Spectral

Start with a sound

Drop a file anywhere, or pick a way in — Open a file you have, Record straight in, or Sketch by humming an idea.

WAV, MP3, FLAC, OGG, AAC, AIFF, M4A · MP4, MOV, WEBM and more — video keeps its picture while you edit the sound · 5.1 and 7.1 keep every channel

Your library — on this device

keep video

Hum an idea

Tempo BPM Beats/bar Count-in bars Click

Spectral by NodeDSP Ready

Upgrade

Analyzing audio...

M—

S—

I—

LRA—

TP—

Channels Tools → all channels

Melody Tighten 0% Tune 100%

Stem Mixer

Time Trip Spray Motion Shimmer

Grid — BPM Loop

Spectral Erase 0 regions Cut -60 dB

Modulation Automation no lanes

0:00 / 0:00

×1.00

fmt

Master Section

AI Assistant

Assistant

Drop an audio file to get started. I'll analyze it and help you process it with natural language commands.

Spectral Audio Processing — by NodeDSP

Email Password

Enter your email and we'll send a link to reset your password.

If that email is registered, a reset link is on its way. Check your inbox.

Choose a new password.

New password

Display Name Email Password

NodeDSP License Key

Users

Tools

History — what was done to this audio

History

Settings

Plan: free

Theme

LLM Provider

Anthropic API Key Model

OpenAI API Key Model

Ollama URL Model

API Access (MCP)

Drive Spectral from any MCP agent. Endpoint: POST /mcp, header X-API-Key.

Copy now — shown only once:

Spectral Plans

The editor, engine, loops, and metering are free forever. Subscriptions unlock the AI assistant, stem separation, higher quality, and more capacity.

Slice to Sampler

Cut the file into slices, audition them on pads, and export a kit — per-slice WAVs plus an SFZ instrument mapped chromatically from C1.

Stem Separation

Isolate individual instruments from the full mix. Today this runs on NodeDSP cloud (a dedicated worker wakes up for your job); on-device separation is coming and will start instantly.

Model

Separating...

Export Audio

Format Sample Rate

Bit Depth

MP3 Bitrate

Export selection only No selection active

Normalize (EBU R128)

Ringtone

Uses your selection (max 30s) with fades & loudness for phone speakers; no selection = first 30s.