BBM

• You can look at the Spec of the Data Here.
• Better Binary Model (BBM) is an Insanely lightweight 3D Model and Geometry format (upto 1.3% of the original file-size). With support for Lossy, and Lossless Compression, and Compression algorithms.

-For Game Engines, Embedded Systems and Microcontrollers with limited resources and/or no storage.

• It is highly versatile, supporting OBJ/PLY/STL conversion, encryption, compression, and multiple models.
• Designed for efficient geometry storage in game engines and real-time applications.
• No encryption information is stored in the Header (keeps header minimal).

---

Features:

• Easy-To-Understand Header.
• Standard Compression Algorithms or None at all.
• Auto Compression Mode (mode 6, selects smallest result automatically).
• Multi-Model Support (multiple models per file).
• Insanely small file-sizes (up to 98%+ reduction).
• Lossless Mode (exact geometry, no precision loss).
• Quantized Mode (ultra-small with near-identical visual quality).
• Bit-Packed Geometry (exact-bit index + coordinate packing).
• Separate Vertex Streams (X/Y/Z for better compression).
• Dump Information on Models at Generation (acts like a key).
• Ability to encrypt your Models (not headers).

---

Geometry Modes:

• lossless
  • Exact vertex data
  • Stream-separated (X/Y/Z)
  • Optimized for compression
• quantized [bits]
  • Fixed-bit coordinate storage (e.g. 12-bit, 14-bit)
  • Much smaller, slight precision tradeoff

---

Compression Modes:

0 = None
1 = BZ2
2 = LZ4 (optional)
3 = ZStandard (optional)
4 = ZLib
5 = LZMA
6 = Auto (selects smallest result)

---

Compiling Models:

│------------------Required Field-------------------│ │------------------------------Optional Field------------------------------│
python generator.py [input] [output] [compression] [dumpKeys] [encryptionMode] [encryptionKey] [geometryMode] [quantBits]

---

Example Command(s):

# Lossless (recommended)
python generator.py model.obj model.bbm 6 false None None lossless

# Quantized (smaller)
python generator.py model.obj model_q.bbm 6 false None None quantized 12

# Encrypted
python generator.py model.obj secure.bbm 6 true aes "MyKey123" lossless

# Multi-model folder
python generator.py models/ army.bbm 6 true xor "TankArmyKey" lossless

---

Header Information:

00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F | Decoded Text 
------------------------------------------------|-----------------
42 42 4D 01 ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? | BBM.............
?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? | ................ 
6D 6F 64 65 6C 00 00 00 00 00 00 00 00 00 00 00 | model...........

Bytes 00->02 = Format Tag ("BBM")
Byte  03     = Format Version / Source Type
Bytes 04->07 = Vertex Count
Bytes 08->0B = Face Count
Bytes 0C->0D = Compression Mode
Bytes 0E->0F = Number of Models
Bytes 10->17 = Vertex Data Length
Bytes 18->1F = Face Data Length
Bytes 20->2F = Model Name

Notes:

• Vertex/Face data may be:
  • Delta encoded
  • Bit-packed
  • Quantized (if enabled)
• Header is NOT compressed or encrypted

---

Rendering Model:

│---------------Required Field----------------│ │--------Optional Field--------│
python renderer.py [inputBbmFile] [modelNumber] [encryptionMode] [encryptionKey]

---

Important Notes:

• Mode 6 will automatically choose the smallest compression algorithm.
• Lossless mode may compress better than quantized due to higher pattern repetition.
• Quantized mode produces the smallest raw data, but may compress less efficiently.
• LZ4 and ZStandard are optional dependencies.

---

Roadmap:

• Support for BBModel / JSON / BJSON
• Normal / UV compression
• Animation / skeletal support
• Advanced topology encoding (triangle strips / edge compression)