Nodes/IronLeo78 MTV/IronLeo78 2in1 Revitalizer
ComfyUI Node

IronLeo78 2in1 Revitalizer

All-in-one: loads a first-pass latent + conditioning saved by the IronLeo78 MTV, upscales the video latent (the optional Minimax H3 Latent Upscaler 3D when installed, else a built-in latent resize), refines it (guider + sampler + sigmas), decodes video and audio, builds the video (link the video output to a Save Video node: this node does not save it). by IronLeo78

By ironleo78-sys·Created 26 days ago·Updated 17 days ago· 3
IronLeo78 2in1 Revitalizer
    • video
    • images
    • audio
    • latent
    • status
    ◄latent▾►
    ◄conditioning(match the latent)►
    ◄freeze_audiotrue►
    ◄unet_name(none found in models/diffusion_models)►
    ◄vae_video(none found in models/vae)►
    ◄vae_audio(none found in models/vae)►
    ◄lora_ontrue►
    ◄lora_name(none)►
    ◄lora_strength1.00►
    ◄attention_backendcomfy kitchen attention►
    ◄upscale_ontrue►
    ◄upscale_model(none found in models/latent_upscale_models)►
    ◄megapixels0.4►
    ◄temporal_chunkingfalse►
    ◄force_unloadfalse►
    ◄upscale_devicecuda►
    ◄upscale_precisionfp16►
    ◄seed1978►
    ◄sampler_nameres_multistep►
    ◄sigmas0.9231, 0.8780, 0.8000, 0.6316, 0.3158, 0.0000►
    ◄LoRa_1_enabledfalse►
    ◄LoRa_1None►
    ◄LoRa_1_strength1.00►
    ◄LoRa_2_enabledfalse►
    ◄LoRa_2None►
    ◄LoRa_2_strength1.00►
    ◄LoRa_3_enabledfalse►
    ◄LoRa_3None►
    ◄LoRa_3_strength1.00►
    CategoryIronLeo78/MTV

    Inputs (29)

    NameTypeDefaultDescription
    latentCOMBOThe first-pass latent (<clip>_video.latent). Lists ComfyUI/output (newest first) and uploaded files. Use the Upload button to add files from anywhere.
    conditioningCOMBO(match the latent)'(match the latent)' loads the <clip>.conditioning saved with the selected latent. Pick another file to use a different clip's conditioning.
    freeze_audioBOOLEANtrueON: the audio latent gets a zero noise mask (as in the MTV), so the refine pass does not re-noise the song.
    unet_nameCOMBO(none found in models/diffusion_models)Diffusion model (UNETLoader).
    vae_videoCOMBO(none found in models/vae)H3 video VAE.
    vae_audioCOMBO(none found in models/vae)H3 audio VAE.
    lora_onBOOLEANtrueThe 'LoRa OFF/ON' switch of the workflow (model-only LoRA, e.g. the 4-step turbo).
    lora_nameCOMBO(none)The LoRA of the workflow (the 4-step turbo). A saved set puts the Model Loader's Turbo LoRA file here.
    lora_strengthFLOAT1.00-100–100Strength of the LoRA above. A saved set puts the Model Loader's Turbo LoRA strength here (v0.1.11).
    attention_backendCOMBOcomfy kitchen attentionModel Attention Backend. 'comfy kitchen attention' falls back to PyTorch attention by itself when it is not available.
    upscale_onBOOLEANtrueOFF: refine at the saved size (no Separate / Upscaler / Concat). ON: with the optional Minimax H3 Latent Upscaler node + a model file installed, that trained upscaler is used; without them the latent is resized by this node and the refine pass redraws the detail (the status says which).
    upscale_modelCOMBO(none found in models/latent_upscale_models)Only used when the optional Minimax H3 Latent Upscaler node is installed (models/latent_upscale_models). Without it the built-in latent resize is used.
    megapixelsFLOAT0.40.1–2Target size of the upscaled video, in megapixels (max 2.0).
    temporal_chunkingBOOLEANfalse—
    force_unloadBOOLEANfalseUnload the upscale model to CPU afterwards to free VRAM.
    upscale_deviceCOMBOcuda3 options: cuda, rocm, cpu
    upscale_precisionCOMBOfp163 options: fp16, bf16, fp32
    seedINT19780–18446744073709550000Use the clip's own seed to repeat it (the refine pass should use the same seed).
    sampler_nameCOMBOres_multistep44 options: euler, euler_cfg_pp, euler_ancestral, euler_ancestral_cfg_pp, heun, heunpp2, +38
    sigmasSTRING0.9231, 0.8780, 0.8000, 0.6316, 0.3158, 0.0000The refine schedule, numbers separated by commas, going down from left to right (N numbers = N - 1 steps). Default: the 5-step refine.
    LoRa_1_enabledoptBOOLEANfalseLoRa_1 ON / OFF (default OFF). Applied after the LoRA above.
    LoRa_1optCOMBONoneLoRa_1: LoRA file from ComfyUI/models/loras/ (same as the Model Loader's LoRa_1).
    LoRa_1_strengthoptFLOAT1.00-100–100LoRa_1 strength (default 1.0).
    LoRa_2_enabledoptBOOLEANfalseLoRa_2 ON / OFF (default OFF). Applied after the LoRA above and after LoRa_1.
    LoRa_2optCOMBONoneLoRa_2: LoRA file from ComfyUI/models/loras/ (same as the Model Loader's LoRa_2).
    LoRa_2_strengthoptFLOAT1.00-100–100LoRa_2 strength (default 1.0).
    LoRa_3_enabledoptBOOLEANfalseLoRa_3 ON / OFF (default OFF). Applied after the LoRA above and after LoRa_2.
    LoRa_3optCOMBONoneLoRa_3: LoRA file from ComfyUI/models/loras/ (same as the Model Loader's LoRa_3).
    LoRa_3_strengthoptFLOAT1.00-100–100LoRa_3 strength (default 1.0).

    Outputs (5)

    NameTypeDescription
    videoVIDEO—
    imagesIMAGE—
    audioAUDIO—
    latentLATENT—
    statusSTRING—