Video
Multi-camera Mix - 3
The Editing Process
This is where the work really begins! You need to mix together, for the duration of the performance, one video tape from each camera plus one sound tape, maintaining perfect synchronisation. It sounds pretty difficult, but by keeping to the system I describe it's not as painful as you might imagine.
Divide the work into sections
Even if you have enough of disk space, it's difficult to work on the whole of a performance at once, so it's best to break it down into sections. The show I worked on was in revue format, consisting of sketches and songs, none more than 10 minutes duration. I was thus able to work on one or two items at a time, record the result back to tape, then move on to the next. For a performance such as a play, with fewer breaks, you will have less flexibility.
You may be limited by available disk space (becoming ever less of a problem as disk sizes continue to rocket), and also by the maximum size of a singleAVI file. This latter limit depends on the capture system you use. Systems based on Video for Windows (VfW) can't use single files greater than 2 GB. At DV data rates that's 9.5 mins, or using analogue at a lower rate of 2MB/sec it's 16 mins. Microsoft's DirectShow (Active Movie) architecture allows "unlimited" files in today's terms, but you may still be limited by the disk architecture.
DirectShow is currently almost exclusive to Windows built-in DV drivers, typically used with OHCI 1394 ports, but if your disks are formatted using FAT32 you can't have a single file larger than 4 GB - that's 19 mins of DV. Only by using NTFS, which itself requires Windows 2000, can you use single DV files that run longer than 19 mins.
To work out how long a section you can manage at once, you need disk and/or file space to capture shots from all the cameras, plus enough to create the finished video. You also need space for the PA sound recording, but that is small in comparison to the video files. As an example, if you use DV with a maximum file size of 4 GB then the longest section that can be worked on easily is 19 mins.
To
work on 19 mins. worth of the show in one go, you need to capture files
from both the static cameras (8 GB), plus the hand-held (if it was in
use for half the time that's another 2 GB). You also need 4 GB space to
create the final video. That's 14 GB of disk space (right). If
you have more cameras, you'll need proportionately more, but if your editor
has the ability to play an assembled video directly from the timeline
(e.g. MediaStudio Pro 6 with an OHCI Firewire card), then you
don't need space for the final output file.
Do your sums according to your setup, then you'll be able to organise how you're going to work with the material to best effect.
A bit more ground-work
Before you start capturing your video, you have a few more things to do.
If you've used more than one source for the auditorium sound, decide by listening which is best. This will become part of the final sound-track, the other sound-tracks will only be used for synchronising. We'll assume this track is on one of the fixed cameras, and I'll refer to it as "camera-1".
You need to get a good sound level for this sound-track when capturing, and also a good level for the PA sound when you record that to disk. Use your audio editor, or whatever sound-recording tool you prefer to use, to run some test recordings of both these audio sources. Adjust the input controls to get a good, but not overloaded, level. Make a note of the settings for each source. If you have a sound mixer with presettable buttons, these come in very handy here.
However, if you are using DV the sound is captured digitally at whatever level it was recorded in the camera. Any adjustments have to be made at the editing stage.
For video, you should also try to get as close a colour match as possible between the different cameras. Again this can only be done at the capture stage with analogue equipment, and with amateur gear it is usually a compromise, but with care you can get a reasonable match. If you have a camera with a colour-bar generator, or took the step of recording a colour card on each camera, you can make use of the vectorscope in MediaStudio's capture program. Otherwise you can do it "by eye" using a well-lit and brightly coloured scene. Whatever the method, find the best brightness/contrast/colour/hue settings on your capture card for each source tape, to get the results as close as possible. Make a note of the exact settings for each, so you can return to them for each capture session.
Returning to sound, decide what sample rate you are going to use, and use it both when capturing sound with video, as well as recording the separate PA sound. You may be limited in choice by the technology, e.g. DV will be either 32 KHz or 48 KHz, so record analogue sound to match. It's also pointless to record in stereo unless the source really is stereo. For example, if you're lucky enough to have a true stereo PA system then you can make use of a stereo recording, otherwise just record in mono.
Site sections:
