Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

DRM Decryption Investigation

Related Issue: #73

Overview

This investigation documents Kobo’s DRM (KDRM) system and how it encrypts purchased books. The native Kobo firmware stores all necessary decryption keys in its SQLite database.

The goal is to document how to decrypt books on device so that they can be viewed by an alternative reading software.

Key Findings

Required Data Sources

To decrypt KDRM-protected books, we need data from two sources:

  1. Device Serial - From /mnt/onboard/.kobo/version (first comma-separated field)
  2. User ID - From the user table in KoboReader.sqlite
  3. Content Keys - From the content_keys table in KoboReader.sqlite

Version File

The device serial is stored in the version file:

/mnt/onboard/.kobo/version

Format: SERIAL,version,version,version,version,platform_id

The first field before the comma is the device serial number.

Database Tables

user Table - User Credentials

SELECT UserID FROM user;

Returns:

  • UserID: UUID format (e.g., 12345678-abcd-1234-abcd-123456789abc)

content_keys Table - File Encryption Keys

SELECT elementId, elementKey
FROM content_keys
WHERE volumeId = ?;

Structure:

  • volumeId: Book identifier
  • elementId: File path within EPUB (e.g., OEBPS/image/cover.jpg)
  • elementKey: Base64-encoded encryption key for that specific file

KDRM Encryption System

Characteristics:

  • File-level encryption (not archive-level)
  • Individual files within EPUB/KEPUB are AES-ECB encrypted
  • Uses known hash keys for device ID derivation

Key Hierarchy

The key derivation follows the approach documented in the obok project:

flowchart TD
    A[Hash Key] --> C[Concatenate]
    B[Device Serial] --> C
    C --> D[SHA256 hash]
    D --> E[Device ID<br/>64 hex chars]

    E --> F[Concatenate]
    G[User ID] --> F
    F --> H[SHA256 hash]
    H --> I[Extract second 16 bytes]
    I --> J[User Key]

    J --> K[Decrypt with AES-ECB]
    L[Encrypted Content Key] --> K
    K --> M[Content Key]

    M --> N[Decrypt with AES-ECB]
    O[Encrypted File Content] --> N
    N --> P[PKCS7 unpad]
    P --> Q[Decrypted File]

Known Hash Keys

Kobo uses a set of known hash keys for device ID derivation. The decryption process tries each until one produces valid output:

  • 88b3a2e13
  • XzUhGYdFp
  • NoCanLook
  • QJhwzAtXL

Decryption Process

Step 1: Read Device Serial

# Read from /mnt/onboard/.kobo/version
# Parse first comma-separated field
serial = "N123456789012"

Step 2: Get User ID from Database

# SELECT UserID FROM user
user_id = "12345678-abcd-1234-abcd-123456789abc"

Step 3: Derive User Key

# Try each hash key until decryption succeeds
for hash_key in KOBO_HASH_KEYS:
    device_id = SHA256(hash_key + serial).hexdigest()
    user_key = SHA256(device_id + user_id).hexdigest()[32:]  # Second 16 bytes

Step 4: Decrypt Content Keys

For each encrypted file in the book:

encrypted_content_key = base64_decode(content_keys[element_id])
content_key = AES_ECB_decrypt(encrypted_content_key, user_key)

Step 5: Decrypt File Content

encrypted_content = read_from_epub_zip(element_id)
decrypted_content = AES_ECB_decrypt(encrypted_content, content_key)
final_content = pkcs7_unpad(decrypted_content)

Step 6: Reconstruct EPUB

Create new ZIP with:

  • Decrypted files replacing encrypted ones
  • Unencrypted files copied as-is
  • Original directory structure preserved

References