Image Processing Reference
a little bit better, as will S-VHS recordings, but if you have the higher-quality production
master tapes available you will achieve better results if you work from them.
You should avoid any unnecessary file conversions due to the risk of coding artifacts'
being introduced. The source material might be presented in a format that cannot be oper-
ated on in the way you want. Perhaps you want to change the clip length or splice two seg-
ments together that are presented in different formats. Converting to an uncompressed
format will allow you to edit the content and then recompress again with minimal arti-
facts. This is not ideal, but it may be necessary.
Editing Styles and Implications of Different Transitions
If the film editor knows something about the compression system used to create the fin-
ished program, some editing decisions can be made to avoid compression artifacts. The
edited program will suffer far fewer problems when it is delivered through a low-bit-rate
Video editing is only feasible on video that is coded with all I-frames. DV25, for
example, is a motion JPEG format, and there is no dependency between one frame and the
next. This is a format that can be edited without any problems. Editing of any format
where there is a cadence or structure to the frames should be avoided. In these cases, you
should convert the format first and then edit. If you must restore the original format, then
do so when the editing is complete. For example, removing 3-2 pulldown from telecine
footage before editing is vital. Otherwise, the cadence will be compromised during edit-
ing and the pulldown cannot be removed properly after that.
Cutting from one scene to another is an example of an editing technique that is sympa-
thetic to the compression process. Because the audio is dealt with separately, so-called
L-shaped edits are feasible. Figure 32-2 shows an L-shaped edit where audio and video are
cut at different times.