Digital Video Tutorial
DV terminology
Everyone starting out in digital
video finds themselves faced with a bewildering array
of new buzz-words, acronyms, and terminology. This glossary
aims to help you through the confusion by explaining
many of the commonly used terms.
Rather than just a
glossary of words in alphabetical order, the terms are
introduced in the order you would encounter them when
working though a typical video project. For quick reference
there is also an alphabetical index
and an Inspiration Map.
Linear editing
is the process of copying video between multiple VCR
machines, possibly including special effects processors,
etc. It can be manually or computer controlled. It
is fast but requires expensive equipment for good
results.
Non-linear editing (NLE)
is the process of manipulating video stored as digital
computer files. Special video editor software is used.
This allows video to be visually constructed in a
manner similar to cutting and sticking cine film.
This tutorial focuses on NLE.
|
Video computer files in the PC world
are usually either AVI
or MPEG
types. The process of getting video into a computer
from an external device such as a camcorder is called
capturing.
There are two main methods of capturing,
either analog
or digital.
You can use analog capture with almost
any device; you can only use digital capture with digital
devices that have a Firewire
output. Both methods require appropriate hardware installed
in the computer.
Analog capture normally requires a
capture card,
but MPEG capture can be done using an external device
on a USB or parallel port.
Digital camcorders use the Digital
Video (DV)
standard. This defines not only the size of the tape,
but also the format of the digital data. All DV camcorders
produce digital data in the same format. DV and Mini-DV
are the same except for the physical size of the cartridge
and hence the running time. Digital-8
cameras also use the same data format, but use a tape
that is physically the same as Video-8.
Firewire
(or IEEE1394) is a high-speed serial data connection,
used by almost all DV camcorders. It needs a Firewire
port on the computer (often as a plug-in card), but
there is now a trend in high-end computers for Firewire
to be built-in.
DV captured by Firewire is stored
in AVI files (when using Microsoft Windows). There are
two types, Type-1
and Type-2.
An AVI file can contain multiple
streams, usually one video and one audio. DV is a
data stream that itself contains video and audio.
In a Type-1 AVI, the whole DV stream is stored unmodified
as one AVI stream. In a Type-2 AVI the DV stream is
split into separate video and audio data, which are
stored as two streams in the AVI.
The advantage of Type-1 is that
the DV data does not need to be processed, and is
stored in its exact original format. The advantage
of Type-2 is that it is compatible with video software
that is not specially written to recognize and process
Type-1 files.
There are several limits that affect
the maximum size of an AVI file.
The original specification for AVI
files limits their size to 4GB, but for software reasons
the maximum workable size is 2GB.
This corresponds to about 9.5 minutes of DV format
video.
OpenDML
is a standard that extends the AVI specification to
effectively remove the file-size limit. If video software
and drivers support OpenDML then the size of AVI files
is only limited by the operating system and size of
disks.
File sizes are also limited by the
disk format used by the operating system.
FAT32,
supported by Windows 95 onwards, does not allow any
single file larger than 4GB.
NTFS,
supported by Windows NT & Windows 2000, allows
effectively unlimited files and partitions.
Once video has been captured to disk
files, it can be edited.
This is done using a special
video editor
program, and the majority of non-linear editors, such
as Video Studio (we will be using this!) or MediaStudio Pro
(more advanced - for professionals!) or Adobe
Premiere (Adobe's non-linear editor), are based on the idea of
a timeline.
This presents a visual workspace on the screen, representing
time from left to right. The finished video is composed
in the way a cine film might be composed by laying out
film clips on a bench. In the NLE process the clips
are disk files containing captured video.
Different editors have different capabilities,
but most will include transitions
allowing various effects when going from clip to clip,
filters
allowing the look of the video to be altered, and overlays
allowing multiple clips to be combined in various ways.
When everything has been laid out
as required in the timeline, the editor program renders
a new video file which contains the result. The original
files are not altered, the computer combines information
from the files, transitions, effects, and overlays and
computes a new file. Depending on complexity, this can
require large amounts of computing power resulting in
a long rendering time
(sometimes more than 10 times running time). The faster
the CPU the faster this process runs.
If the result is intended for recording
to tape (e.g. DV), the rendered file must be played
on the computer, and an output taken to a VCR or recording
camcorder. In the case of DV, the output is again the
Firewire connection, allowing the pure digital data
to be recorded onto tape. If a camcorder is to be used
as the external recorder, its Firewire connection must
be able to record input as well as replay output. Such
a capability is called DV-in
and not every camcorder supports it, especially if it
is made in Europe.
Both the capture and output-recording
processes are simpler and more accurate if the computer
can control the tape functions of the camcorder/VCR.
OHCI is
a standard for sending controlling commands over the
Firewire link. If supported by the digital camcorder/VCR
and by the computer software, then no manual control
is required and the external device effectively becomes
a computer peripheral.
There are many more technical terms
used in connection with video on computers, but we hope
this has helped to explain many of the frequently encountered
terms.
|