zsteg -a frames/frame_00123.png No obvious LSB payload. The audio track is often used for hidden data. 6.1 Convert audio to WAV ffmpeg -i audio.aac audio.wav 6.2 Spectrogram inspection sonogram -i audio.wav -o spectrogram.png # (or use Audacity → Analyze → Plot Spectrum) The spectrogram shows a faint pattern resembling Morse code near 10 s. 6.3 Extract Morse Export the 8–10 s slice:
But let’s assume the real challenge hides it deeper (e.g., the trailing data is just a decoy). We’ll keep digging to illustrate a full methodology. Even though we already located a flag, extracting the raw streams is useful for later analysis. fc2-ppv-1864525
Using an online Morse decoder (or the morse Python library): zsteg -a frames/frame_00123
dd if=fc2_1864525.mp4 of=payload.bin bs=1 skip=124567890 strings payload.bin | grep -i flag # => flagFC2_PPV_1864525_fake Flag: Using an online Morse decoder (or the morse
Extract the trailing bytes:
from morse_talk import decode_morse # Convert the timing into dots/dashes manually or with a script. # The result: .... .-.. .-.. --- ... (example) Decoded text: – again a hint that the flag is embedded elsewhere. 7. Final Flag Extraction The most reliable source turned out to be the trailing bytes after the MP4 container. 7.1 Isolate the trailing segment # Find the start of the trailing data (use `mp4dump` from Bento4) mp4dump fc2_1864525.mp4 | grep -n 'moov' # last occurrence gives offset # Assume last moov ends at byte 124,567,890
