ComfyUI Node
🐧 DualSamplerCustomAdvanced
Samples a packed audio-video latent (e.g. MiniMax H3) with a separate sigma schedule per stream.
🐧 DualSamplerCustomAdvanced
- noise_video
- noise_audio
- guider
- dual_sampler
- video_sigmas
- audio_sigmas
- av_latent
- output
- denoised_output
- video_output
- audio_output
- denoised_video_output
- denoised_audio_output
CategorySuperNodes/Sampling
Inputs (7)
| Name | Type | Default | Description |
|---|---|---|---|
| noise_video | NOISE | — | |
| noise_audio | NOISE | — | |
| guider | GUIDER | — | |
| dual_sampler | DUAL_SAMPLER | — | |
| video_sigmas | SIGMAS | — | |
| audio_sigmas | SIGMAS | Use the same number of steps as video_sigmas. | |
| av_latent | LATENT | Packed video + audio latent. |
Outputs (6)
| Name | Type | Description |
|---|---|---|
| output | LATENT | — |
| denoised_output | LATENT | — |
| video_output | LATENT | — |
| audio_output | LATENT | — |
| denoised_video_output | LATENT | — |
| denoised_audio_output | LATENT | — |