned? If you are fighting with bandwidth, you might consider dynamic technologies (such as Dynamic Converter) that could provide an output adjusted to connected users.
                  • 6. Re: PDF Optimization (web-optimized)
                    Srinath Menon-Oracle
                    Hi ,

                    Adding to Jiri's points , I think you should look at using Ghostscript as the distiller / conversion engine which has pretty good PDF Optimization features . Since that is the key requirement that would be the first option I would suggest .

                    • 7. Re: PDF Optimization (web-optimized)
                      Thanks for responses.

                      Our PDF's can contain a combination of text and Rich Image files.
                      PDF's size could be as big as 100MB, and in the past web optimized PDF resulted out to be 5MB 9max).

                      There are no specific perimeters which we are looking for. Only if PDF size could be reduced.
                      I am a little skeptical around use of 3rd party tools like (Ghostscripts) due to security restrictions but would prefer any OOB Oracle WebCenter technology which can accomplish same?

                      • 8. Re: PDF Optimization (web-optimized)
                        Srinath Menon-Oracle
                        Hi Amit ,

                        The Inbound Refinery and PDF Export will pull configuration information from the ix.cfg and px.cfg files. These values are used to determine what settings are used during the conversion of a document. Any of the default values in the ix.cfg and px.cfg can be overridden by adding the new value into the IBR's intradoc.cfg. For example, to improve the graphics output by setting the compression to none, add the following entry info the <IBR install dir>/bin/intradoc.cfg file


                        A list of all the available options can be found here: http://download.oracle.com/docs/cd/E14154_01/dev.835/e12885/e02_ixoptions.htm

                        To find out exactly what those commands are, go into the Inbound Refinery system, then to the Administration/System Audit Information page.
                        - Remove all Active Sections in the Tracing Sections Information section of the page, then add refinery and ref-config using the drop down box
                        - Check the box for Full Verbose Tracing and then press the Update button.

                        Go into the View Server Output link at the top of that page and press the clear button to clear out the logs.

                        Next go into the Content Server and check in a copy of the document that will be processed by the inbound refinery. Wait for that document to be released and then go back to the Server Output page on the Inbound Refinery system.

                        The log will show you exactly what command the IBR system used to convert the document. Search for the text "Launching exsimple" and it will show the format of the command used. It will list the source file, the output file, and the configuration settings used. This information could be used as a template to test the various conversion options quickly and easily.

                        Hope this helps .