How many times have you opened an ebook and noticed awkward hyphenations or other conversion errors? I still see this in the majority of the ebooks I buy and it’s clear these are the result of someone not paying attention during the conversion process. They may be minor annoyances but they reflect poorly on the publishers who produce them.
I recently had a chance to talk about this problem with Patrick Martinent, the CTO at Newgen KnowledgeWorks. They have a terrific platform called Silk Evolve that helps automate and reduce the errors when going from PDF to EPUB. The following Q&A is a preview to what you can expect to hear in Patrick’s session at next month’s TOC NY conference.
Tell us about Silk Evolve and why publishers need to know about it.
Silk Evolve is Newgen’s proprietary integrated product for high-quality, professional conversion of PDF to EPUB; operators with limited or no knowledge of EPUB, PDF, or HTML can produce EPUBs in less than one- tenth of the time that an expert would normally take.
Publishers always complain about lack of quality in the EPUB output format and issues such as hyphenation, word spacing, special characters, super scripting and sub scripting are much-publicized quality issues found in ebooks.
The key to getting the best quality output in the shortest possible time is in striking the right balance between automation and human judgment. The design theme of Silk Evolve is to automate as much as possible and to keep the operator focused only on those issues where his judgment is required, which in turn assures better quality and output.
What are some of the benefits Silk Evolve offers over other PDF-to-EPUB conversion tools?
There are many companies/products in the market that claim to have an automated solution, but these fall short in the aspect of quality and ability to handle complex books. Often, the generated EPUB needs hours of manual correction. This manual correction requires skilled people who are adept at HTML & CSS coding. Silk Evolve produces professional grade EPUB2 or EPUB3 books which very rarely require any further tweaking.
What user skills are required to become proficient with Silk Evolve?
As mentioned earlier, operators with limited or no knowledge of EPUB, PDF, or HTML can produce EPUBs in a fraction of the time required by an expert.
Yes. Silk Evolve generates EPUB3-compatible files that implement many of the standard new features. In particular, the native audio and video support and images in PDFs can be converted to SVG for better scaling. Math equations are rendered as MathML 8.
The Silk Evolve process has two stages: automation and manual. How much content is typically converted without the need for much manual intervention?
With Silk Evolve, PDF is converted to EPUB not by one but by multiple cycles of automation and user review. Though this sounds more manual, it actually results in fewer errors. We call it inprocess review as the user gets a chance to review the result at each stage. In each cycle, Silk Evolve automates and accomplishes a clearly defined task. But while processing, there might be some complex or ambiguous situations where the program is not sure of its results. So, after automation, Silk Evolve will highlight only those areas the user needs to review. As mentioned before, the idea is to strike the right balance between automation and human judgment.
Does Silk Evolve learn the specifics of a particular book series so that manual steps on the first book can become automated in future ones from the same series?
Of course, that is why we call it Silk Evolve: first, it is an evolutionary platform which integrates our learning from tens of thousands of books converted; second, the operator is equipped with many semi-automatic routines that will allow her to deal with unexpected scenarios. Furthermore, it is possible for advanced operators to write some scripts that Silk Evolve can execute and make global changes in one go.