The four phases of spell casting (when it comes to sounds) are “prepare”, “loop”, “cast”, “impact”.
Prepare and Loop go together, as they are played prior to the spell being cast.
Cast is when the spell actually happens.
Impact is when the spell makes contact with the target (ally, enemy, ground, object, etc)
Data for sounds in Spells
data "PrepareSound" "Spell_Prepare_Buff_Gen_L1to3" // Initially played at beginning of Prepare phase
data "PrepareLoopSound" "Spell_Prepare_Buff_Gen_L1to3_01_Loop" // Loops after PrepareSound, continues looping until CastSound
data "CastSound" "Spell_Cast_Buff_Blur_L1to3" // Plays when the spell is cast
data "TargetSound" "Spell_Impact_Buff_Blur_L1to3" // Plays when the spell impacts an object, character, ally etc
There is also a Vocal Component to go with VerbalIntent
data "VocalComponentSound" "Vocal_Component_Disadvantage"
data "VerbalIntent" "Buff" // Not 100% sure if this affects the actual speech
data "SpellFlags" "HasVerbalComponent" // Make sure this flag is on your spell if you want the verbal sound with it
Vocal Component Sounds
Verbal Intent Options
And finally a control for the loudness
data "SpellSoundMagnitude" "Big" // Small, Normal, Big, None
Data sounds in a status
data "SoundStart" "Misc_Status_AbsorbElementsActive_Start" // Plays when the status is first applied.
data "SoundLoop" "Misc_Status_AbsorbElementsActive_MO" // Plays on a loop throughout the status
data "SoundStop" "Misc_Status_AbsorbElementsActive_End" // Plays when the status is removed.
There is also a Vocal component for statuses.
data "SoundVocalStart" "NONE" // Plays when the status is first applied.
data "SoundVocalLoop" "LAUGHTERMANIACAL" // Plays on a loop throughout the status
data "SoundVocalEnd" "RECOVER" // Plays when the status is removed.