- Our game will have multiple voice lines with X amount of variations (“hello”, “hi!”, “hey!”)
- Those lines will be shared by different characters (child, girl, guy, etc)
- We’re planning to have 2 locales.
Given that creating a huge amount of keys and/or having to parse the VO content in order to randomize/switch things by code could take a long time to do it right, I’m thinking on not using this at all and going with the “sound effect” approach (so I can use multi instruments for randomizing and parameters for switching between characters, etc). Example: I can have a “VO_Hello” event and randomize different lines and can change the character with a parameter.
Is it reasonable to do this? I suspect that, even if there are other things to take care about (like handling locales differently), the amount of effort could be much lower.
I might be missing some things here, so I’m looking forward to hearing your thoughts!
It is reasonable, yes. Plenty of games have done it.
By implementing your dialogue in FMOD Studio, you’ll more easily be able to use FMOD Studio’s features in conjunction with your dialogue, reducing the amount of time required to iterate on your design.
This increased ease comes at the cost of flexibility, however. As you have noted, it means you will not be able to use audio tables to handle localization; you will have to build any localization into the project yourself. It also means an audio asset will only be able to play in an event if you specifically add it to that event first.
Every game project has different needs and requirements, so only you can judge whether this trade-off is worth it for yours.
Thanks Joseph. Yes, in the end it’s all about the project needs. I really like how easily FMOD handles locales, and I see it could work well for specific dialogues (no randomization at all), but I also find other ways to work with locales, perhaps by using a parameter.