Ubunchu Chapter Two, in the pipeline

oThose keeping track of Seo’s blog will know that he released chapter 2 of the popular Ubuntu manga comic “Ubunchu”, Those who can read Japanese should head over there and enjoy.

For all those waiting on the English versions (both edited Left to Right and un-edited Right to Left versions) stay tuned. We’ve pretty much got it all translated baring a few last minute tweaks and now I’m just putting it all together into svgs, editing, editing etc etc.

So today I will explain what we have done for this project so far:

Release

The first important part of this project was getting the release under Creative Commons license which allows derivative works. It would have been better to get it without the Non-Commercial clause; but maybe we can work on that with the publishers. Getting the PSD (Photoshop) files from the artist is also very important (see editing).

Transcription

Now it’s time to take each of the pages and transcribe the original text into a translation tool, we used google docs spreadsheets (we could have used Launchpad, but CC-NC prevents us from using lp). Someone who knows Japanese hopefully makes sure that all the text is labeled by Page, Panel and Person. All text is transcribed including sound effects and narrator notes.

First Translation

Screenshot-1Because we don’t have anyone that speaks Japanese and English well enough we have to do our translations in two steps. First the translator (Hito) takes the Japanese and produces a rough English translation called “Engrish”, he also makes translator notes in another column which gives cultural background and other note worthy information.

Second Translation

Now those who speak English very well (that would be me and Arturo) go through each of the Engrish translations and produce reasonable English that fits in context and in the boxes provided on the page. We are doing a number of passes between each of us because it’s important to get the flow of the English and make sure the comic doesn’t loose authenticity.

Image Editing

ScreenshotWe need to take the original Photoshop files and remove all Japanese. Because Seo was forward thinking most of the text is on separate layers which can be hidden, any mistaken or other text is carefully removed too using Gimp and any places where the original art was deleted for the text aura is repaired. These files are saved as gimp xcf files and a resized export is made to png.

LtR Editing

Lots of people won’t agree with the Left to Right editing, but since we’re making sure both versions are available I don’t think they have much to complain about. I like reading Left to Right. So flipping the new xcf files will produce a good result most of the time, you will find instances where there was some English or written name which now needs to be flipped or redrawn. In the case of chapter 02 I had to redraw the blackboard textures and redo one of the screenshots in order to get text the right way round.

Vector Production

I create the final result by importing each page image into an SVG file. I do this so that the language can be in xml and to make export to pdf easier. It also allows us to export a pot file which can be used to create translations with existing tools. Once I’ve got each page imported, I take the text and position it on each page and in each panel. The editing touches are to select font and redraw sound effect words (which are not usually from fonts).

Binding

My script outputs a PDF file and a source tar, but I also have a workbook version of the PDF that allows me to print off several copies for last week’s UDS in Barcelona. Very few people got a copy and it involves taking each page going from the first and last towards the middle and alternating between inside and outside pages. People who have done this to make books will recognise the pattern.

Hopefully we’ll be done with this soon and if you’d like to help translate it into your language we will make sure the link to the google doc is also released so everyone can join in and make a column for their language.

No Responses to “Ubunchu Chapter Two, in the pipeline”

  1. WindPower says:

    Would you mind sharing those “translator notes” (or the whole spreadsheet) with the rest of the world? :) I’d like to contribute to the french translation, I’m a native french speaker.

  2. doctormo says:

    http://spreadsheets.google.com/ccc?key=rp0ZWqR4oF45eCucUEtenPg

    Treat it like the spreadsheet wiki it is, with respect and create a column as required, try not to edit the English unless you’ve spoken to us first.

  3. WindPower says:

    I did a bunch of them. Would you please check and see if I’m doing everything correctly, before I continue?

  4. Zac says:

    Looking forward to it. Thanks people.

  5. Stani says:

    If you want *multipage* open source vector editing program, have a look at sk1:
    http://sk1project.org/modules.php?name=Products

    I’ve made it available through my PPA:
    http://pythonide.blogspot.com/2009/05/sk1-vector-graphics-editor-is-now.html

    sk1 has it roots in the old skencil project, just like inkscape with sodipodi. sk1 can import svg and inkscape uses uniconvertor under the hood, which was developed by the sk1 developers to convert between vector formats.

  6. doctormo says:

    I think I may want to keep using inkscape as my editor and use something else for the multi-page organisation.

  7. MenTaLguY says:

    I wonder if this would be a good time to mention minicomic, which is the Ruby stuff I use to bind The Last Bus. It’s currently in the middle of an ugly rewrite, but now that it looks like you’ll be doing more of this, it might be in our mutual interests to see about finishing it up…