
Welcome
Well, that's probably the most work I've ever gotten done over the course of a single Ludum Dare. Seance is my 20-somethingth collaboration with @alexrose, and it very well may be our best yet. It's definitely our most ambitious. A fairly complex story (especially by gamejam standards) written by Alex and mapped out to facilitate work between @lowri-bowden on the graphics, our 7 voice actors, and me on the audio editing/processing side. Over 400 lines of dialogue auditioned, edited, and processed. I personally handled 2+ hours of voiceover takes resulting in ~350 lines of dialog. There were some additional audio efforts from @sonnybone, who took care of processing his own voice for the detective, and @alexrose, who did some additional work on our friendly neighborhood malevolent being's voice. Scope creep is real, but we powered through.
Voice Casting
This was the first time we've attempted to create a game with anywhere NEAR this amount (or quality) of voice acting. We needed to find people who could record themselves at a sufficient quality, could get the tone of a character, and give a compelling delivery of their lines. We started off with a fairly broad call on Twitter that resulting in Mike Bithell(?!) saying he'd be down for a quick take or two.
Once we had Mike on the list we doubled down on trying to find the best voices we could for the other roles. We got some long-time Ludum Darers (@sonnybone, @sheepolution, @rincs, @naetoid) and some friends that we turned into first-timers (@chrissiecodes, @loriston4shoriston) to jump in. Everyone killed it with their roles. I think I've heard each character referred to as someone's favorite character (#RendellGang4Lyfe), which shows that we didn't end up with any weak points.
Additional thanks to @pomo and @blay09 for their work over the weekend trying vocal takes and helping make connections with some of our actors.
Art
Voice did a bunch of the heavy lifting in this game, but art was instrumental in pushing the quality forward - both in the final result, but also in giving people a visual to tune their vocal deliveries to. This was our second time working with @lowri-bowden and she absolutely killed it.
Music
I had a bunch of ideas for how to handle the music, but the rest of the work ended up eclipsing the stuff that I had in mind. We ended up with about a minute-and-a-half of music which, when spread over the course of an hour-long playthrough, means you hear the same thing a LOT. We added some dynamic controls to allow some variation depending on who you're talking to, which helps out considerably, but I had planned to lean into this dynamic control a lot more. I wish I hadn't focused as much on my (failed) solo entry in the early part of the weekend or that I had figured out some efficiencies in the vocal editing earlier so I could have spent more time focusing on the music.
I still REALLY like the Soundtrack that I created, I just wish there was more of it.
Audio Automation
It took MOST of the jam period, but eventually I got a suitable workflow for processing the 2+ hours of audio takes I had to process. Initially I started with a VERY manual process to ghostify all the voices (pre-verb in particular required extra steps) - make sure the voice ROUGHLY matched the tone and volume of the other voices, find the take I wanted, normalize it, bounce it, reverse it, add processing for pre-verb, un-reverse it, add additional processing, export as wav, convert to ogg. About halfway through the first characters takes I realized this was COMPLETELY unsustainable. I ended up developing an Audacity macro that allowed for me to automate the bulk of this process.
Audio Editing
Once I had the macro sorted out, I still had the much smaller (but still fairly large) job of vocal matching and mining the raw audio for the best takes. This was COMPLETELY new territory for me, so I had to learn as I iwent. Initially it took an ABSURD amount of time for me to properly process, edit, and bounce the clean audio takes. After hours and hours of doing things inefficiently, I eventually got a workflow that simplified things. By the final set of takes, I was able to power through half-hour of dialogue in right about an hour. If I had figured out my process earlier in the weekend I would have saved SO much time, but I got there. If we do something this absurd again in the future, I'll be prepared.
The End
Overall, I'm incredibly happy with how this complex of a project ended up pulling together. Huge shout-out to the entire team - everyone did their part to elevate the quality of what we attempted. Extra huge shout-out to @lowri-bowden for killing it on the art and helping formulate this story. And hugest shoutout to @alexrose for creating such a clear structure for all of us to be able to do our parts effectively and then tying it all together into a cohesive project at the end.
Check out our Game
Check out the Soundtrack