How does PDF/VT relate to 'V'ariable data and 'T'ransactions

Question

Could someone explain to me how PDF/VT 'supports' Variable and Transactionsl data ?  I'm new to PDF/VT and I'd expected that it somehow made it easier to insert variable and transactional data, but I haven't found anything of that nature.  I do see that VT supports Document Part Hierarchy and metadata.  Is this the improvement related to VT - better ability to manage the documents within a single PDF file?
There also appear to be a number of other improvements, such as direct ripping of PDF without translating to Postscript first,  JDF-aware metadata, etc.  These improve print operations, but I don't see how they improve variable data handling.
Any pointers would be appreciated.
Thanks,
bob

Answer 1

Good questions. I'll endeavor to provide a response.
The most important thing to understand about PDF/VT is that in and of itself, it is a final form file format. It represents the fully composited output stream of a VDP job. There are no “variables” in PDF/VT itself. PDF/VT provides the protocols for final form PDF to optimally respresent such an output stream.
PDF/VT-1 is built on top of PDF/X-4. All PDF/VT-1 files are totally valid PDF/X-4 files. PDF/VT-2 is based upon PDF/X-4p, PDF/X-5pg, and PDF/X-5g files - in the latter two cases possibly referring to Reference XObjects in external PDF/X-4 files.
There are two major areas in which PDF/VT supports VDP (and so-called “transpromo”) print streams:
Data optimization - space and RIP performance. Repeating objects, whether considered “fixed” or “variable” by whatever generates the PDF/VT, are represented via Image XObjects, Forms XObjects, and for PDF/VT-2 also Reference XObjects. Objects are defined once and then referenced many times. This can dramatically cut down the size of PDF files used for VDP. Furthermore, in conjunction with RIP technology optimized for processing PDF/VT files, such as the technology provided by products using the Adobe PDF Print Engine, use of XObjects is automatically recognized and triggers caching mechanisms that allow the RIPed results of XObjects to be cached in the RIP's high speed memory avoiding unnecessary repeated RIPing of common content. PDF/VT metadata also provides for “hints” to assist the RIP in such caching. This type of optimization makes it possible to use PDF for printing extremely graphically-rich color VDP jobs on very high speed (hundreds of pages per minute) digital presses now being developed and released to the market by digital printer manufacturers.
Metadata.In addition to providing caching hints, PDF/VT provides metadata support for tagging pages or sets of pages for operational considerations. For example, one could tag each “set” of pages associated with the statement for each customer of a financial institution and via such metadata be able to directly access the pages associated with one statement (“view or print the statement for Jane Doe of Anywhere, New Jersey”) or multiple statements (“view or print the statements for all left-handed, herniated, Californians who own a Toyota” assuming of course you have all that metadata on the pages associated with particular account statements). As another example, you could tag via metadata for pages of certain types such as “front covers,” “inside pages,” and “rear covers” where each page type is printed on a different paper stock or size and possibly on a different printer (one could print “all front covers” for “all statements for accounts in ZIP codes 11010 through 11199”). For printing purposes, typically these types of selective printing requests are handled via JDF which instructs the RIP to select and process pages based upon such PDF/VT metadata and other JDF commands.
Direct printing of PDF without conversion to PostScript has been occurring for a number of years now, especially starting with traditional offset print platemaking and now becoming much more prevalent for high-end digital presses, typically those used for VDP jobs. (Direct PDF isn't yet on typical desktop devices. PDF is still printed to PostScript from Acrobat on most such low-end devices.) PDF/VT really forces the need for direct PDF RIPs, but wasn't the only requirement for same. PDF/X-4 and PDF/X-5 were defined very much as a response to the needs for content at the highest level of abstaction (live transparency, color management, layers, etc.) for direct PDF printing.
Hopefully this has begun to answer your questions. Let me know if there is further detail that you want or need.
          - Dov Isaacs (Chair, ISO TC130/WG2/TF2 PDF/X and Co-Chair, ISO TC130/WG2/TF3 PDF/VT)