Eve library

ISIS v2 Voice Line Pack Control-Conditioned Workflow Graph

Guide

Revision 1

On this page

Generated: 2026-03-03T12:08:02.637Z Checklist: MXA-0544 (ISIS2-00742) Asset Family: voice-line-pack Phase: Phase 4 | Milestone: M3 Launch (2026-07-27 to 2026-10-30)

Objective#

Define a control-conditioned workflow graph for voice-line-pack that applies pose/depth/segmentation guidance where applicable (with optional canny/lineart enrichment), while preserving deterministic synthesis defaults and mastered-dialogue output contracts.

Deterministic Defaults#

Setting Value
modelName xtts_v2_voiceset_a.safetensors
promptTemplateId voice-line-pack-control-conditioned-v1
styleProfileId voice-line-pack-control-guidance-v1
seed 2026073104
durationSeconds 8
speakingRate 1
pitchShiftSemitones 0
conditioningStrength 0.68
voiceStability 0.74
sampleRateHz 48000
targetLufs -18
outputFormat wav
outputChannels 1
outputBitDepth 24
controlApplicabilityMode where-applicable
controlSignalFusion weighted-sum
providerCapabilityMode required
minimumEnabledControlsPerRun 1
poseStrength 0.36
depthStrength 0.62
segmentationStrength 0.54
cannyStrength 0.48
lineartStrength 0.45
articulationAlignmentMode consonant-priority

Input Contract#

Input Type Required Default
pose_map image no n/a
depth_map image no n/a
segmentation_map image no n/a
edge_map image no n/a
lineart_map image no n/a
use_pose_control boolean yes false
use_depth_control boolean yes true
use_segmentation_control boolean yes true
use_canny_control boolean yes false
use_lineart_control boolean yes false
pose_strength float yes 0.36
depth_strength float yes 0.62
segmentation_strength float yes 0.54
canny_strength float yes 0.48
lineart_strength float yes 0.45
control_applicability_mode enum yes "where-applicable"
control_signal_fusion enum yes "weighted-sum"
provider_capability_mode enum yes "required"
control_profile_id string yes "VOICE_LINE_CONTROL_PROFILE_V1"
control_timeline_fps integer yes 24
articulation_alignment_mode enum yes "consonant-priority"
cue_id string yes "VOICE_LINE_M3_CONTROL_CONDITIONED_01"
script_text string yes "Commander, reroute auxiliary power to deck three and hold for extraction."
speaker_id enum yes "narrator_f"
language_locale enum yes "en-US"
emotion enum yes "urgent"
speaking_rate float yes 1
pitch_shift_semitones float yes 0
duration_seconds integer yes 8
text_prompt_hint string no "Keep diction clear and syllable timing stable while integrating control-guided prosody and contour emphasis."
conditioning_strength float yes 0.68
voice_stability float yes 0.74
speaker_style_reference_id string no "VOICE_LINE_STYLE_ANCHOR_V1"
voice_consistency_profile_id string no "VOICE_LINE_CONSISTENCY_PROFILE_V1"
prompt_template_id string yes "voice-line-pack-control-conditioned-v1"
style_profile_id string yes "voice-line-pack-control-guidance-v1"
sample_rate_hz integer yes 48000
target_lufs float yes -18
output_format enum yes "wav"
model_name string yes "xtts_v2_voiceset_a.safetensors"
seed integer yes 2026073104

Graph Nodes#

Node ID Class Role Output Count
n1 LoadImage pose-control-input 1
n2 LoadImage depth-control-input 1
n3 LoadImage segmentation-control-input 1
n4 LoadImage canny-control-input 1
n5 LoadImage lineart-control-input 1
n6 VoicePoseControlAdapter pose-control-adapter 1
n7 VoiceDepthControlAdapter depth-control-adapter 1
n8 VoiceSegmentationControlAdapter segmentation-control-adapter 1
n9 VoiceTransientControlAdapter canny-control-adapter 1
n10 VoiceContourControlAdapter lineart-control-adapter 1
n11 VoiceControlSignalMixer control-signal-fusion 1
n12 VoiceLinePromptComposer prompt-construction 1
n13 VoiceLineModelLoader model-load 1
n14 VoiceLineSynthesisNode audio-generation 1
n15 VoiceLineCleanupDenoise audio-post 1
n16 AudioLoudnessNormalize audio-post 1
n17 SaveAudio artifact-output 0

Graph Edges#

Edge ID From To
e1 n1:0 n6.control_map
e2 n2:0 n7.control_map
e3 n3:0 n8.control_map
e4 n4:0 n9.control_map
e5 n5:0 n10.control_map
e6 n6:0 n11.pose_control
e7 n7:0 n11.depth_control
e8 n8:0 n11.segmentation_control
e9 n9:0 n11.canny_control
e10 n10:0 n11.lineart_control
e11 n12:0 n14.prompt
e12 n11:0 n14.control_signal
e13 n13:0 n14.model
e14 n14:0 n15.audio
e15 n15:0 n16.audio
e16 n16:0 n17.audio

Topological Node Order#

n1 -> n12 -> n13 -> n2 -> n3 -> n4 -> n5 -> n10 -> n6 -> n7 -> n8 -> n9 -> n11 -> n14 -> n15 -> n16 -> n17

Research Alignment#

Workflow Path Control Nodes Audio Nodes Declared Control Inputs Node Classes
AnimateDiff with ControlNet libs/isis/workflows/src/workflows/video/animatediff-controlnet.json 3 0 AnimateDiffCombine, AnimateDiffLoader, AnimateDiffMotionModelSettings, CLIPTextEncode, CheckpointLoaderSimple, ControlNetApply, ControlNetLoader, ControlNetPreprocessor, EmptyLatentImage, KSampler, LoadImage, LoadImageSequence, LoopProcessor, SaveImage, TemporalConsistencyProcessor, VAEDecode, VHS_VideoCombine
Image to Video Animator libs/isis/workflows/src/workflows/video/image-to-video.json 0 0 AnimateDiffCombine, AnimateDiffLoader, AnimateDiffMotionModelSettings, BLIP_Caption, CLIPTextEncode, CheckpointLoaderSimple, ColorGrading, ImageResize, KSampler, LoadImage, LoopProcessor, RepeatLatentBatch, SaveImage, TemporalConsistencyProcessor, VAEDecode, VAEEncode, VHS_VideoCombine
Voice Performance Synthesis libs/isis/workflows/src/workflows/audio/voice-performance-synthesis.json 0 3 AudioDeesser, KokoroModelLoader, KokoroPromptBuilder, KokoroTextToSpeech, SaveAudio

Acceptance Criteria#

  • Graph is acyclic and topologically sortable.
  • Pose/depth/segmentation control paths are explicit and operate where applicable without blocking unsupported lanes.
  • Canny/lineart supplementary controls are optional and routed through explicit adapter nodes when provided.
  • At least one control mode can be enabled per run while preserving deterministic seed/model/prompt/style defaults.
  • Articulation alignment and control-signal fusion policies are first-class contract inputs.
  • Output contract aligns to MXA-0541 required output classes for voice-line-pack.

Traceability#

  • Checklist item: MXA-0544
  • Ticket key: ISIS2-00742
  • Source references:
    • docs/domains/isis/remediation-v2/ISIS_REMEDIATION_V2_TICKETS.json
    • docs/domains/isis/remediation-v2/ISIS_V2_VOICE_LINE_PACK_REQUIREMENTS_BRIEF.json
    • docs/domains/isis/remediation-v2/ISIS_V2_VOICE_LINE_PACK_BASELINE_WORKFLOW_GRAPH.json
    • docs/domains/isis/remediation-v2/ISIS_V2_VOICE_LINE_PACK_IMAGE_CONDITIONED_WORKFLOW_GRAPH.json
    • docs/domains/isis/remediation-v2/ISIS_V2_CONTROL_TAXONOMY.json
    • docs/domains/isis/remediation-v2/ISIS_V2_WORKFLOW_AUTHORING_GUIDE.json
    • libs/isis/workflows/src/workflows/video/animatediff-controlnet.json
    • libs/isis/workflows/src/workflows/video/image-to-video.json
    • libs/isis/workflows/src/workflows/audio/voice-performance-synthesis.json
    • docs/domains/isis/remediation-v2/ISIS_V2_AUDIO_CONDITIONING_SUPPORT.md
    • docs/domains/isis/remediation-v2/ISIS_V2_OUTPUT_PACKAGE_TAXONOMY.json
    • ISIS_TODOS_2.md