Making a visual novel is a battle. Here at 4LS, it often is true both literally and figuratively. We have three contesting factions that constantly get into fights with each other when trying to cooperate: writing, art and directing. That's because making a VN is a game of resource control, and the only way to not lose is to not play. A VN is essentially a collection of pieces that we usually call "assets", assembled into a whole by the magic of code. Due to it being the most "expensive" asset to produce, art is treated as the finite resource. VNs are most often built around the idea of conserving art resources, reusing them where possible and cheating when you can. Everyone has to take into account the finiteness of art when working.
Thus, the bread and butter of any VN of KS's type are Two (or more) People Talking™ and the inner monologue of the protagonist. These are cheap to visualize: character sprites, backgrounds, you're done. However, whenever anything else than one of these happens (say, two characters kissing), there is a problem that has to be solved. Every line of the game script needs to be visualized, so it's up to the director to figure out the specific solution. The director can either bypass the issue, leaving the kiss to the reader's imagination (lame), or he could try to manipulate the existing art resources to create a passable representation of a kiss (uncanny), request a piece of art of the kiss drawn by the artists (expensive), or ask the writer to consider whether the kiss is really needed (problematic).
The necessity of resource-conservative visualization creates interesting problems to VN writers. We have to hold back ourselves with certain kinda of narrative description and place much more emphasis on dialogue and the protagonist's "inner monologue". As an inexperienced writer, I find this constraint very challenging to maneuver around and often end up making myself, our director, our artists or all three unhappy.
Now, directing emphasizes function over intrigue. An example from the development of KS: certain alternative outfit sprites that some characters get are drawn alternate poses as well, instead of drawing the outfit over the old sprite base. This was because drawing a new pose altogether was appealing to the artists. However, it made delta fly off the handle. The entire expression/pose palette, the "language" he had accustomed to using to the communicate the atmosphere, body language and mood of the characters was now unusable when they were in the alternate outfits. Similarly, artists might like to draw pictures of scenes or events that don't necessarily need them (in the opinion of the writers or director). This is always time not spent on drawing pictures that would be needed. And of course, sometimes the artists are just asked to do the impossible.
So, it's a battle. Writers want to do all kinds of quirky stuff, but then the director who wants to do all sorts of cool things with that will request bazillion pieces of art from the artists who want to draw altogether different quirky stuff and sooner or later, someone is driven to a panic attack, cursing everyone to the ninth hell or a week of heavy drinking. Essentially, this inherently conflicting setup solves itself in accordance to the attitudes and personalities of the creators, and their willingness to cause suffering to themselves and each others. Someone has to yield, either the writer who has to kill his brainchild, the artist who has to slave away more hours, the director who has to come up with alternative visualization or all three, when visualization is bypassed. I think we, when making KS, have a fairly high standard of visual representation, that is, we try to get as much stuff visualized as humanly possible (and often more). We also have a certain setup of personalities that amplifies this, leading us to often solve our own struggles in the way we do. We rarely lower the bar of our visualization standard, and usually prefer to either alter the script or get more art assets instead.
Now, this is not the only, and in my opinion, ultimately not even a very good way to go about making a VN. As with so many other things in this project, it just is the production model we ended up with and are sticking with. I think that a more experienced, a more functional team could have a better handle on how to deal with the problems I described. Meanwhile, we make do, and grind day after day with this project. At least when one of these arguments happen, it doesn't anymore lead to a catastrophic shitstorm like it used to. What do you know, even we learn.
» Discuss this post on the forums