Don't Be Afraid of the Dark

Platform: hackerlab | Category: Steganography | Type: Challenge | Difficulty: Easy | OS: NA | Author: XESXOR | Date: 2025-12-22 | Status: Solved Techniques: Alpha channel value analysis, Binary mask generation from alpha differences, Invisible data in transparency layer

Summary

Task: PNG image that appears completely black. Solution: Analyzed alpha channel variations (254 vs 255) and created binary mask to reveal hidden flag text.

Don't Be Afraid of the Dark — hackerlab.pro

Description

"Don't be afraid of the dark, it's the place where you can hide" - August Verona

The task contains task.zip with an image task.png (1240x720 PNG with RGBA).

Analysis

Step 1: Initial Inspection

The image appears completely black. All pixels have RGB = (0, 0, 0).

 python read-only • copy allowed Copy
from PIL import Image
import numpy as np

img = Image.open("task.png")
arr = np.array(img)

print(f"Mode: {img.mode}")  # RGBA
print(f"Size: {img.size}")  # (1240, 720)
print(f"RGB unique values: {np.unique(arr[:,:,:3])}")  # [0] - all black

‍​‌‌​​​​‌​‌‌​​‌‌​​‌‌​​‌‌​​‌‌​​‌​​​​‌‌​​​​​‌‌​​‌‌​​‌‌​​‌​‌​‌‌​​‌​‌‍

Step 2: Hint in the Description

The phrase "Don't be afraid of the dark" hints that data is hidden in the "darkness". Since RGB channels are empty, we check the alpha channel.

Step 3: Alpha Channel Analysis

 python read-only • copy allowed Copy
alpha = arr[:, :, 3]
unique_alpha = np.unique(alpha)
print(f"Unique alpha values: {unique_alpha}")  # [254, 255]

# Pixel count
print(f"Alpha=254: {np.sum(alpha == 254)}")  # 33881 pixels
print(f"Alpha=255: {np.sum(alpha == 255)}")  # the rest

Findings:

• 2 unique alpha values: 254 and 255
• 33881 pixels with alpha=254 (slightly transparent)
• Remaining pixels with alpha=255 (fully opaque) ‍​‌‌​​​​‌​‌‌​​‌‌​​‌‌​​‌‌​​‌‌​​‌​​​​‌‌​​​​​‌‌​​‌‌​​‌‌​​‌​‌​‌‌​​‌​‌‍

The difference between 254 and 255 is invisible to the human eye but encodes hidden information.

Solution

Create a binary mask: pixels with alpha=254 become white, the rest become black.

 python read-only • copy allowed Copy
#!/usr/bin/env python3
from PIL import Image
import numpy as np

# Load the image
img = Image.open("task.png")
arr = np.array(img)

# Extract alpha channel
alpha = arr[:, :, 3]

# Create binary mask
# alpha=254 -> white (255) - hidden content
# alpha=255 -> black (0) - background
mask = np.where(alpha == 254, 255, 0).astype(np.uint8)
‍​‌‌​​​​‌​‌‌​​‌‌​​‌‌​​‌‌​​‌‌​​‌​​​​‌‌​​​​​‌‌​​‌‌​​‌‌​​‌​‌​‌‌​​‌​‌‍

# Save the result
Image.fromarray(mask).save("hidden_alpha.png")
print("Saved: hidden_alpha.png")

Result

The resulting image shows the flag text in white on black background:

CODEBY{REDACTED}

Theory

Alpha Channel Steganography

The alpha channel defines pixel transparency:

• 255 = fully opaque
• 0 = fully transparent
• 1-254 = partially transparent

The difference between adjacent values (e.g., 254 and 255) is practically invisible visually but can encode binary data:

• alpha=255 → bit 0
• alpha=254 → bit 1 ‍​‌‌​​​​‌​‌‌​​‌‌​​‌‌​​‌‌​​‌‌​​‌​​​​‌‌​​​​​‌‌​​‌‌​​‌‌​​‌​‌​‌‌​​‌​‌‍

In this case, a simple approach was used: pixels with alpha=254 form visible text when visualized.

Why This Works

1. The human eye cannot distinguish 0.4% difference in transparency
2. Most image viewers ignore the alpha channel when displaying
3. Black RGB (0,0,0) makes any transparency variations invisible ‍​‌‌​​​​‌​‌‌​​‌‌​​‌‌​​‌‌​​‌‌​​‌​​​​‌‌​​​​​‌‌​​‌‌​​‌‌​​‌​‌​‌‌​​‌​‌‍