Issue link: https://digital.copcomm.com/i/1546054
110 M PS E . O R G 3, we did a lot of outdoor, literal field recording. We did huge weapons recording sessions and built up a pretty robust library of content. We didn't really need to go back and reinvent the wheel for weapons for Borderlands 4. We did need more of the nuanced Foley for the order soldiers, as well as new layers for gore. So we did a lot of recording inside of the studio instead of going outside to do huge multi-mic weapon sessions. We recorded a lot of individual elements, for instance, for every single Vault Hunter or vehicle. We aimed to record at least one or two organic layers for those. So we deployed a lot of individual sessions for those types of things. EM: Other than the volume of work, how did this being an open world game affect some of your team's choices? JP: How we handled acoustics had a big effect on our plans for that. In previous titles where we had smaller levels, we would pre-bake acoustics data, and then at runtime, we could look up information about, "if you're standing there, these are the properties." That wouldn't work in an open world game. We needed an acoustic system that, in real time, would make assessments and good choices about how the environment should sound. We invested a lot of time building a ray tracing system that could estimate how close walls were, what the roughness factor was, what materials they were around to absorb sound, and how enclosed the space was. That system was a big change from previous titles. JL: Yeah, there's a handful of workflow changes we had to do to support the open world. We had to rethink our population of ambiences, and like Julian said, how acoustic system would scale when you've got kilometers of space that are changing every day. With a baked system, we didn't know how good the results would sound out of that. I did a prototype of an acoustic system to show Mark and Brian. The idea was "What if we did this based on the feedback you had about the old stuff, and what you wanted to hear aesthetically?" Then I worked backwards to figure out the simplest way I can achieve this, and then added in complexity as more information came in. Having a real-time system was super important because there's no way I'm going to try to chase my tail, placing volumes onto everything in the game. I don't want our team to chase that down. So a dynamic system was really the way to go. Ambiences were very similar—instead of placing emitters manually, I worked with our programmers to figure out how we could be more procedural. How can we make rules to say, "If these meshes or particle systems are here and have these properties, they should be making this sound." We try to cover as many broad strokes as possible, and then when fine-tuning is really needed like a really bespoke dungeon area, then we could get more nitty gritty and hand place those sounds. MP: The other challenge is the amount of time a player is traversing. For ambiences, we used opportunistic Borderlands 4 gameplay