Troubleshooting K3 chinese font support spotty?

[willyu34]

Hi everyone,

I am just wondering if any international owners are having the same kind of issue as me.

I got a K3 on firmware 3.0 (515460094) and when I read traditional Chinese mobi and txt files, approximately 10% of the letter doesn't show up (shows up as a square, missing character). I checked the file on my PC and they are fine. The text file I can confirm as unicode UTF-8. The mobi is a traditional chinese EPUB that I made converted with Calibre. Calibre can read it no problem.

I heard that K3 is having some problem with Korean font too?

It should be a software issue, but I am not sure if I am the only one or not.

thanks,

[ichbindasauge]

Are the missing characters, the ones that are different in simplified characters? Maybe they only included simplified characters? Wouldn't make any sense though if they are working with Unicode. I don't have a K3 myself to try out.

[willyu34]

Originally Posted by ichbindasauge

Are the missing characters, the ones that are different in simplified characters? Maybe they only included simplified characters? Wouldn't make any sense though if they are working with Unicode. I don't have a K3 myself to try out.

K3 does support both traditional and simplified chinese. It's in the specifications. The traditional text I used, the text does show up as traditional and not simplified, but some common character seems to drop out. I haven't tried a simplified Chinese so I can't comment on that.

I think I get what you mean because I know that one simplified chinese character usually covers more letters than a traditional character. So your guess that perhaps it's missing that "extra" covered character. However, I don't think it's the case..

[Kanru]

Do you have any sample for testing?

[HelenaJole]

Originally Posted by willyu34

I heard that K3 is having some problem with Korean font too?

I don't think there are any problems with it actually displaying (not that I've heard, anyway), it's just that the font is hideous.

[willyu34]

Hi Kanru,

I have included the text file I used here. This is dl'ed from big5.qidian.com and I have made sure it's UTF-8.

You should notice those "squares" for missing character on almost every page.

anyone else who's a chinese reader wtih K3 give it a try. It would be greately appreciated.

thanks.

[ichbindasauge]

Can you or somebody else provide a screen shot using K3?

[chrichen]

I experienced the same thing as well. After trying different conversion tools but failed, I sent the UTF-8 txt files to "name"@free.kindle.com and they all got converted properly, both Simplified and Traditional. But the titles still may show boxes. Seems like their 'native support' is not exactly what I had in mind.

I don't recall seeing the following when I ordered the K3, but now K3's description on Amazon has the following:
"Support for Non-Latin Characters
...
To view your personal documents with non-Latin characters on your Kindle, send your file as a Microsoft Word document (DOC) attachment to your Kindle's e-mail address ("name"@free.kindle.com)...Loading TXT files containing non-Latin characters over USB is currently not supported as some characters may not display properly."

[Kanru]

Maybe it's just the font has not enough characters.

Because on my Kindle DXG w/ Font Hack the sample file renders pretty well.

[belem]

Press any keys in Home UI, then it will show Search box.

1. Type ;debugOn , press Enter key
2. Type ~changeLocale zh-CN , press Enter key
3. Restart Kindle in Settings

Should be fine to show simplified Chinese.

[Kanru]

OK. I think the myth is busted.

According to https://wiki.mobileread.com/wiki/MOBI there is a “locale” field in the mobi header record. By default if the metadata doesn't specify the locale, it is set to “0x0009”. But a .mobi created by Amazon or Mobipocket the field is set to “0x0404” correctly for Traditional Chinese (zh-tw), or “0x0804” for Simplified Chinese (zh-cn).

I have tried to make Calibre output the correct locale and it works! Try add a metadata <dc:language>zh-tw</dc:language> to the epub file then convert again.

[essebibi]

Hallo Kanru, could you explain in details the steps?

For example, I have a chinese pdf that I see correctly on the Kindle 3, but I want to convert it into a mobi file...Obviously it shows many "boxes" as discussed.

- When Calibre converts the pdf in mobi, in the directory I can see the metadata.opf file with the <dc:language tag>. But at this point I think it's too late: the conversion already has been done. Am I right?

- If I look into the mobi file (hexadecimal) I can see the 0xFD09 (65001 = UTF-8) locale bytes

- How can I tell Calibre to convert the file using the correct locale as you suggested?
There's a place where I can force it?

Many many thanks!

[essebibi]

...Needn't anymore...I decided to do the fonts Hack...Xie xie

[Kanru]

Use the locale hack would solve the issue.

Why the "send to device" by Calibre action corrupts the file is still a mystery.

[essebibi]

As told, with the fonts hack all now is working properly and the hack install is quick and easy.

Many thanks to NiLuJe and the other people that prepared all the packages!