If anyone could identify the first kanji (6th character) for me that would be great.



Although its a horrible dvd subtitle font I've managed to identify some really dense complicated ones using Jim Breen and my Kanji Learners Dic, but for some reason this one's got me stumped.

Looks like 旅. Its keyword is trip (frame 1048)

Great, thanks! I was totally misreading it as flower on the top, with something like sword or power on the left. (I'm not nearly up to trip yet...)

Cheers, Robin

Here's another one, if anyone has a spare minute: the first kanji on the bottom row. Again it's not that complicated, but I think I'm misreading it, or I'm missing something about the stroke order.



25% through so far...!

If you're on a Mac, JEDict is a great program for times when you can't recognise a kanji. If you can recognise one ot two of the primitives, then it becomes easy to look up.
http://jedict.com/

I'm not sure about equivalent programs for Windows or Linux, but there are bound to be some that let you look up kanji by a combination of radials. In the case of JEDict, most of the Heisig primitives can be found, and even if you can only find part of a primitive, it's good enough to filter the list of kanji and then scan it for the one you want.

Thanks rgravina, I do have both a mac and windows, and I hadn't tried JEdict before. That "grapheme" filter function is really useful. Much appreciated!

I'm, still stuck though, the only one remotely close one it came up with was 事 matter 1156, which is not right, so if anyone else has any ideas...

That's 雪　（ゆき） snow, Heisig frame 1143.

Hurrah, thanks! I see my error now

Embarrassingly, I have a DVD of 春[kana]no[/kana]雪, the movie of the Yukio Mishima novel, sat on my shelf beside me - with the kanji in huge type. Oh well...

If you've got a DVD with Japanese subtitles, you might want to look into getting a DVD ripping program and just ripping the subtitles off into a seperate file, I done this once and when I got up to somewhere I was stuck, I just searched for the things around it that I knew until I found it, then just copied it into Firefox and ran Rikaichan over it (You could use and dictionary though.)

I use PAdict (http://padict.sourceforge.net) on my PalmPilot. It has a sort of character recognition mode that works pretty well, as long as you get your stroke order right.

ayoung24 Wrote:I use PAdict (http://padict.sourceforge.net) on my PalmPilot. It has a sort of character recognition mode that works pretty well, as long as you get your stroke order right.

Great, I'll check that out too. If it's anything like JEdict's character recognition, you have to get it bang on. On the other hand, Windows' recognition is amazing. If you're sure of the order it comes to the top of the list, but when you're guessing (because you haven't learned that primitive yet, and there's it has a stroke that always comes last, ot whatever) it still usually manages to come up with the one you want further down the list. In fact, for complicated characters where everything is crammed in closely, you can even make a scribble for some of the more cramped elements, and it still manages to get it more often that not.

krusher Wrote:If you've got a DVD with Japanese subtitles, you might want to look into getting a DVD ripping program and just ripping the subtitles off into a seperate file, I done this once and when I got up to somewhere I was stuck, I just searched for the things around it that I knew until I found it, then just copied it into Firefox and ran Rikaichan over it (You could use and dictionary though.)

That's pretty much what I was doing. But the ripping program I'm using is OCR-based, and needs to be trained on each new character set/font. Not such a big deal for english, but a bit more of a task for kanji! Not counting the kana, I think I trained it for about 900 kanji over the space of an hourlong movie. Of course you only need to train each character once, so it was painfully slow at the beginning, but very fast by the end.

If you or anyone knows of a ripper that works on Japanese without needing to be trained like this for every different subtitle font, please share. I'm still really glad I did it though, as I've learned loads about as-yet-unfamiliar primitives and stroke orders, and I think I've improved my kanji writing speed too.

The good news is I've completed that phase (220 subtitle frames) without having to ask for any more help. And now I've moved onto the translation phrase. I'm chuffed to bits that I seem to be able to do this, largely with the aid of JEdict and my Lammers grammar book, with my furigana dic and kanji learners dic as backups. And I'm surprised that it is actually enjoyable (although still mentally tiring).

I really am a language beginner (further behind than with kanji!), so for me I get a real sense of achievement. That is, 4 minutes (22 frames) of translation so far, which rather amazingly match back to the spoken version and the story context at the end of the process. It's pretty clear though, I have a long way to go until I can transcribe spoken Japanese without the benefit of subtitle files.

For me, doing stuff like this, and manga/novel translation for my own enjoyment, then ultimately being able to watch untranslated tv and read novels, are my goals. Not that it wouldn't be great to visit japan or live there, but it's a secondary goal.

Still got to see what my teacher thinks of the translation though - my confidence might get dented again!

Robin

You can uses this website, it does AJAX Kanji recognition:
http://chasen.org/~taku/software/ajax/hwr/

That's so cool, thanks. Imagine if [kana]FABURISU[/kana] incoporated something like that into the review page!!