Synchronization Basics

[ LiB ]

As I said, synchronization is the process of keeping two or more audio or visual systems running at the same speed and with the same relationship to each other. Keeping audio in sync with video means playing both together from the same point without either one moving faster or slower than the other. It is crucial to do this with a high degree of accuracy and repeatability , as every single edit and mix move you perform must be matched up with a certain action onscreen.

Two things are necessary for accurate sync. The first is a positional reference, or the "where-in-time." The second is a playback rate, or "how fast." I like to use the analogy of cars on a two-lane highway. The positional references are the mile markers on the side of the highway that tell you where on the road you are. The playback rate is the speed of the car, for instance 65mph. If we know what mile marker we're at and how fast we're traveling, it's easy to get another car moving along in sync with our car. Think of the audio as one car and the video as a second car in the next lane. If either car speeds up or slows down, the two cars will no longer be in sync. The drivers of these two cars must be in constant communication with one another in order to maintain the correct speed and position. With Pro Tools and digital audio, this communication is handled in two ways, one for positional information and the other for speed reference.

Digital audio is based on a sampling frequency that determines how often the analog audio is digitally sampled and turned into numeric information. The source of this sampling frequency is referred to as the digital clock and is required for proper operation of any digital audio device. If the sampling frequency or clock speed is changed, then the speed at which audio is played back will change accordingly . If the clock speed increases , digital audio will play back faster and higher in pitch. If the clock speed slows, then audio will play back slower and lower in pitch. Thus, the sampling frequency can control the speed of the audio. This is how the speed reference for synchronization is communicated when using Pro Tools.

Positional Reference

Unique numbers must be assigned to each segment of audio and video in a project in order to know exactly where you are at any given point in either medium. For instance, bars and beat numbers could be used to identify locations in a piece of music. In visual systems, each frame of film or video must be given a unique number. This is accomplished through the use of SMPTE (Society of Motion Picture and Television Engineers) time code.

Every frame of film or video in an AV project may have a specific SMPTE time code number assigned to it. SMPTE time code is the standard method by which we get positional references in audio, film, and video (see Figure 3.1). SMPTE time code represents time in an eight-digit number with colons separating each time value for hours, minutes, seconds, and frames (00h:00m:00s:00f). A frame of SMPTE can be as small as ¹ / ₃₀ th of a second or as large as ¹ / ₂₄ th of a second, depending on how fast the film or video is moving. I'll discuss the different frame rates in more depth in just a bit. For now, understand that SMPTE time code numbers are positional referencesthey tell you where you are in a soundtrack or film.

Figure 3.1. A SMPTE readout displaying 1 hour , 11 minutes, 45 seconds, and 24 frames.

Playback Rate

The second thing you need to know to achieve accurate synchronization is how fast the frames are moving. In my highway analogy, this relates to the speed of the two cars. In film and video, the playback rate is the frame rate. In digital audio, the playback rate is the sample rate or sampling frequency. The speed relationship between the audio and video must remain the same in order to be in sync. Otherwise, the audio will either slip ahead or behind the video.

Several standard frame rates for film and video are commonly used. They are as follows :

24 frames per second for Film
25 frames per second for European or PAL Video
29.97 frames per second for NTSC Color Video
30 frames per second for audio only or old B&W Television
23.976 frames per second for HD Video that converts to NTSC Video

In digital audio, the rate at which sound is played back is not measured in frames but in samples. Each sample of audio is a unit of information, or slice of time, much in the same way that each frame of film or video is a single snapshot of time, although an audio sample is exponentially smaller in duration. The rate at which samples are played back (the sample rate ) is usually expressed in kilohertz (abbreviated kHz); multiplying the kilohertz number by 1,000 results in the number of samples per secondfor example, 44.1kHz represents a rate of 44,100 samples per second.

There are several sample rates commonly used in audio production. They are as follows:

44.1kHz. The standard music sample rate. This is used to create the commercial music CDs we buy everyday.
48kHz. The video and film standard. Professional digital video formats such as Sony's Digital Betacam use this sample rate.
88.2kHz. Hi-res music sample rate for audio CDs. Easily converts to 44.1 kHz.
96kHz. Newer hi-res sample rate for better high-frequency response. Can be used for video since it is a multiple of 48kHz and is easier to convert.
192kHz. Super hi-res sample rate for demanding recordings such as orchestral and jazz. Currently impractical for film and video due to large file sizes, but may be used in the near future.

To sum up, in order to synchronize audio and video, you must know the positional information, in the form of SMPTE time code, and playback speed. Audio and video must start from the same position and play back at the same rate in order to maintain perfect sync. This might sound simple, but as you will see, it can get quite complicated once you consider all the variations in SMPTE formats and playback rates.

[ LiB ]