Another post to update you with news etc. 🙂
(** IMPORTaNT ** This time it will affect the future of the translation process ** IMPORTANT **)
- Cleaned version of Chapter 3 has been released (several naming change).
- Perhaps you have noticed the title in the Chapter 7 released today is different as the usual title. Well, this time I used another translation method (Edited MT). Please read the details below.
- Besides Chapter 7, I will also release some more chapters (probably to Chapter 10) while I’m waiting for the polling result.
- So during the wait for the results, for this week my schedule would be only releasing those sample Chapters that was translated with the Edited MT method, and editing+cleaning Chapter 4 to 6. I think it’s a win-win solution 🙂 I have time to finish the editing, and while you wait, you can still enjoy the story (although with a less accurate translation).
As I mentioned last week, I did some coding to help me with the translation pace. But since I seldom refer to machine translation when I translate manually, it turns out that what I code would not help me speeding up at all.
Throwing my effort away just like that would surely be a waste, so I did an experiment today. And the result is the edited MT version of Chapter 7. Basically this is what I do in this Edited MT method:
- I put all the machine translation together in a file grouped by paragraphs (I use 4 machine translator: Baidu, Bing, Google and SysTrans).
- Then I read all 4 translations for the paragraph, combining the different translations to get the whole picture in my head, also trying to guess what the original text might be saying.
- Afterward, I rewrite the whole paragraph with my own words based of my assumptions of the translation.
Now, some reminder. Doing this translation method, I didn’t refer to the original Chinese text at all. So, if all the machine were translating it incorrectly, my translation would also be incorrect, of course. And since this version currently only serve as a sample for you to preview, I also didn’t do any spell checking or proof reading at all.
When I did the translation today, I tracked the time needed. And with my current manual translation speed (400 characters / hour), this method is about 3x faster than the manual translation method (it was because of this much difference that I decided to post it and asking your opinion).
Personally, I prefer to read a good translation, like ‘Coiling Dragon’ translated by Ren at WuxiaWorld (I have read some translations that were so stressful to read since it was obvious that the translator also used machine translator, but he only did little editing, just enough to make it more humanly) or Zhan Long (although there are times when it was obviously using machine too. Well, there are several people translating it and they are trying to keep their publishing quota 🙂 ). I’m also aware that most of you might have a busy RL schedule and doesn’t care whether it’s a good/bad translation, as long you can grasp the story.
That’s why, I’m asking for your feedback. The translation are made for you to read, not just for my personal use. Here are the options:
- Manual Translation Method
- Slower pace, on average only 1-2 chapters / week (on a really relaxing week, I might manage to release 3 chapters at most).
- The Draft version itself is already going through the check spelling and proof reading process before I published it (all the Draft version that I published so far had also been going through this process before I published them).
- More accurate translation, since I’m reading the Chinese text directly. Sometimes, it might require some term changes, but the overall accuracy should be above 90%.
- The editing and cleaning process will follow later on after the Draft has been published. In this process, I recheck the translation with the original text, and also do some rewording to make the sentence easier or better to read.
- Edited MT Method
- Faster pace, my estimation would be a minimum 5-7 chapters / week (and on really relaxing week, I might probably even release two chapters daily).
- No spell checking or proof reading at all.
- Less accurate translation, since I won’t cross checked it with the original Chinese Text, so the accuracy depends on the machine translation accuracy.
- No editing or cleaning. I just concentrate on releasing the next chapters and leave the published chapters as it is.
- Semi Combined Method (well, I thought of it just now when I was writing this post)
- Fast, but not as fast as the second method. 5-7 chapters / week would be the maximum number of chapters I could currently publish (I still have my RL work).
- The Draft version will go through spell checking and proof reading process.
- Accuracy would be the same as Edited MT method since I won’t cross checked the Draft version. (Unless the paragraph is too confusing to translate, then I might check the original text)
- Instead of re-translating the chapter manually for the editing/cleaning process, I will only cross check the translation. Unless the meaning were totally wrong, I might only do minor changes or even no changes at all. And this editing/cleaning process would be done after the Draft is published, since I would focus more on releasing new chapters. (That’s why I called it SEMI Combined method)
Reminder: Just to avoid any misunderstanding with the Semi Combined Method, please note once again, I will cross check the translation with the original text AFTER the draft is published (during the editing/cleaning phase) or only if the machine translation are too confusing.
P.S. Until it’s decided, there are no next/previous chapters navigation in the post with the machine translation. You have to go back to the machine translation index to read other chapters (when they are available).
Please guys vote for your preference. I would only take you two clicks 🙂 Thank you!
Which Translation Method Should I Use?
- Semi Combined (56%, 62 Votes)
- Edited MT (28%, 31 Votes)
- Manual Translation (16%, 18 Votes)
Total Voters: 111