Wednesday, December 6, 2006

JPEG: Hidden Secrets of Your Wedding Photos

Wedding photographers drift from one photographic moment to the next, trying to capture the bride and groom at their best on their wedding day. Posting on wedding sites and blogs, sending pictures via cell phones, wedding couples rush to share their newly photographed wedding pictures to friends and relatives. Whether it's email, a CD, or even the internet, wedding pictures tend to travel and occasionally end up in odd places. The most popular wedding photography format is the JPEG format. Brides, grooms and even wedding photographers might not realize their wedding photographs could compromise a wedding couple’s privacy. JPEG has quite a few hidden levels of data and information that most users do not realize exist.


JPEG's History

The development of the JPEG standard started as far back as 1983. In April 1983, the International Organization for Standardization (ISO) started working on a process of incorporating photographic-quality graphics onto text terminals.

In 1986, ISO joined with the International Telegraph and Telephone Consultative Committee (CCITT) to form the Joint Photographic Experts Group (JPEG). The CCITT is now known as ITU-T (International Telecommunication Union - Telecommunication Standardization Sector).

The committee issued the JPEG standard in 1992 and it was approved in 1994 as ISO 10918-1. The JPEG format typically uses a lossy data compression method to transform an image. This compression method allows the image file size to be reduced, but at the expense of image quality. The higher the compression used, the higher the degradation in image quality.


JPEG's Technology (To skip this technical study, go to JPEG's Hidden Data)

In a typical JPEG 24 bit encoding process, the beginning starts with a color space transformation. The RGB values that make up an image are converted into a different color space called YCbCr. The Y channel represents brightness, while the other two channels combine to represent hue and saturation.



The human eye is capable of distinguishing more detail in the Y channel, so the encoding process is designed with this information in mind. The next step involves taking the other two channels and downsampling them, which results in the reduction of their sampling rate. This process reduces the data size. The channels are then individually run through a DCT (discrete cosine transform). This process involves taking each channel and dividing it up into 8x8 pixel tiles and running the DCT to compress the data.

The next process involves taking the compressed channels and quantizing them. Quantization is the bulk of the data compression, reducing the amount of high frequency information present in the channels that the human eye has difficulty in distinguishing. Note, that the artifacts typically present in highly compressed JPEG images that the user sees is a result of the compression done in the quantization process.

Afterwards, the result is entropy coded to compress the data further. Entropy coding is a lossless compression method, which compresses the data, without any data loss. The first pass is compressed with RLE (Run Length Encoding). Lastly, the resulting data is then compressed again with Huffman Encoding.


The IJG (Independent JPEG Group) developed a standard called JFIF (JPEG File Interchange Format). This format was necessary to create files from JPEG data. JFIF was originally designed by Eric Hamilton. This format allowed JPEG data to be shared across platforms. This format boasted many features, some of which include: JPEG compression, PC / Mac / Unix compatibility, YCbCr color space, any many other features. Today’s typical wedding photograph JPEG is a JFIF, or Exif.

The Exif (Exchangeable Image File Format) is almost universally used by wedding photographers using digital cameras. Exif was created by the JEIDA (Japan Electronic Industry Development Association). This specification uses the existing JPEG file format, adding a special feature: metadata tags. This is where a JPEG file can contain many levels of information.

JPEG's Secrets - The Metadata

Metadata is embeddable by both JFIF and Exif, which allow for various forms of encapsulated information in a JPEG file. Other specifications that embed metadata include: IPTC headers (International Press Telecommunications Council and XMP (Extensible Metadata Platform) from Adobe.

Metadata is intended to help the user to further define the characteristics of a photograph or digital image. In certain cases, metadata information may be of a sensitive nature. This is a small sampling of information that may be embedded in a JPEG file:

  • Copyright Information

  • Image manipulation software used: (Photoshop, Paintshop, Corel Draw)

  • Text comments

  • GPS location (Global Positioning Satellite)

  • Serial numbers

The Exif data can be extremely extensive. This is a partial list of data that is typically embedded by a digital camera:

  • Text descriptions

  • Camera make

  • Camera model

  • Date and Time

  • Camera software version

  • Exposure mode

  • ISO mode

Image manipulation programs usually add additional metadata to JPEG images. JPEG users need to be aware of the possible information transmitted through JPEG files. The embedded data can contain privacy information not intended for public use. Metadata, thumbnails and Exif data can combine to create large file sizes. Removing metadata can reduce JPEG file sizes down by as much as fifty percent. Your wedding websites can also benefit from faster user downloads, and reduced file traffic by optimizing your JPEG images.

Whether its privacy concerns or file size issues, be cautious of JPEG’s clandestine data before you publish your wedding pictures! Stay tuned for our upcoming follow-up article on how to prevent JPEG privacy data exposure, and insider tips on how to manipulate JPEG images like a pro.