RunningHub workflow page: https://www.runninghub.ai/zh-cn/post/2098775138884300801?inviteCode=rh-v1111
Bilibili tutorial video: https://www.bilibili.com/video/BV1vFYz6HECU/
# SelfLift — 4K-target post-processing
## Purpose and preparation
Resolve the matching custom-node/runtime versions and assets. Load authorized media and inspect active inputs, actual schedules, effective dimensions and every save destination before a small test.
Only image 164 is active at slot 1; image 150 is bypassed despite its retained slot 0 link. Check prompt ordinal interpretation.
The generation selector requests 16:9,1.03MP,multiple32; links override cached1344×768. Duration 5 feeds the alignment expression, requesting124 frames, nominal about 5.167 seconds at 24fps. Exact raster and duration remain unverified.
The active LightX2V LoRA is named 4step and has strength about 0.65, but BasicScheduler stores simple/8/denoise1 before H3SigmaRefiner(extra_steps 1,start 0.7,end 0,cosine). Do not present this as a measured four-step run or infer total NFE. LowVRAMAttention and ChunkFeedForward nodes are bypassed.
SelfLift stores CFG 1,transition 6,lowres_scale 0.4,rho 0,w_min 0.5,w_max 1,BF16 learned latent upscaler and highres_tiling=false. Negative conditioning is ZeroOut. Preview 1024/12fps is not export resolution/FPS. Native generated audio feeds assembly/interpolation; source drive/final/reference-audio ports are unconnected and conditioner mux_audio is unused.
Decoded images pass through SeedVR2, linked target2160, then a resize request3840×2160, Lanczos, keep_proportion=resize, divisible_by2, CPU. Audio bypasses spatial processing. These are post-decoding targets, not native generation dimensions or measured results. Cached resize diagnostics are not render evidence.
Interpolation requests 2x,motion auto,threshold0.24,keep_audio on. This is temporal, not spatial scaling. For generation presets,24→48fps is configured intent, not measured cadence/frame count/duration. Combined VIDEO and explicit images/audio are connected; precedence requires implementation review. VIDEO output feeds SaveVideo with auto format/codec.
## Files and selected assets
- [Workflow JSON]()
- [Selected cover]()
- `minimax_h3_hybrid_fl2va_ref2va_b25-49-int8.safetensors`
- `qwen3vl_32b_minimax_h3_int8_convrot.safetensors`
- `minimax_h3_audio_vae_fp32.safetensors`
- `minimax_h3_video_vae_fp16.safetensors`
- `minimax_h3_fl2v_lightx2v_turbo_4step_v0.1_comfy.safetensors`
- `seedvr2_ema_3b_fp8_e4m3fn.safetensors`
- `minimax_h3_latent_upscaler_3d_bf16.safetensors`
- `taeh3.safetensors`
## Version, evidence and rights
This is a local descriptive draft, not a published upgrade. Files01–08 are byte-identical to originals; only the standalone ninth file's historical video-preview field is cleared in its new copy. Originals remain unchanged.
No model inference, visual/audio acceptance, wall-clock benchmark or license verification was performed. Covers are selected assets, not accepted render evidence. Weights and external input media are not bundled. Check workflow, nodes, models, LoRAs, source media and cover rights separately; no blanket commercial license is asserted.
[Source-provided tutorial](https://www.youtube.com/watch?v=R3W2ab1_TzY) · [Source-provided workflow entry](https://www.runninghub.cn/post/2098689766622580737/?inviteCode=rh-v1111)
Links are not live-verified. The workflow is supplied through AI-KSK.