Tsunami MT, Version 6.0

NeocorTech adds more Japanese text editing, learning features to their translation software

I've always been wary of automatic translation software, especially when Japanese is the source or target language. Even though translation programs have come a long way in the last few years, they are far from perfect.

I was skeptical when I heard the hype concerning Tsunami MT, NeocorTech's English to Japanese translation software for 32-bit Windows. But after a closer look, I found that Tsunami MT is by far the best PC translation tool I've ever used.

The technology underlying Tsunami is impressive. The translation engine was developed by Japan's Kamejima Co., which has over a decade of translation software experience and holds 70% of the automatic translation market in Japan. And it shows.

Tsunami's translations are quite fast, regardless of the topic. I translated a number of files on number of subjects, ranging in length from 500 to 2,000 words. The longest took just over three minutes to complete.

The translations, while better than expected, sometimes suffer from an affliction common to all translation software: words and phrases are converted too literally. Certain words won't translate at all; these are left in English. The program will sometimes choose a Japanese word when it should have been converted to Katakana. In one document about the World Wide Web, Tsunami translated "Web" as ami (a net or web), rather than keeping it as it or converting it to Katakana. If you're using Tsunami to translate HTML documents, be aware that it converts certain tags (e.g.&ltHEAD> and &ltLINK>) to Katakana.

There really is no way to make the translations perfect. Tsunami's manual gives tips on how to improve accuracy. And you can configure how the program translates verbs, nouns, adjectives and even the way it displays interrogatives. All this is helpful, but you'll find that output will have to be edited from within Tsunami or in a Japanese word processor or text editor.

Tsunami also contains a number of useful utilities: the front-end processor, the Kanji Reader and e-mail support.

Tsunami's front-end processor (FEP) allows you to easily edit translated documents from within the program. However, it doesn't work like most FEPs I've tried. In fact, I struggled with it until I read the manual. After that, I was able to quicly master it.

The Kanji Reader is a powerful tool for learning Kanji. You simply highlight a character or a compound and press a hotkey. A dialog box appears containing the reading and English translation. The Kanji Reader can only handle single Kanji or two-character compounds. Try it with longer compounds and you'll get an error message.

With a mouse click, Tsunami can e-mail your translated document and /or the original English. It's simple, too. Just click the e-mail button and in the dialog that pops up enter the address of your recipient and a subject for the message. This feature is perfect for sending a translation to a colleague for proofreading. The only drawback is that the recipient will need to have Japanese e-mail software or a program capable of rendering Japanese characters.

One aspect of the program I didn't like was that you can't import files created in popular word processors. The only way to get documents into Tsunami is to copy and paste, manually type them or import text files. This wipes out all formatting. You'll have to open the file in a Japanese word processor or text editor to reformat it.

Tsunami MT professional will help cut translation time and costs. If used properly, and with a professional translator, it can be an excellent bridge between your business and Japan.

- Scott Nesbitt

 ęCopyright 1997 Multilingual Communications & Technology. All Rights Reserved


Click here for Neocor in the News Page

Click here for NeocorTech's Home Page

Microsoft Internet Explorer

Powered by BackOffice

Last modified by NeocorTech Webmaster: March 10, 1998.
Copyright 1998 NeocorTech LLC. All rights reserved.
Ne.o.cor.tex : n. The dorsal region of the cerebral cortex.