ComfyUI Node
MD: Audio Auto Master Pro
A ComfyUI node in MD_Nodes/Audio Processing with 44 inputs and 8 outputs.
MD: Audio Auto Master Pro
- audio
- audio
- analysis_details
- yaml_config
- waveform_before
- waveform_after
- spectrum_plot
- dynamics_plot
- lufs_history_plot
◄target_lufs-14.0►
◄profileStandard - Balanced all-purpose mastering►
◄output_modeMastered Audio►
◄enable_ai_helpertrue►
◄genre_hint►
◄ollama_urlhttp://localhost:11434►
◄ollama_modelqwen2.5:14b►
◄debug_mode1 - Info►
◄enable_profilingfalse►
◄yaml_config►
◄export_yamlfalse►
◄input_gain_db0.0►
◄spectral_tilt0.00►
◄vocal_tamer_strength0.00►
◄harmonic_exciter_drive0.00►
◄fix_sub_mud_db0.0►
◄fix_kick_thump_db0.0►
◄highpass_freq0►
◄lowpass_freq0►
◄do_eqtrue►
◄eq_bass_target9.5►
◄eq_high_target5.5►
◄eq_adaptivetrue►
◄max_iterations_eq5►
◄do_deesstrue►
◄deess_amount_db-10.0►
◄do_mbctrue►
◄mbc_crossover_low300►
◄mbc_crossover_high3000►
◄mbc_crossover_order8►
◄mbc_low_thresh_db-24.0►
◄mbc_low_ratio2.5►
◄mbc_mid_thresh_db-22.0►
◄mbc_mid_ratio2.5►
◄mbc_high_thresh_db-20.0►
◄mbc_high_ratio2.0►
◄do_limitertrue►
◄limiter_threshold_db-1.0►
◄soft_clip_drive1.00►
◄stereo_width1.00►
◄fast_modefalse►
◄skip_initial_analysisfalse►
◄mix1.00►
CategoryMD_Nodes/Audio Processing
Inputs (44)
| Name | Type | Default | Description |
|---|---|---|---|
| audio | AUDIO | AUDIO INPUT • Purpose: Unprocessed audio waveform to master. • Requirement: Standard ComfyUI AUDIO dict. | |
| target_lufs | FLOAT | -14.0-30–-6 | TARGET LOUDNESS • Purpose: The final perceived loudness target (LUFS). • Options: -14.0 (Streaming), -23.0 (Broadcast). ⭐ Recommended: -14.0 |
| profile | COMBO | Standard - Balanced all-purpose mastering | MASTERING PROFILE • Purpose: Automatically sets dozens of DSP parameters. • Options: 'Standard', 'Diffusion Repair' (fixes AI noise), 'Podcast'. ⭐ Recommended: 'Diffusion Repair' for raw audio generation outputs. |
| output_modeopt | COMBO | Mastered Audio | OUTPUT MODE • Purpose: Defines what audio is sent to the output node. • Options: 'Mastered' (Final result) or 'Delta' (Only what was changed). ⭐ Recommended: Mastered Audio. |
| enable_ai_helperopt | BOOLEAN | true | AI CO-PILOT • Purpose: Queries a local Ollama LLM to fine-tune EQ based on analysis. • Requirement: Ollama must be running locally. ⭐ Recommended: True for experimental/creative runs. |
| genre_hintopt | STRING | GENRE HINT • Purpose: Text clue to help the AI Co-Pilot make better EQ decisions. | |
| ollama_urlopt | STRING | http://localhost:11434 | OLLAMA URL • Purpose: Endpoint for the local LLM API. |
| ollama_modelopt | STRING | qwen2.5:14b | AI MODEL • Purpose: Model used for Co-Pilot reasoning. |
| debug_modeopt | COMBO | 1 - Info | LOGGING VERBOSITY • Controls console logging and AI explanation detail. |
| enable_profilingopt | BOOLEAN | false | ENABLE PROFILING • Track execution time of LLM vs DSP stages. |
| yaml_configopt | STRING | YAML CONFIG • Purpose: Paste exported settings here to override all GUI controls. | |
| export_yamlopt | BOOLEAN | false | EXPORT YAML • Purpose: Outputs the final computed settings as YAML text for saving. |
| input_gain_dbopt | FLOAT | 0.0-36–36 | INPUT GAIN (dB) • Pre-processing volume adjustment. |
| spectral_tiltopt | FLOAT | 0.00-12–12 | SPECTRAL TILT • Extremely sensitive macro EQ. • +0.05 = Brighter, -0.05 = Warmer. |
| vocal_tamer_strengthopt | FLOAT | 0.000–2 | VOCAL TAMER • Purpose: Dynamically cuts harsh 1-3kHz resonances common in AI voices. |
| harmonic_exciter_driveopt | FLOAT | 0.000–2 | HARMONIC EXCITER • Purpose: Tube-style saturation for warmth. Use sparingly (0.05 - 0.20). |
| fix_sub_mud_dbopt | FLOAT | 0.0-36–0 | FIX SUB MUD • Purpose: Low shelf cut (75Hz) to remove boominess. |
| fix_kick_thump_dbopt | FLOAT | 0.00–12 | FIX KICK THUMP • Purpose: Targeted narrow boost (90Hz) to restore punch. |
| highpass_freqopt | FLOAT | 00–1000 | HIGHPASS FILTER • Cut frequencies below this point (Hz). |
| lowpass_freqopt | FLOAT | 00–22000 | LOWPASS FILTER • Cut frequencies above this point (Hz). |
| do_eqopt | BOOLEAN | true | ENABLE ADAPTIVE EQ • Auto-balance the spectrum to targets using Librosa FFT analysis. |
| eq_bass_targetopt | FLOAT | 9.50–20 | EQ BASS TARGET • Desired low-end energy distribution. |
| eq_high_targetopt | FLOAT | 5.50–20 | EQ HIGH TARGET • Desired high-end energy distribution. |
| eq_adaptiveopt | BOOLEAN | true | ADAPTIVE MODE • Dynamically scale EQ adjustments based on input deviation. |
| max_iterations_eqopt | INT | 51–20 | EQ ITERATIONS • How many analysis/adjustment passes to reach perfect balance. |
| do_deessopt | BOOLEAN | true | ENABLE DE-ESSER • Dynamically reduces harsh 'S' sounds in the 7kHz range. |
| deess_amount_dbopt | FLOAT | -10.0-60–0 | DE-ESS AMOUNT (dB) • Maximum intensity of sibilance reduction. |
| do_mbcopt | BOOLEAN | true | ENABLE MULTIBAND COMPRESSOR • Enables independent 3-Band dynamics processing. |
| mbc_crossover_lowopt | FLOAT | 30040–1000 | MBC CROSSOVER LOW • Frequency split point between Bass and Mids. |
| mbc_crossover_highopt | FLOAT | 30001000–16000 | MBC CROSSOVER HIGH • Frequency split point between Mids and Highs. |
| mbc_crossover_orderopt | INT | 82–8 | CROSSOVER SLOPE • Higher numbers create sharper frequency separation. |
| mbc_low_thresh_dbopt | FLOAT | -24.0-60–0 | LOW BAND THRESHOLD • Level at which bass compression engages. |
| mbc_low_ratioopt | FLOAT | 2.51–20 | LOW BAND RATIO • Severity of bass compression. |
| mbc_mid_thresh_dbopt | FLOAT | -22.0-60–0 | MID BAND THRESHOLD • Level at which mid compression engages. |
| mbc_mid_ratioopt | FLOAT | 2.51–20 | MID BAND RATIO • Severity of mid compression. |
| mbc_high_thresh_dbopt | FLOAT | -20.0-60–0 | HIGH BAND THRESHOLD • Level at which treble compression engages. |
| mbc_high_ratioopt | FLOAT | 2.01–20 | HIGH BAND RATIO • Severity of treble compression. |
| do_limiteropt | BOOLEAN | true | ENABLE LIMITER • Engages the final brickwall lookahead limiter to prevent clipping. |
| limiter_threshold_dbopt | FLOAT | -1.0-24–0 | LIMITER CEILING • Maximum allowed True Peak level (-1.0 is standard safety margin). |
| soft_clip_driveopt | FLOAT | 1.000.8–1.5 | SOFT CLIP DRIVE • Pre-limiter saturation gain. Higher = Louder/Dirtier, Lower = Clean. |
| stereo_widthopt | FLOAT | 1.000–2.5 | STEREO WIDTH • 1.0 = Original, >1.0 = Wider (Haas effect), <1.0 = Narrower. |
| fast_modeopt | BOOLEAN | false | FAST MODE • Skips intermediate LUFS normalization passes for a speed boost. |
| skip_initial_analysisopt | BOOLEAN | false | SKIP PRE-ANALYSIS • Skips initial chart generation to save time. |
| mixopt | FLOAT | 1.000–1 | GLOBAL MIX • Final Dry/Wet blend parameter (1.0 = 100% Processed). |
Outputs (8)
| Name | Type | Description |
|---|---|---|
| audio | AUDIO | — |
| analysis_details | STRING | — |
| yaml_config | STRING | — |
| waveform_before | IMAGE | — |
| waveform_after | IMAGE | — |
| spectrum_plot | IMAGE | — |
| dynamics_plot | IMAGE | — |
| lufs_history_plot | IMAGE | — |