Skip to main content

Decode Watermark

Detect and extract watermark information from audio/video files to verify content authenticity and origID.

Endpoint​

POST /api/watermark/decode

Authentication​

Required: Bearer token (API key)

Authorization: Bearer origid_xxxxxxxxxxxxxxxx

Request​

Headers​

HeaderValueRequired
AuthorizationBearer {api_key}Yes
Content-Typemultipart/form-dataYes

Body Parameters​

Send as multipart/form-data:

ParameterTypeDescriptionRequired
fileFileWatermarked audio or video file to analyzeYes

Supported File Types​

  • Audio: .wav, .mp3, .m4a, .flac
  • Video: .mp4, .mov, .avi, .mkv

File Size Limits​

  • Maximum file size: 500MB

Response​

Success Response - Watermark Detected (200 OK)​

{
  "hasWatermark": true,
  "origId": "606885",
  "username": "Revolver"
}
FieldTypeDescription
hasWatermarkbooleanWhether a watermark was detected
origIdstringThe OrigID embedded in the watermark
usernamestringUsername associated with the OrigID

Success Response - No Watermark (200 OK)​

{
  "hasWatermark": false,
  "message": "No watermark detected in this file"
}

Error Responses​

400 Bad Request​

{
  "error": "File is required"
}

401 Unauthorized​

{
  "error": "Authentication required"
}

500 Internal Server Error​

{
  "error": "Decoder service not configured"
}

Usage Examples​

cURL​

curl -X POST https://origid.ai/api/watermark/decode \
  -H "Authorization: Bearer origid_xxxxxxxxxxxxxxxx" \
  -F "file=@/path/to/watermarked_audio.wav"

JavaScript/TypeScript​

const file = document.querySelector('input[type="file"]').files[0];
const formData = new FormData();
formData.append('file', file);

const response = await fetch('https://origid.ai/api/watermark/decode', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer origid_xxxxxxxxxxxxxxxx',
  },
  body: formData,
});

const data = await response.json();

if (data.hasWatermark) {
  console.log('OrigID:', data.origId);
  console.log('Owner:', data.username);
} else {
  console.log('No watermark detected');
}

Python​

import requests

headers = {
    'Authorization': 'Bearer origid_xxxxxxxxxxxxxxxx',
}

files = {
    'file': open('/path/to/watermarked_audio.wav', 'rb'),
}

response = requests.post(
    'https://origid.ai/api/watermark/decode',
    headers=headers,
    files=files
)

result = response.json()

if result['hasWatermark']:
    print(f"OrigID: {result['origId']}")
    print(f"Owner: {result['username']}")
else:
    print("No watermark detected")

Processing Time​

Decoding is synchronous and typically completes in:

  • Audio files: 1-3 seconds
  • Video files: 3-10 seconds

Detection Accuracy​

  • Watermarked files: 99.9% detection rate
  • Non-watermarked files: Accurately identified as unwatermarked
  • Tampered files: May detect watermark with reduced confidence

Rate Limiting​

  • Default limit: 100 requests per hour per API key
  • Decoding is FREE - does not consume credits

Use Cases​

Content Verification​

Verify that content originated from a specific creator:

async function verifyContent(file, expectedOrigId) {
  const result = await decodeWatermark(file);
  
  if (!result.hasWatermark) {
    return { verified: false, reason: 'No watermark found' };
  }
  
  if (result.origId !== expectedOrigId) {
    return { verified: false, reason: 'OrigID mismatch' };
  }
  
  return { verified: true, owner: result.username };
}

Content Provenance​

Track content distribution and detect unauthorized use:

async function trackContentProvenance(file) {
  const result = await decodeWatermark(file);
  
  if (result.hasWatermark) {
    // Log provenance information
    logProvenance({
      origId: result.origId,
      owner: result.username,
      timestamp: new Date(),
      source: 'user_upload'
    });
  }
  
  return result;
}

Notes​

Instant Results

Unlike encoding, decoding is synchronous and returns results immediately without polling.

Free Decoding

Decoding watermarks is free and does not consume account credits. Only encoding (watermarking) consumes credits.

File Handling

The uploaded file is temporarily stored for decoding and deleted immediately after processing. We never retain your decoded files.

Next Steps​