Table of Contents
Advertisement
Hard Disk Speed—Throughput
Compression
Rev 1.00
The HRDV16/8 has inputs for both composite video and S-Video. You may only hook up
one camera to each input, either composite or S-Video. You can hook up different types of
cameras to different inputs. When you play back data, a composite output or an S-Video
output may be used to send a signal to your composite or S-Video monitor.
The speed of a hard disk is often measured by its seek time, which tells you how quickly
the drive head can move from one section of its storage to another. For video recording and
playback, a much more important speed measurement is the throughput of the hard disk.
The fastest hard disks are currently capable of speeds as high as 13 MBs at the outer edge
of the disk, while typical hard drive speeds vary from 3 to 7 MBs. The throughput is higher
at the outer edge of the disk because the speed of a point on an outer track is faster than
the speed of a point on an inner track for a disk with a fixed rotation rate.
Is this fast enough for NTSC or PAL standard quality video? For digital video using the world
standard CCIR-601 sampling rate, every frame of video is sampled at a resolution of 720
horizontal by 480 vertical pixels. Each pixel requires two bytes of data—one for the
luminance and one for the chrominance. There are 30 frames per second. A little arithmetic
gives a total data rate of approximately 20 MBs.
To achieve this speed, you need several disks ganged together, along with an array of
hardware and software to manage them. The cost of this kind of multi-disk arrangement
would be too expensive for most applications. A far more practical solution is to compress
the data, so that it can be sent at the disk throughput rates of available hard drives.
Compression is a mathematical process that reduces the amount of data in a given file.
There are many different types of compression processes (or algorithms) tailored for the
type of data they are designed to compress. Compression schemes that are optimized for
a specific type of data usually do not work as well for other data types. For example,
compression schemes that are suited for text files may not be suitable for image
compression.
Compression algorithms can be divided into reversible and non-reversible algorithms.
When data is compressed and then decompressed using a reversible algorithm, the result
is exactly the same as the original data. This type of compression is effective for data with
a lot of redundancy, such as text, but it provided little reduction in data size for
nonredundant data such as digital video data.
To reduce the data size of video images, non-reversible compression algorithms have been
developed that provide higher compression ratios than reversible algorithms.
Nonreversible algorithms do not give you back exactly the same data when they are
203
HRDV Reference Guide
Document 900.0261
04/05
Advertisement
Table of Contents
