
Powered By RTX 4090
Just released open-source ACE Step music generation and Float digital human
Performed track separation to help the digital human better recognize vocals
Thus accurately matching lip movements.
Local 8G VRAM is sufficient to run
Detailed usage explanation in the comments
晨曦之光
2025-06-24 Update
晨曦之光
2025-06-24 Update
Workflow introduction
Just released open-source ACE Step music generation and Float digital human
Performed track separation to help the digital human better recognize vocals
Thus accurately matching lip movements.
Local 8G VRAM is sufficient to run
Detailed usage explanation in the comments
Nodes Information
24
LoadImage
ACEStepGen
AudioSeparation
CR Text
Float
FloatProcess
GenerationParameters
GetNode
ImageCASharpening+
ImpactSwitch
Int
LoadAudio
LoadFloatModels
MultiLineLyrics
MultiLinePromptACES
Note
PreviewAudio
PrimitiveFloat
RH_Translator
SaveAudio
SetNode
VHS_LoadVideo
VHS_VideoCombine
workflow>1