Skip to content

TIF.md

Latest commit

 

History

History
282 lines (216 loc) · 7.46 KB

TIF.md

File metadata and controls

282 lines (216 loc) · 7.46 KB
Raw

TIFF (.TIF/.TIFF) Format

Overview

TIFF (Tagged Image File Format) is a flexible, adaptable file format for handling images and data within a single file. TIFF was originally created by Aldus Corporation in 1986 and is now controlled by Adobe Systems. It remains one of the most widely supported and versatile image formats.

Key Characteristics

  • Dimensions: Variable (no limits in specification)
  • Color Depth: 1 to 64 bits per channel
  • Compression: Multiple options (uncompressed, LZW, PackBits, JPEG, ZIP, etc.)
  • Magic Bytes:
    • Little-endian: 0x49 0x49 0x2A 0x00 ("II*\0")
    • Big-endian: 0x4D 0x4D 0x00 0x2A ("MM\0*")
  • File Extension: .TIF or .TIFF
  • Platform: Cross-platform (Windows, Mac, Linux, Unix)
  • Year: 1986 (ongoing revisions)
  • Developer: Originally Aldus Corporation, now Adobe Systems

Historical Context

TIFF was designed to become the standard for storing scanned images and has grown to support a wide variety of image types including:

  • Bilevel (black and white)
  • Grayscale
  • Palette color
  • Full color (RGB)
  • CMYK (for printing)
  • Lab color
  • YCbCr

The format's flexibility made it popular for:

  • Desktop publishing
  • Document imaging and archiving
  • Medical imaging (DICOM uses TIFF)
  • Geospatial data (GeoTIFF)
  • Professional photography

File Structure

Basic Structure

TIFF files are organized using a tag-based structure:

Offset    Size    Description
------    ----    -----------
0x00      2       Byte order (II = little-endian, MM = big-endian)
0x02      2       Version (always 42, 0x002A)
0x04      4       Offset to first IFD (Image File Directory)
var       var     Image data and IFDs

Image File Directory (IFD)

An IFD contains information about the image:

  • Image dimensions
  • Color space
  • Compression type
  • Bits per sample
  • Photometric interpretation
  • Strip/tile organization
  • And many other optional tags

Common TIFF Tags

  • ImageWidth (256): Width of the image in pixels
  • ImageLength (257): Height of the image in pixels
  • BitsPerSample (258): Number of bits per channel
  • Compression (259): Compression scheme used
    • 1 = No compression
    • 2 = CCITT Group 3
    • 5 = LZW
    • 32773 = PackBits
    • 7 = JPEG
    • 8 = Deflate (ZIP)
  • PhotometricInterpretation (262): Color space
    • 0 = WhiteIsZero
    • 1 = BlackIsZero
    • 2 = RGB
    • 3 = Palette color
    • 4 = Transparency mask
    • 5 = CMYK
  • StripOffsets (273): Byte offset of each strip
  • SamplesPerPixel (277): Number of channels
  • RowsPerStrip (278): Number of rows per strip
  • StripByteCounts (279): Byte count for each strip

Color Modes

TIFF supports numerous color modes:

Bilevel (1-bit)

  • Black and white images
  • Each pixel is 1 bit
  • Common in document scanning

Grayscale (8-bit or 16-bit)

  • Shades of gray
  • 8-bit: 256 levels
  • 16-bit: 65,536 levels (for high dynamic range)

Palette Color

  • Indexed color using a color palette
  • Up to 256 colors typically

RGB (24-bit or 48-bit)

  • Full color images
  • 24-bit: 8 bits per channel (R, G, B)
  • 48-bit: 16 bits per channel (for high color depth)

CMYK (32-bit or 64-bit)

  • For printing and prepress
  • Cyan, Magenta, Yellow, Black channels

RGBA (32-bit or 64-bit)

  • RGB with alpha transparency channel

Compression Methods

TIFF supports multiple compression algorithms:

No Compression (Type 1)

  • Uncompressed pixel data
  • Large file sizes but fast to decode
  • Lossless

PackBits (Type 32773)

  • Simple run-length encoding
  • Same algorithm used in MacPaint
  • Lossless
  • See MAC.md for PackBits details

LZW (Type 5)

  • Lempel-Ziv-Welch compression
  • Lossless
  • Good compression for most images
  • Patent issues resolved (expired)

JPEG (Type 7)

  • Lossy compression
  • Used for photographic images
  • Can achieve high compression ratios
  • Quality/size tradeoff

Deflate/ZIP (Type 8)

  • Lossless compression
  • Similar to PNG compression
  • Good compression for most images

CCITT Group 3/4

  • Fax compression
  • For bilevel (1-bit) images
  • Lossless

Decoding with PIL/Pillow

In the Vintage Image Viewer, TIFF files are decoded using PIL's native TIFF support:

def decode(file_path):
    # Use PIL's native TIFF support
    img = Image.open(file_path)

    # Convert to RGB if needed for consistent display
    if img.mode in ('RGBA', 'P', 'LA'):
        img = img.convert('RGB')
    elif img.mode not in ('RGB', 'L', '1'):
        # Convert any other modes to RGB
        img = img.convert('RGB')

    return img

Supported Modes

PIL/Pillow automatically handles:

  • All standard TIFF compressions
  • Multiple color modes
  • Big-endian and little-endian byte orders
  • Multi-page TIFF files
  • Various bit depths

Implementation Reference

See vintage_image_viewer.py:

  • TIFImageDecoder.decode() - TIFF decoder using PIL

Advantages

  • Universal support: Widely supported across platforms and applications
  • Flexible: Supports many color modes and bit depths
  • High quality: Lossless compression options preserve image quality
  • Metadata: Extensive tag system for storing image information
  • Multi-page: Can store multiple images in one file
  • Professional: Industry standard for publishing, printing, archiving

Limitations

  • File size: Uncompressed or losslessly compressed files can be large
  • Complexity: Tag-based structure more complex than simpler formats
  • Compatibility: Some TIFF features not supported by all readers
  • No animation: Unlike GIF, does not support animation

Common Use Cases

  • Document scanning: Standard format for scanned documents
  • Professional photography: RAW conversions and master files
  • Printing and prepress: CMYK color space support
  • Archival: Lossless storage of important images
  • Medical imaging: DICOM standard uses TIFF
  • Geospatial: GeoTIFF for geographic data

Relationship to Vintage Formats

TIFF has historical connections to some vintage formats supported by this viewer:

PackBits Compression

  • TIFF compression type 32773 uses PackBits
  • Same algorithm as MacPaint (.MAC format)
  • See MAC.md for details

LZW Compression

  • More advanced than RLE compression in .ART, .PIC, .PCX
  • Replaced simple RLE in professional applications

File Identification

Quick identification of TIFF files:

# Check magic bytes (little-endian)
xxd -l 4 file.tif
# Should show: 4949 2a00  (II*\0)

# Or big-endian:
# Should show: 4d4d 002a  (MM\0*)

# Python detection
with open('file.tif', 'rb') as f:
    magic = f.read(4)
    is_tiff = (magic == b'II*\x00' or magic == b'MM\x00*')

Extensions and Variants

BigTIFF

  • Extended TIFF for files > 4GB
  • Uses 64-bit offsets instead of 32-bit
  • Supported by modern software

GeoTIFF

  • TIFF with georeferencing information
  • Used for satellite imagery and maps
  • Stores coordinate system and projection data

TIFF/EP

  • TIFF for Electronic Photography
  • Used in digital cameras
  • Supports camera-specific metadata

DNG (Digital Negative)

  • Adobe's open RAW image format
  • Based on TIFF/EP
  • Stores camera RAW data

Specifications

  • Official Specification: TIFF Revision 6.0 (1992)
  • Current Maintainer: Adobe Systems
  • Standards: ISO 12639:2004 (Electronic imaging)

Related Formats

  • PNG: Simpler lossless format with alpha transparency
  • JPEG: Lossy compression for photographs
  • BMP: Windows bitmap (simpler, uncompressed)
  • PCX: Earlier DOS format (.PCX)
  • GIF: Indexed color with animation

Last updated: 2025-11-04