gchq/CyberChef · error · Error

Invalid marker while parsing JPEG at pos

Error message

Invalid marker while parsing JPEG at pos ${stream.position}: ${marker}

What it means

Thrown by extractJPEG while walking JPEG segments. Every JPEG marker is two bytes beginning with 0xFF; if the first byte read is not 0xFF the stream is misaligned or the data is not JPEG. This throws a plain Error (not OperationError), so it surfaces as an internal/unexpected failure rather than recipe output.

Solutions

1. Confirm the input actually starts with the JPEG SOI marker (0xFFD8) before calling extractJPEG.
2. Verify the offset points at a real JPEG segment boundary (typically the SOI).
3. If the file is corrupt, re-acquire it or route it to the correct format extractor.

Example fix

// before
extractJPEG(arbitraryBytes, 0);

// after
if (arbitraryBytes[0] === 0xff && arbitraryBytes[1] === 0xd8) {
    extractJPEG(arbitraryBytes, 0);
}

function looksLikeJPEG(bytes, offset = 0) {
  return bytes.length - offset >= 2 && bytes[offset] === 0xff && bytes[offset + 1] === 0xd8;
}
if (!looksLikeJPEG(bytes, offset)) throw new Error("Input is not a JPEG (no SOI marker)");
extractJPEG(bytes, offset);

const isJPEGSOI = (bytes, offset = 0) =>
  bytes.length - offset >= 2 && bytes[offset] === 0xff && bytes[offset + 1] === 0xd8;

try {
  extractJPEG(bytes, offset);
} catch (err) {
  if (/Invalid marker while parsing JPEG/.test(err.message)) {
    // not a JPEG or corrupt; route to a different extractor
  } else throw err;
}

When it happens

Trigger: Calling extractJPEG(bytes, offset) on data that is not a JPEG, with an offset that lands mid-scan, or on a truncated/corrupted file where segment boundaries are wrong. Note the message interpolates `marker` (a Uint8Array), so the value prints as a comma-separated sequence like '0,1'.

Common situations: Wrong file type routed to the JPEG extractor; offset computed incorrectly by a caller; bit-flipped or crafted/malicious file; truncated download where a segment size jumped past real data.

Related errors

• Unable to parse JPEG successfully
• Invalid block type while parsing DEFLATE stream at pos
• Invalid Huffman Code length while parsing DEFLATE block at…
• Caught in probable infinite loop while parsing Huffman Block
• Not a valid RTF file

AI-assisted analysis of gchq/CyberChef@4290ea7539 (2026-08-13). Data as JSON: /api/errors/b58656b0054b3f20. Report an issue: GitHub.

Appendix: source

Thrown at src/core/lib/FileSignatures.mjs:2634

            extractor: null
        }
    ]
};


/**
 * JPEG extractor.
 *
 * @param {Uint8Array} bytes
 * @param {number} offset
 * @returns {Uint8Array}
 */
export function extractJPEG(bytes, offset) {
    const stream = new Stream(bytes.slice(offset));

    while (stream.hasMore()) {
        const marker = stream.getBytes(2);
        if (marker[0] !== 0xff) throw new Error(`Invalid marker while parsing JPEG at pos ${stream.position}: ${marker}`);

        let segmentSize = 0;
        switch (marker[1]) {
            // No length
            case 0xd8: // Start of Image
            case 0x01: // For temporary use in arithmetic coding
                break;
            case 0xd9: // End found
                return stream.carve();

            // Variable size segment
            case 0xc0: // Start of frame (Baseline DCT)
            case 0xc1: // Start of frame (Extended sequential DCT)
            case 0xc2: // Start of frame (Progressive DCT)
            case 0xc3: // Start of frame (Lossless sequential)
            case 0xc4: // Define Huffman Table
            case 0xc5: // Start of frame (Differential sequential DCT)
            case 0xc6: // Start of frame (Differential progressive DCT)

View on GitHub (pinned to 4290ea7539)