Reputable Script Organization

Janaury 3, 2025

Hey, welcome to 2025! A new year, but work continues apace on Sleuthhounds: Cruise. This year looks to be a banner one for Cruise as it’s starting with the recording of the game’s dialog. Most characters have been cast and, rounding out 2024, I even already recorded the lines for three of them and have two fully integrated.

For the recording sessions, I’ve been doing what I can to make the recording scripts make sense for the people coming in. The story of Cruise takes place over the course of five acts with a variety of scenes from different acts happening in the same locations. The reason the locales are important is because of how the game’s code is structured. Each room in the game also represents a given programming source file. For example, the dining room has one source file that has all the code for that room in it. The atrium has one source file with all of its code. The brig… you get the idea.

This organization of the game code by room has been very conducive to programming the game. There are some exceptions where a scene starts in one room and tansitions to one or two others that causes a little bit of programming awkwardness, but otherwise it’s a good structure. Well, for programming, but not for recording dialog.

The problem for recording is that the different rooms have chunks of code from all across the game’s five acts. So, you might have a room that, in game code order, has a scene from act 1, then act 3, maybe a 2, then an act 1 again. Compounding this, players can make choices during the game that impact how the non-player characters respond to them. So a given scene may have two or three variations depending on if an NPC is friendly or annoyed with the player.

For the earlier Sleuthhounds games I had created a tool that extracted all the lines of dialog to be recorded in game code order. Those games were small enough and simple enough that there wasn’t a lot of jumbliness in the recording scripts produced. Any pieces of the recording scripts that were out of order, I could easily explain to the people that came in to record.

However, the size and complexity of Cruise means that a recording script presented in game code order is all over the place. Scenes from the end of the game might be right next to scenes from the beginning. Scenes where a character is angry with the Sleuthhounds might be right next to ones where they’re best of friends. It’s all rather chaotic.

To help reduce this problem, I’ve created another helper tool. After extracting the game code order recording scripts, I can now load them in this new Script Organizer tool. It breaks the “raw” scripts into a list of distinct scenes that I can then drag-and-drop around. This allows me to quickly reorganize the recording script to be more-or-less in story order. While I can’t quantify it, I know that this has made explaining the scenes to the voice talent much easier. It also largely separates sections of the game where an NPC is happy with players from sections where they're disgruntled based on the player’s choices. This helps the voice talent stay in the same mindframe for a character, rather than constantly switching between different emotional states.

To be sure, there’s still some jumbliness to the scripts. There are some scenes where there’s simply no good order to put them in. Or there are sections where only a line or two need to have nuanced differences based on the character reputations and so would be even more awkward to break out separately than have right in the middle of a scene. We still have to work around those in the recording booth, but overall having the recording script generally in story order is hugely advantageous.

Understandably, it takes a bit of time to reorganize the recording scripts from game code order to story order, even with a tool. I’ve been working on that a lot lately, and only have two more characters to complete that process for. All of this is in preparation for bringing the voice talent in throughout January to record. I have quite a number of audio sessions already scheduled, so it’s going to be voices, voices, voices for the next month or two.