What’s been going on:

1. I am still working on Versailles no Bara/Rose of Versailles, which I first mentioned long ago.
2. Clannad is … still in progress >_> (ask IX!)
3. I’ve been doing more reading, and will soon get off my ass to do more writing on editing and typesetting.

I mention #3 here to give myself some pressure so I don’t keep putting off these long-delayed posts. If I’m not going to be releasing subs, I should at least write useful things to justify the existence of this blog …

First: a follow-up to The Reason … Is Because. Expect this within the week.

In an earlier post, I commented that Eve no Jikan presents some rather interesting challenges in timing. Here, I’d like to discuss one of these challenges in more detail and explain how we tackled them.

Rapid lines

Even before Eve no Jikan, from shorts like Mizu no Kotoba, Director Yoshiura demonstrates a fondness for sequences of short, quick, jarring clips; the camera snaps from position to position quickly, and audio cuts in and out. This tends to create scenes with very short dialogue lines; one such scene is shown below (at around 0:31:39 in the movie).

One thing that irked me about the typesetting and timing in that clip is how some of the lines merge together visually, especially around 3 or 4 seconds into the video. Lines with similar lengths and very short screen time appear almost similar as they appear one after the other, and viewers would hardly notice the change if they were focusing one’s eyes elsewhere on the screen.

Separate dialogue lines

I had a short discussion with Hale on the issue. His preferred method is to leave a short time gap (1–2 frames) between the lines, creating a “flashing” effect that alerts the viewer to the line change subconsciously. This is one way of doing it, but I didn’t feel comfortable removing 2 frames of screen time from a line that was already only about 12 frames long (~0.5 seconds). In addition, I personally dislike the “flashing” effect and try to avoid or eliminate it at every opportunity.

Some inspiration struck; DVD subtitles do often put dialogue lines for two characters on screen, indicating them with prefixed hyphens (although typographically they ought to be dashes). What if I do that, but use separate line timing to do the demarcating?

Transitions between lines are much clearer now, but the lines are still as short and difficult to read. What can we do about that?

Line pairing

Here, I pulled in a timing technique I’ve been using on and off with other projects: line pairing. Some short dialogue lines don’t get much screen time before the next line appears. Keeping these lines on screen while the next line appears allows slower viewers to read the next line without missing the previous line. This works well in certain dialogue situations, e.g. “Why? [‘Nande?’ Usually a very short line]” “Because you didn’t wait for me at all.”

With line pairing, the result looks much more readable:

Visual cueing is tricky business. Timers need to ensure that line changes are visible from the viewers’ peripheral vision, yet at the same time we don’t want to distract the viewer with excessive visual signals as well. Do let us know other alternative solutions you’ve seen in use.

Slight update, for those about to kill me for spamming their RSS. I have a *job* now, and would like to keep my subbing work as surreptitious as possible. I also intend to write about more topics on Taipu, not all of them subbing-related. Therefore I’ll be moving many of my typesetting posts over to ZUM instead.

Over the next few weeks I’ll be going through the migrated posts and cleaning up the formatting to better fit this blog. Drop me a note if you spot any inconsistencies. If you’d like to contribute a guest post on typesetting, let me know as well.