OreCruncher
commented
8 years ago Implementation will do a survey in a 6 block radius (not square) around the player to figure out the biome demographic. Based on the results the volumes for the sounds from the biomes will be scaled based on their representation in the survey. For example, if there is a 50/50 split between forest and river biomes the Forest sounds will play at about 50% volume and the River sounds will play at 50% volume. (The actual scale factor calculated is a little complex and the 50% values serve to illustrate the point.)
Biome transitions can be abrupt when playing sounds. The sound for the old biome fades out and the new biome fades in. When there is a world with smaller biome clusters, such as what can happen with ATG the sounds transitions can be a bit edgy.
The idea is to do a biome survey around the current player position and to blend the sounds from all applicable biomes. The way the sounds would be blended would be based on the presence weight of the biome within the region applied as a scaling factor for the biome sound volume.
This would not apply to spot sounds - candidates will only come from the players current biome. It will only apply to sounds that could be considered a "sound track" for the biome.