How Results are Displayed
Searches result in the listing of pairs of related Japanese and English.
For example, searching for the English
motorcycle'' might result in,
The response would be completely different if you weren't using a raw
Japanese text encoding (you're using
now). Click here
to see what this page would look like if you were in
gif image mode (the default).
- /Shoei (motorcycle helmet manufacturing company)/
- $BFsNX/two wheeled vehicle (bicycle, motorcycle, etc)/
If you didn't have Japanese support
The Japanese text would have to be sent as an image,
and so therefore wouldn't
know about the colors (foreground/background) and size of your browser's font.
In using this method (which you aren't now), the
default appearance of the Japanese text images in query results is white
text on a transparent background (if your viewer supports the GIF89a
standard, a black background otherwise), with characters 26 pixels square.
You can change these defaults by selecting the appropriate items in the
form on the main dictionary
page, if you enter it in image mode.
In using this method (which you aren't now), you
can select ``white foreground'' or ``black foreground'' from the
Japanese text colors selection on the main page. The background
will be transparent (i.e. use your window's background color) if your
viewer supports it, the foreground's opposite otherwise. You can also
choose to have a background colour present instead of a transparent background
to do some funky things, but there's no accounting for taste so don't blame
me. Going one step further, you can also chose to have the image inverted,
that is, the foreground colour can become transparent, with the background
colour taking on the former foreground colour. Note however that if you
define a background colour other than black, then the colours will be reversed
(think reverse video), not the transparency To get a transparent foreground,
select black as the background colour and check the ``invert colors'' checkbox.
Also on the main page, you can select from a number of font sizes:
18 pixel, 26 pixel, 48 pixel, 64 pixel and 96 pixel.
Selecting a size is a compromise between a number of things:
The kanji for nihongo (``the Japanese language'')
can be rendered by this server as small as
or as large as
You'll probably find that one is a bit easier to see than the other.
As it is, since you're in
mode, the fonts and such be set
by you at your client.
Visually pleasing integration with your local (i.e. English) fonts.
characters from larger fonts take more time to
transfer, although it seems that most of the time is in the per-image
transfer overhead. In practice, a 48 bit font should result in
images about four times larger than a 16 bit font, but access
doesn't seem to be all that much slower.... the main time cost is in the
The fastest, of course, is using a standard encoding method, such as
you're using now.
Getting a closer view
In raw Japanese mode, there's no need to view Japanese text as images,
but sometimes it might be nice to see a complex character written in a
really large font. Toward that end, you can click on the
Anchors to Large Japanese in the main window. This will turn each
kanji entry of a dictionary query result into an anchor to a gif of a
48-point image. You can then call up the large images at your leisure.
You can have this feature turned on for you automatically using the
customization features of this server.
There's also the option of having the server emit Japanese text as images
in a vertical manner. In the server's case, this is done by rotating each
character counter-clockwise 90 degrees before generating the image. While
this leaves you the work of rotating the whole image clockwise using
an image manipulation program, doing it this way saves from creating
a messy presentation on the screen. As this process is a bit
calculation-intensive, you'll need to wait longer for the images to be
generated as they load into your web browser.
Comments appreciated [Return to Index]
[Return to Main Page]
(this page's master source last modified 10 years, 3 months ago)
This reply to request 121,218,176 made just for you Wed Nov 22nd 2017 4:09am JST [load currently averaging 19118 requests/day over a 235-second sample]