audio-processing

Effortlessly Separate Audio Tracks with Demixing Actions

25 Apr 2025
Effortlessly Separate Audio Tracks with Demixing Actions

Demixing is a powerful service designed to enhance audio processing by enabling developers to separate various audio components from a mixed track. Whether you are a music producer looking to isolate instruments for remixing or an audio engineer needing specific elements for analysis, Demixing offers a seamless solution. By employing advanced technology like the Hybrid Transformer architecture, this service provides high-quality separation of vocals, drums, bass, guitar, and piano, making it a valuable tool for anyone working with audio.

Imagine being able to extract just the vocals from a song for a karaoke app or isolating the guitar track for a music lesson. Demixing not only simplifies these tasks but also saves time, allowing developers to focus on creating innovative audio applications.

Prerequisites

To get started with Demixing, you'll need a Cognitive Actions API key and a basic understanding of making API calls.

Separate Audio Stems

The Separate Audio Stems action is at the core of the Demixing service. It enables users to utilize the Demucs model to break down mixed audio tracks into distinct components, such as instruments and vocals. This action addresses the challenge of extracting specific audio elements from a track, making it easier to manipulate or analyze them individually.

Input Requirements

To use this action, you need to provide the following input:

  • audio: A URI link to the input audio file you wish to process.
  • stem: Specifies the audio source to separate (e.g., vocals, bass, drums, guitar, piano, or all). The default is set to "drums".
  • outputFormat: Determines the format of the output audio file, with options including 'mp3', 'wav', and 'flac'. The default is 'mp3'.

Example Input:

{
  "stem": "drums",
  "audio": "https://replicate.delivery/pbxt/JyyTIJ1fAXvSeIHsSUB1UI07ipZvbjx9MOSgYS2MRjmRQIvR/JBlanked%20-%20Cobie%20Sample.mp3",
  "outputFormat": "mp3"
}

Expected Output

The expected output is a downloadable link to a zip file containing the separated audio tracks, allowing for easy access and further processing.

Example Output:

https://assets.cognitiveactions.com/invocations/8d798b58-74e7-473c-b0bc-99fd84eafe74/9980b64d-cc12-4731-829b-33ac9de969bd.zip

Use Cases for this Specific Action

  • Music Production: Producers can isolate specific instruments to create remixes or mashups without the need for the original multi-track recordings.
  • Educational Purposes: Music teachers can extract specific parts of a song to help students learn individual instruments.
  • Audio Analysis: Researchers can analyze the frequency content or dynamics of individual audio components for various studies.
  • Karaoke Applications: Developers can create apps that provide karaoke tracks by removing vocals from popular songs.
import requests
import json

# Replace with your actual Cognitive Actions API key and endpoint
# Ensure your environment securely handles the API key
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
# This endpoint URL is hypothetical and should be documented for users
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute"

action_id = "127d68ee-fb24-487f-8d37-d50963747a6f" # Action ID for: Separate Audio Stems

# Construct the exact input payload based on the action's requirements
# This example uses the predefined example_input for this action:
payload = {
  "stem": "drums",
  "audio": "https://replicate.delivery/pbxt/JyyTIJ1fAXvSeIHsSUB1UI07ipZvbjx9MOSgYS2MRjmRQIvR/JBlanked%20-%20Cobie%20Sample.mp3",
  "outputFormat": "mp3"
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json",
    # Add any other required headers for the Cognitive Actions API
}

# Prepare the request body for the hypothetical execution endpoint
request_body = {
    "action_id": action_id,
    "inputs": payload
}

print(f"--- Calling Cognitive Action: {action.name or action_id} ---")
print(f"Endpoint: {COGNITIVE_ACTIONS_EXECUTE_URL}")
print(f"Action ID: {action_id}")
print("Payload being sent:")
print(json.dumps(request_body, indent=2))
print("------------------------------------------------")

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json=request_body
    )
    response.raise_for_status() # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully. Result:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body (non-JSON): {e.response.text}")
    print("------------------------------------------------")

Conclusion

The Demixing service, with its Separate Audio Stems action, significantly enhances audio processing capabilities for developers. By allowing for precise separation of audio components, it opens up a wide range of possibilities in music production, education, and analysis. As you integrate these actions into your applications, consider the various use cases that can benefit from audio separation. Start leveraging the power of Demixing today to elevate your audio projects to new heights!

← Back to All Articles