Editor for EPUB w/SMIL on Windows?

[karyan]

I would like to create EPUBs with SMIL for media overlay - but on a Windows platform. For readers, I have used both AZARDI and Readium's Chrome add-on to test existing samples, but have not found an editor that I can use to create new works.

I wish I could use SIGIL, but I can't seem to produce an EPUB that either reader recognizes as having audio. My next attempt will be with Tobi (from DAISY consortium), but before diving in I was hoping that someone could tell me that this is even possible and I'm not wasting even more time.

Thanks.

UPDATE: I did play around with Tobi a bit. On the plus side, it's a lot easier for syncing the audio to text than manually writing the SMIL file using Audacity to find the start/stop points. The negative is that the xhtml files needs to be pretty much perfect before starting this step - they can't be edited in Tobi (from what I can tell). The resulting EPUB worked (mostly, except if I tried to jump around within a chapter the audio-to-text goes out of sync by a few seconds) in Readium/Chrome, but doesn't work in AZARDI.

[karyan]

So...I've been playing around with Tobi/Sigil/Notepad. It's been a bit frustrating.

Tobi is great for creating the .smil files and it exports the .epub and audio files nicely. BUT I can't edit either the .xhtml or .smil files directly in Tobi. So, it makes in impossible to adjust text selection based on the narrator's phrasing. Not a problem if I only wanted to do whole paragraph selection - which I don't.

So, export everything, make the adjustments manually and re-import, right? Not exactly. Tobi doesn't re-import the .smil files, so each correction to the .xhtml files means re-doing ALL the audio-to-text.

Is there an easier way? (Right now, I'm looking at manually editing the content.opf file.)

[Toxaris]

Since there is no real epub 3 editor really available, you're stuck at doing it manually.

[mrmikel]

There is not much incentive to create one either. There are few devices/apps/programs and those using them are likely to have problems resulting in endless tech support for the creator of such a program.

[karyan]

It's not that I object to doing some of the work manually. The problem is that even that doesn't work. The only thing that I've found that works is syncing at the paragraph level - which is NOT what I want to do.

I know I'm not explaining this well. Basically, I need a way to arrive at a working EPUB using SMIL. As I said, Tobi "loses" the existing SMIL when importing a EPUB. Sigil "corrupts" the EPUB 3 (trying to make it "correct") when I need to make small tweeks in the XHTML.

I was hoping that someone could actually explain how they edit the SMIL/XHTML etc. to arrive at a working EPUB using SMIL.

[Toxaris]

Well, since Sigil is an ePUB2 editor it is not that strange that Sigil tries to correct an ePUB3 file.

If you know of a working example, you can always take it apart. For now you are stuck with doing it manual.

[karyan]

Originally Posted by Toxaris

Well, since Sigil is an ePUB2 editor it is not that strange that Sigil tries to correct an ePUB3 file.

If you know of a working example, you can always take it apart. For now you are stuck with doing it manual.

It's not a question of having a working example. Several are available. (Just don't save them after viewing in Sigil or they will be "corrected", too. Or import into Tobi.) Nor am I saying that either Tobi or Sigil are behaving other than in a way that they are intended to do.

What I am asking is, can ANYONE explain (at a high level) how THEY (personally) produced a working EPUB using SMIL?

I have done this - painfully - using Tobi and Sigil - but every small change to the XHTML loses ALL the SMIL information so I have to start again.

[DiapDealer]

Yep. I'm going to be the lazy guy who asks someone to explain what the heck the acronym SMIL refers to instead of googling it my own big myself.

[karyan]

Originally Posted by DiapDealer

Yep. I'm going to be the lazy guy who asks someone to explain what the heck the acronym SMIL refers to instead of googling it my own big myself.

SMIL (pronounced “smile”) = Synchronized Multimedia Integration Language.

In this case, I use it to match different parts of an mp3 file with sections of text so that that text is highlighted as the corresponding audio plays.

[mrmikel]

Let it be noted that SMIL is only opened by media players and media software. But he wants an XHTML editor.

Adobe GoLive supports SMIL. But it probably wants to make web pages, not epubs.

[AlPe]

Originally Posted by karyan

What I am asking is, can ANYONE explain (at a high level) how THEY (personally) produced a working EPUB using SMIL?

I feel your pain. Essentially the best tool out there is Tobi, and you need to integrate it into your workflow to get the most of it.

[Self-Promotion deleted - MODERATOR]

I developed an automated tool where you pass in the text fragments and the audio files, and you get back the SMIL files. Clearly this is integrated in our ebook workflow --- e.g., the text fragments are <span>-marked automatically at word/phrase/paragraph level.

Some CC-licensed examples you can freely download and examine:

Divine Comedy, 7h, ~15K SMIL fragments, Italian: http://www.smuuks.it/index.php/en/pr...f-lino-pertile

The Curious Case of Benjamin Button, 1h, ~1K SMIL fragments, English: http://www.smuuks.it/index.php/en/pr...enjamin-button

PM if you want to discuss additional (private) details.

[karyan]

Originally Posted by AlPe

I feel your pain. Essentially the best tool out there is Tobi, and you need to integrate it into your workflow to get the most of it.

Thank you. It's good to know that what I want to do is actually possible.

I've made more progress. After fixing the epub, especially the .opf file, with a plain text editor, it can now be load into Tobi with a clean EPUB 3 check and without losing the SMIL files. (And it runs using Readium, but not Azardi.)

So I have a 90K word document synced at the paragraph level and a second version that I started to break into sentences.

[Hitch]

Originally Posted by karyan

Thank you. It's good to know that what I want to do is actually possible.

I've made more progress. After fixing the epub, especially the .opf file, with a plain text editor, it can now be load into Tobi with a clean EPUB 3 check and without losing the SMIL files. (And it runs using Readium, but not Azardi.)

So I have a 90K word document synced at the paragraph level and a second version that I started to break into sentences.

We don't use a tool; we make our SMIL's by hand. We have to synch word-to-word for iBooks, so I can't help you with any shortcuts for synching to paragraphs/lines/larger chunks, although I can't see that it would be spectacularly different. {shrug}. We make the SMIL-synch portion of the ePUB last, for the same reasons that you've highlighted herein; no editors for the whole package.

Moreover, there are so few readers for this type of product, although Alpe's company apparently made the investment to create a tool, I wouldn't do it at mine. If we had a much larger demand for this type of work, of course, we would, but the demand for it, at least at my shop, is negligible. We did 24 titles in the fall-winter of last year, and another....5? 7? this year, so we have made about the same number as Alpe's shop, which I think probably is indicative of demand. The 24 last winter was anomalous, I would normally not expect our shop to do more than 10 a year, not at this point in time, not until more reading devices/apps support it.

Best of luck to you!

Just my $.02.

Hitch

[AlPe]

To put things in perspective: we started a partnership with a publisher of audiobooks (i.e., CD or MP3 digital download), and the 30 title mentioned above are part of an initial batch of ~70 titles, which are being released bi-weekly.

Meanwhile, other three (Italian) publishers have become interested in the product, so I guess the wave has been started. Plus, I think that the potential market is huge, especially for "language X as second language" titles.

Originally Posted by Hitch

not at this point in time, not until more reading devices/apps support it.

You are correct. Current Reading Systems are bad (in general), and simply atrocious when it comes to supporting audio (in EPUB3), especially for audio-ebooks like ours. I even wrote a JS library ( https://github.com/pettarin/rb_smil_emulator ) to emulate Media Overlay support in iBooks, which does not provide it in reflow mode...

Fortunately, w.r.t. RS support, I know that things are going to change for the best pretty soon...

[daniel.weck]

Hi all,
Tobi can (re)import text-only as well as text+audio EPUB3 and DAISY talking books (both standards are based on SMIL, but use different XML document formats: HTML5 versus DTBOOK). The SMIL text-audio synchronisation data is indeed preserved during the import process (i.e. when creating a new Tobi project).

As correctly stated though, the DTBOOK / HTML5 markup cannot be edited directly from within the application, making it hard to adjust the synchronisation granularity "on the fly" / at authoring time (e.g. split a large paragraph into several sentences, or even down to the word level for children read-aloud books). We have plans to introduce granularity control / structure editing, early next year. Until then, painstaking roundtrip export+reimport is the only workaround. We generally advise to "prepare" the text structure in advance (which is quite a common practice in production workflows), but we totally appreciate the need for built-in structure editing, and we will address this caveat.

As for Media Overlays reading systems: the Readium Chrome plugin is currently useful, but not completely satisfactory (e.g. missing previous-next audio navigation, tap-to-play, skippability support, etc. ... features which are also lacking in iBooks and others). Fortunately, great progress is being made in the Readium SDK (the aforementioned shortcomings are already fixed), so hopefully these improvements will shortly be available in the Readium Chrome extension (or more likely: in the Readium-Web library).

Feel free to use the http://daisy.org/tobi forum if you encounter problems when using Tobi, we do monitor user feedback and queries there.

Cheers, daniel