Digitization of Multimedia Assets 

Assets are all kinds of digital files that can become parts of multimedia applications or projects in the course of Media Integration.

In order to become assets, all analog sources must first be digitized.
Two technical parameters play an important role in this context:

• sampling rate
• bit depth

At digitization always a quantization is carried out which leads to falsifications of the original source in form of noise. The amount of falsification (noise) can be regarded as a quality criterion.
-The less the quantization, the higher the quality
- But high quality mostly also leads to high amounts of data

In each individual cast an adequate compromise between quality and storage space required or data rate must be found, according to application.

2. Technical performance of the digitization system
The technical performance of the digitization system is primarily responsible for the maximum resolution to be obtained at digitization. The technical performance is generally defined by the available quantity of levels or potentiality for the (maximally detail-accurate) description of the original analog source information in the form of digital information after the data acquisition.
Resolution can be expressed or described in different ways:
• sampling rate (in kHz or MHz)
• bit depth (quantity of quantization levels in Bit)
• highest spatial resolution (mostly described in elements per unit of length,for example dpi, lpi)
• highest spatial frequency (for example number of bright/dark changes persecond in Hertz)
• maximum data rate (quantity of Bits or Bytes per second, e.g. MBit/sec)
• bandwidth (analog: difference between lowest and highest frequency; digital: maximum data rate needed for transmission in Hertz or Bit/sec)

These descriptions of resolution are mostly being used parallel to each other, but sometimes also alternatively for each other. The use also depends on the kind of analog source material to be acquired (text, drawings, photos, audio, video, etc...).
It does not make sense to always utilize the maximum resolution of the digitization system for data acquisition. This would lead in many cases to a much too high amount of data captured, in comparison with the requirements of the application in
question. It rather makes sense to define the necessary resolution before the data acquisition,
in order to acquire only the amount of data required for the application in question and not more.

Reduction or enhancement of the resolution after data acquisition is possible, butis mostly carried out by an interpolation, that causes an additional loss of qualityby introducing additional noise and/or unsharpness .

Compression

Since the amounts of data gained during the digitization of analog information or signals can get very high, compression is frequently used in the field of digital media (especially for distribution purposes).

Compression here means the reduction of data amounts, without the (complete) loss of the original information (contained in the uncompressed data).
In media production, after digitization, data are compressed first, and then stored in compressed form to save storage capacity. For further editing the data must be decompressed first again. After editing the data can be compressed again and
stored again. Before the re-conversion into analog signals or information the compression must always be removed first, of course.
In connection with compression there are two different basic methods:

Loss-Free Compression

This method of compression has the objective not to change the information content of the data in any way by compressing.
During the decompression the original state of the data before the compression is being restored again to 100 %.
This is accomplished by only removing redundant information from the data and in this way reducing the amount of data to the pure relevant information without any redundancy (redundancy reduction).

A typical field of application for loss-free compression is the compression of software.
Conventional formats are for example:
• arj
• zip
• cap
• sit

Loss-free compression typically reaches compression ratios of approximately 3:1average. This is mostly too low for the distribution of media contents.
In the field of digital media production loss-free compression is used for media content when very high quality is needed (mostly during production, seldom for distribution). In this case compression is often omitted completely, in spite of the
sometimes very high data amounts (for example with uncompressed digital video).

Lossy Compression 

At this method of compression a certain loss of information is being tolerated onpurpose in order to achieve higher compression ratios. At a final decompression a perfect restoration of the original state of the uncompressed data therefore
is not possible.

A decrease of quality of the original is therefore unavoidable in the case of lossy compression.
Lossy compression is carried out by not only removing redundant information but also relevant information from the data. The high art of lossy compression is, to remove only the very part of the relevant data, that is relatively irrelevant for the human perception (irrelevancy reduction).

The weaknesses and limitations of the human perception system are being exploited on purpose in order to achieve higher compression ratios. If we can not perceive sections of the information at all, it does not annoy us much, if these areremoved from the uncompressed data during the lossy compression.

Lossy compression typically can reach compression ratios of approximately 150:1 average and more. Below a certain ratio however, the quality diminishes rapidly. A typical field of application for lossy compression is the compression of media content (especially for distribution). Conventional formats are for example:
   • GIF                                                  • ASF (Windows Media)
   • JPEG                                               • Real-Audio
   • MPEG 1, 2, 4                                    • Real-Video
   • Mp3                                                  • ShockWave and Flash (.swf)

In the field of digital media production lossy compression is often used for media
content when reasonable quality is needed, but the amounts of data are to be
reduced drastically.

from : Prof. Dr. Roland Riempp