In two weeks, it’ll have been three months since the last blog entry, so as exec producer it falls to me to periodically remind people we haven’t gone full tumbleweed. The obligatory update: Dev meetings have regained momentum despite a few pauses (including a dev’s recent relocation to the West—howdy, tenderfoot), but rather than bore you with writer fights over dibs on scenes, a project comment instead.
Over on a forum, people were discussing voiced versus unvoiced visual novels; it was again mentioned that a number of hardcore Japanese visual novel fans refuse to read unvoiced VNs, and one of the more well-known EVNs deliberately imitating JVNs did indeed catch some flak for its lack of voicework. It wouldn’t have been impossible to add it, but the risks outweighed the benefits (most importantly, the long-overdue project’s release date).
One of the allures of VN development is how accessible the tools are to the average person once you have a team able to provide the necessary assets, and how quickly those assets can be employed. Composers tend to have the proper musical software, artists already know their way around their weapons of choice, and the writers as usual have the unglamorous word processors. More importantly, most last-minute changes can be done on the fly, which is a marvel to those of us who used to work in print and remember the terror of opening a box and examining a print job for mistakes.
Voicework is a game-changer. Just as uncanny art can ruin a VN experience faster than bad writing will, or inappropriate musical direction breaks your immersion, mediocre voicework / bad audio engineering can end willing suspension of disbelief or even make good lines sound terrible. It’s telling that a company like JAST will convince a Japanese publisher to license a VN translation for a fraction of its original price, hire a TL team to get a good idiomatic translation and negotiate overseas with the publisher’s programmers to have the VN recompiled with the TL text, but they won’t invest in having those VNs revoiced.
While accounts vary, voicework is one of the most expensive aspects (in both literal costs and production time) of commercial visual novel development. In western video game development, a typical recording session is four hours long. A voice director in-studio requests multiple retakes with different moods and inflections so that editors down the line can choose the one they like best. Because of multiple retakes, the audio engineer’s going to end up actually only using a fraction of that session’s material (unless the actor is acing everything on their first take and convincing the director not to need retakes). It might well end up taking 3x the amount of spoken time to record the dialogue, if not more.
Say you have a best-case scenario: a great actor who can do multiple characters, has a solid home studio setup and delivers perfect takes the first time, every time, on schedule. You’re in the middle of beta testing, this close to release and a tester notes a glaring continuity mistake referenced in the dialogue; multiple times, as a matter of fact. The actor’s on vacation in the Bahamas. You can pinch hit a number of things when you have to, but most of us don’t have Jim Cummings or Billy West on speed dial to voicematch your actor.
There’s also a pragmatic consideration: people who read books (you know, the paper things) envision the characters in their mind, including forming a mental voice to match each of them, because that’s what our brains do when information’s missing. Conveniently, we never miscast that voice, either.
For the time being, most EVNs are unvoiced and the necessities of doujin production will likely keep it that way for a while.