ComfyUI Audio Toolkit (AudioTools)

Welcome to the ComfyUI Audio Toolkit (AudioTools), a comprehensive custom node pack for ComfyUI that brings a full suite of audio generation, processing, and analysis capabilities into your generative workflows.

This toolkit is designed for a wide range of audio tasks, from podcast enhancement and text-to-speech to creative music manipulation and fully automated, batch-processed audio-reactive visual generation.

Node Glossary

I/O & Batch Nodes

• Load Audio Batch (Path)
• Get Audio From List
• Standardize Audio (Format/Channels)

Generation Nodes

• Text to Speech

Utility Nodes

• Concatenate Audio
• Stereo Panner
• Pad With Silence

Processing & Repair Nodes

• Amplify / Gain
• Normalize Audio
• Mix Audio Tracks
• Trim Audio
• Remove Silence
• Noise Gate
• De-Esser
• De-Plosive (Low Cut)
• De-Hum (50/60Hz)
• Parametric EQ for Voice
• Vocal Compressor

Effects Nodes

• Reverb
• Delay / Echo
• Fade In
• Fade Out
• Pitch Shift / Time Stretch

AI Nodes

• Stem Separator (AI)
• Speech Denoise (AI)
• Speech-to-Text + SRT (Whisper)

Analysis & Reactive Nodes

• Loudness Meter (LUFS)
• BPM Detector / Reactive
• Audio-Reactive Envelope
• Show Audio Info

Visualization Nodes

• Display Waveform
• Compare Waveforms

---

Installation

1. Navigate to your ComfyUI custom_nodes directory:

   cd ComfyUI/custom_nodes/

2. Clone this repository:

   git clone https://github.com/lum3on/ComfyUI_AudioTools

3. Install the required dependencies from within the new directory:

   pip install -r ComfyUI_AudioTools/requirements.txt

4. Restart ComfyUI.

• Batch Processing: Most processing, effects, and visualization nodes are now batch-aware. You can load an entire folder of audio files, process them all simultaneously, and visualize the results in one go.

---

Node Reference

📂 I/O & Batch Nodes

Load Audio Batch (Path)

Category: AudioTools/IO

Loads all audio files from a folder path that match a pattern (e.g., *.wav). Can be configured to sort files in various ways, including by modification date to get the newest file.

Parameter	Type	Description
(Input) directory_path	STRING	The full path to the directory containing audio files.
(Input) file_pattern	STRING	The pattern to match files (e.g., *.wav, *.mp3, audio_*.flac).
(Input) sort_order	COMBO	The order to sort files before loading.
(Input) skip_first	INT	Number of files to skip from the start of the sorted list.
(Input) load_cap	INT	Maximum number of files to load. Use -1 for no limit.
(Output) audio_batch	AUDIO	A single padded AUDIO object containing all loaded clips as a batch.
(Output) audio_list	AUDIO_LIST	A list where each item is a separate, unpadded AUDIO clip.
(Output) filenames	STRING	A list of the filenames (string) that were successfully loaded.

Get Audio From List

Category: AudioTools/IO

Retrieves a single audio clip from the audio_list output of the batch loader, allowing for individual processing within a workflow.

Parameter	Type	Description
(Input) audio_list	AUDIO_LIST	The list of audio clips from a batch loader.
(Input) index	INT	The index of the audio clip to retrieve from the list. Wraps around if the index is out of bounds.
(Output) AUDIO	AUDIO	The single audio clip selected from the list.

Standardize Audio (Format/Channels)

Category: AudioTools/Processing

Converts audio to a standard format (mono or stereo) and data type to fix compatibility issues with other nodes that expect a specific layout.

Parameter	Type	Description
(Input) audio	AUDIO	The audio (or batch) to standardize.
(Input) channel_layout	COMBO	Convert audio to mono (single channel) or ensure it is stereo (two channels).
(Output) AUDIO	AUDIO	The standardized audio.

💬 Generation Nodes

Text to Speech

Category: AudioTools/Generation

Converts text into spoken audio using your operating system's built-in text-to-speech engine.

Parameter	Type	Description
(Input) text	STRING	The text to be converted into speech.
(Input) voice_index	COMBO	The system voice to use for synthesis. Voice options depend on your operating system.
(Input) rate	INT	The speaking rate in words per minute.
(Input) volume	FLOAT	The volume of the generated audio (0.0 to 1.0).
(Output) AUDIO	AUDIO	The generated spoken audio clip.

🛠️ Utility Nodes

Concatenate Audio

Category: AudioTools/Utility

Joins two audio clips together end-to-end. The first clip from each input batch is used.

Parameter	Type	Description
(Input) audio_a	AUDIO	The first audio clip (will be the beginning of the result).
(Input) audio_b	AUDIO	The second audio clip (will be appended to the end of the first).
(Output) AUDIO	AUDIO	The new, combined audio clip.

Stereo Panner

Category: AudioTools/Utility

Positions a sound in the stereo (left/right) field. This is applied to all audio clips in a batch.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to pan.
(Input) pan	FLOAT	Stereo position. -1.0 is hard left, 1.0 is hard right, 0.0 is center.
(Output) AUDIO	AUDIO	The panned audio. Mono inputs are converted to stereo.

Pad With Silence

Category: AudioTools/Utility

Adds a specified duration of silence to the beginning or end of audio. This is applied to all audio clips in a batch.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to pad with silence.
(Input) pad_start_seconds	FLOAT	Duration of silence to add to the beginning of the audio.
(Input) pad_end_seconds	FLOAT	Duration of silence to add to the end of the audio.
(Output) AUDIO	AUDIO	The padded audio.

Processing & Repair Nodes

Amplify / Gain

Category: AudioTools/Processing

Adjusts the volume of the audio by a specified decibel (dB) value.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to apply gain to.
(Input) gain_db	FLOAT	Amount of gain in decibels (dB) to apply. Positive values amplify, negative values attenuate.
(Output) AUDIO	AUDIO	The amplified audio.

Normalize Audio

Category: AudioTools/Processing

Normalizes the peak volume of the audio to a target dB level, maximizing loudness without clipping.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to normalize.
(Input) target_level_db	FLOAT	The target peak volume in decibels (dB). A value of 0.0 is maximum, but -1.0 is a common target to avoid clipping.
(Output) AUDIO	AUDIO	The normalized audio.

Mix Audio Tracks

Category: AudioTools/Processing

Combines two audio tracks into one. The first clip from each input batch is used.

Parameter	Type	Description
(Input) audio_1	AUDIO	The first audio track.
(Input) audio_2	AUDIO	The second audio track.
(Input) gain_1_db	FLOAT	Gain in dB for the first audio track.
(Input) gain_2_db	FLOAT	Gain in dB for the second audio track.
(Output) AUDIO	AUDIO	The mixed audio.

Trim Audio

Category: AudioTools/Processing

Cuts a specified number of seconds from the beginning or end of an audio clip. (Note: This node is not batch-aware).

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to trim.
(Input) trim_start_seconds	FLOAT	Number of seconds to cut from the beginning of the audio.
(Input) trim_end_seconds	FLOAT	Number of seconds to cut from the end of the audio.
(Output) AUDIO	AUDIO	The trimmed audio clip.

Remove Silence

Category: AudioTools/Processing

Intelligently analyzes and trims silent sections from an audio clip. (Note: This node is not batch-aware).

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip from which to remove silent sections.
(Input) silence_threshold_db	FLOAT	The volume level (in dB) below which audio is considered silent.
(Input) min_silence_len_ms	INT	The minimum duration (in milliseconds) of silence to be removed.
(Output) AUDIO	AUDIO	The audio with silent sections removed.

Noise Gate

Category: AudioTools/Processing

Silences audio that falls below a specified volume threshold, useful for removing background noise between words.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to apply the noise gate to.
(Input) threshold_db	FLOAT	The volume level (dB) below which the gate will close and silence the audio.
(Input) attack_ms	FLOAT	How quickly (in ms) the gate opens when the signal exceeds the threshold.
(Input) release_ms	FLOAT	How quickly (in ms) the gate closes after the signal falls below the threshold.
(Output) AUDIO	AUDIO	The gated audio.

De-Esser

Category: AudioTools/Processing

Reduces harsh "s" sounds (sibilance) in voice recordings by applying a narrow-band EQ cut at a specified frequency.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to de-ess.
(Input) frequency_hz	INT	The center frequency of sibilance to target (typically 5-8 kHz).
(Input) reduction_db	FLOAT	The amount of gain reduction (in dB) to apply at the target frequency.
(Input) q_factor	FLOAT	The width of the frequency band to affect. Higher Q is narrower.
(Output) AUDIO	AUDIO	The de-essed audio.

De-Plosive (Low Cut)

Category: AudioTools/Processing

Reduces low-frequency pops ("plosives") caused by air hitting the microphone (e.g., from 'p' and 'b' sounds) using a high-pass filter.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to de-plosive (low-cut).
(Input) cutoff_hz	INT	The cutoff frequency for the high-pass filter. Frequencies below this will be rolled off.
(Output) AUDIO	AUDIO	The filtered audio.

De-Hum (50/60Hz)

Category: AudioTools/Processing

Removes electrical power line hum by applying very narrow notch filters at the fundamental frequency (50 or 60 Hz) and its first harmonic.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to de-hum.
(Input) hum_freq	COMBO	The fundamental frequency of the electrical hum to remove.
(Input) reduction_db	FLOAT	The amount of gain reduction (dB) to apply to the hum frequencies.
(Input) q_factor	FLOAT	The narrowness of the filter. A high Q value is needed to target only the hum.
(Output) AUDIO	AUDIO	The de-hummed audio.

Parametric EQ for Voice

Category: AudioTools/Processing

A 3-band equalizer specifically tuned for enhancing vocal clarity, featuring a low-cut, a presence boost, and an "air" band.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to equalize.
(Input) low_cut_hz	INT	Low-cut (high-pass) filter to remove rumble. 80-120Hz is common for voice.
(Input) presence_boost_db	FLOAT	Boost/cut for vocal presence (around 4kHz).
(Input) air_boost_db	FLOAT	High-shelf boost/cut for 'air' and clarity (around 12kHz).
(Output) AUDIO	AUDIO	The equalized audio.

Vocal Compressor

Category: AudioTools/Processing

Evens out the dynamic range of an audio clip, making quiet parts louder and loud parts quieter for a more consistent volume level.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to compress.
(Input) threshold_db	FLOAT	The volume level (dB) at which the compressor starts working.
(Input) ratio	FLOAT	The amount of gain reduction (e.g., 4.0 means a 4:1 ratio).
(Input) attack_ms	FLOAT	How quickly (in ms) the compressor reacts to loud sounds.
(Input) release_ms	FLOAT	How quickly (in ms) the compressor stops after the sound falls below the threshold.
(Input) makeup_gain_db	FLOAT	Volume boost to apply after compression to make up for the reduced level.
(Output) AUDIO	AUDIO	The compressed audio.

✨ Effects Nodes

Reverb

Category: AudioTools/Effects

Adds spatial reverberation to the audio, simulating the sound of a room or space.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to apply reverb to.
(Input) room_size	FLOAT	The perceived size of the reverberant space (0-100).
(Input) damping	FLOAT	How much the high frequencies are absorbed in the reverb tails (0-100).
(Input) wet_level	FLOAT	The volume of the reverberated (wet) signal.
(Input) dry_level	FLOAT	The volume of the original (dry) signal.
(Output) AUDIO	AUDIO	The reverberated audio.

Delay / Echo

Category: AudioTools/Effects

Creates a repeating, decaying echo effect on the audio.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to apply delay to.
(Input) delay_ms	FLOAT	The time (in milliseconds) between each echo.
(Input) feedback	FLOAT	How much of the delayed signal is fed back into the delay line, creating more echoes.
(Input) mix	FLOAT	The balance between the original (dry) and delayed (wet) signal. 0.0 is all dry, 1.0 is all wet.
(Output) AUDIO	AUDIO	The audio with the delay effect.

Fade In

Category: AudioTools/Effects

Applies a linear fade-in from silence to the start of the audio.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to apply a fade-in to.
(Input) duration_seconds	FLOAT	The duration of the fade-in effect in seconds.
(Output) AUDIO	AUDIO	The audio with the fade-in applied.

Fade Out

Category: AudioTools/Effects

Applies a linear fade-out to silence at the end of the audio.

Parameter	Type	Description
(Input) audio	AUDIO	The audio to apply a fade-out to.
(Input) duration_seconds	FLOAT	The duration of the fade-out effect in seconds.
(Output) AUDIO	AUDIO	The audio with the fade-out applied.

Pitch Shift / Time Stretch

Category: AudioTools/Effects

Changes the audio's pitch without changing its speed, and/or changes its speed without changing the pitch. (Note: This node is not batch-aware).

Parameter	Type	Description
(Input) audio	AUDIO	The audio to pitch shift or time stretch.
(Input) pitch_semitones	FLOAT	The number of semitones to shift the pitch up or down.
(Input) tempo_factor	FLOAT	The factor by which to change the tempo. >1.0 is faster, <1.0 is slower.
(Output) AUDIO	AUDIO	The processed audio.

🧠 AI Nodes

Stem Separator (AI)

Category: AudioTools/AI

Uses the Demucs AI model to split a music track into its core components: vocals, bass, drums, and other instruments.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to be separated into stems.
(Input) model_name	COMBO	The Demucs model to use for separation. 'htdemucs_ft' is a good general-purpose choice.
(Output) vocals	AUDIO	The isolated vocal track.
(Output) bass	AUDIO	The isolated bass track.
(Output) drums	AUDIO	The isolated drum track.
(Output) other	AUDIO	All other musical elements combined.

Speech Denoise (AI)

Category: AudioTools/AI

Uses the Demucs AI model to isolate vocals from a recording, effectively removing background noise and non-vocal sounds.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip containing speech to be denoised.
(Input) model_name	COMBO	The Demucs model to use for isolating vocals. It will remove non-vocal sounds.
(Output) AUDIO	AUDIO	The denoised audio containing only the vocal signal.

Speech-to-Text + SRT (Whisper)

Category: AudioTools/AI

Transcribes audio to text using OpenAI's Whisper model. Can optionally generate a timed SRT (SubRip Subtitle) formatted string.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to be transcribed.
(Input) model_size	COMBO	The size of the Whisper model to use. Larger models are more accurate but slower.
(Input) language	COMBO	The language of the speech in the audio. 'Auto-Detect' is an option.
(Input) task	COMBO	Choose between standard transcription or translating the speech directly to English.
(Input) generate_srt	BOOLEAN	If enabled, generates a timestamped SRT subtitle string in the srt_text output.
(Input) srt_max_line_len	INT	(Optional) The maximum number of characters allowed per line in an SRT block.
(Input) srt_max_lines	INT	(Optional) The maximum number of lines allowed per SRT block.
(Input) srt_max_duration_sec	FLOAT	(Optional) The maximum duration in seconds an SRT block can cover.
(Output) text	STRING	The transcribed text as a single string.
(Output) srt_text	STRING	The generated SRT subtitle string with timestamps. Empty if generate_srt is disabled.

📈 Analysis & Reactive Nodes

Loudness Meter (LUFS)

Category: AudioTools/Analysis

Measures the perceived loudness of an audio clip according to the EBU R 128 broadcast standard and outputs the result as a string.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to measure for loudness.
(Output) loudness_info	STRING	A text string containing the measured integrated loudness in LUFS.

BPM Detector / Reactive

Category: AudioTools/Analysis

Estimates the tempo (Beats Per Minute) of an audio clip and generates a frame-synced list of beat events.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to detect the tempo from.
(Input) fps	INT	The target frames per second to sync the beat events list with.
(Output) bpm_info	STRING	A text string with the estimated BPM.
(Output) beat_events	FLOAT	A list of floats, with 1.0 on frames that land on a beat and 0.0 otherwise.

Audio-Reactive Envelope

Category: AudioTools/Analysis

Analyzes the volume envelope (RMS) of an audio clip and outputs it as a frame-by-frame list of floats, perfect for driving animations.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to analyze for its volume envelope.
(Input) fps	INT	The target frames per second to sync the envelope list with.
(Input) smoothing	FLOAT	Amount of smoothing to apply to the envelope. 0 is no smoothing, 1 is maximum smoothing.
(Output) envelope	FLOAT	A list of floats (normalized 0-1) representing the audio's volume for each frame.

Show Audio Info

Category: AudioTools/Visualization

An output node that displays technical details about an audio clip or batch, such as sample rate, duration, batch size, and more.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip to display information about.
(Output) info	STRING	A text string containing the technical details of the audio.

Visualization Nodes

Display Waveform

Category: AudioTools/Visualization

Generates and displays an image of the audio's waveform. This node is batch-aware and will produce one image for each audio clip in the batch.

Parameter	Type	Description
(Input) audio	AUDIO	The audio clip(s) to visualize.
(Input) width	INT	The width of the output image in pixels.
(Input) height	INT	The height of the output image in pixels.
(Input) line_color	STRING	The color of the waveform line (hex code).
(Input) bg_color	STRING	The background color of the image (hex code).
(Input) show_axis	BOOLEAN	Whether to display the time and amplitude axes.
(Output) IMAGE	IMAGE	An image (or batch of images) of the waveform.

Compare Waveforms

Category: AudioTools/Visualization

Creates an overlay image of two waveforms, making it easy to see the difference before and after processing. This node is batch-aware.

Parameter	Type	Description
(Input) audio_a	AUDIO	The first audio clip (or batch) to compare.
(Input) audio_b	AUDIO	The second audio clip (or batch) to compare.
(Input) width	INT	The width of the output image in pixels.
(Input) height	INT	The height of the output image in pixels.
(Input) color_a	STRING	The color for Audio A's waveform (hex code).
(Input) color_b	STRING	The color for Audio B's waveform (hex code).
(Input) bg_color	STRING	The background color of the image (hex code).
(Input) show_axis	BOOLEAN	Whether to display the time and amplitude axes and a legend.
(Output) IMAGE	IMAGE	An image (or batch of images) comparing the two waveforms.

Dependencies

• torch & torchaudio
• demucs
• openai-whisper
• pyttsx3
• matplotlib
• librosa (for advanced analysis)
• pyloudnorm (for LUFS measurement)

---

Changelog

Recent Changes:

Version 1.1.06

• fixed compatibility of the Speech-to-text Node with https://github.com/niknah/ComfyUI-F5-TTS
• added "Standardize Audio Format/Channels" Node to fix future compatibility issues

Version 1.1.05

• added srt output to the Speech-to-Text Node
• fixed some edge case errors in the Speech-to-text Node

Version 1.1.01

• added tooltips on hover to almost all in/outputs
• added licence + disclaimer

Version 1.1.00

• init