Drop or non-drop? Pull-up or pull-down? 30 or 29? And what is 23.976? All these are questions that strike terror into the hearts of post-production audio engineers, sound designers and even composers.
Anyone who works with sound for television has his or her own nightmare stories about dealing with this issue. One of my favorites, from the days when analog multitrack ruled the post-production world, involved a studio that striped 29.97 SMPTE on one track and 60 Hz Nagra resolve tone on another, and then tried to get their synchronizer to lock to both. The hapless 5-centimeter machine bucked until the tape snapped.
Even now, in the age of digital, we still cannot escape problems with sync. I just went through this, working on a score for a 15-minute film on Beta SP video that was being posted by an expensive transfer house in New York. They asked me to provide the music in ordinary DAT format and to tell them at what frame of the film to start the audio.
When I got the videotape back, the audio started right on the money, but it ended about 40 frames after the picture did. For the next three days, the transfer house and I argued about mismatched frame rates. We used up my monthly allotment of cell phone air time, and then some, arguing about how to make the audio and the video come together. They tried different DAT machines, different video machines and different operators. They tried transferring it from DAT to another videotape, and pulled it up and down, and probably in and out as well. Nothing worked.
Ultimately, I determined that the DAT machine was not locked to house sync. All of the fretting about frame rates was unnecessary. As long as everything in the house was running from the same master clock, there was no problem.
For many of us, sync issues are routinely solved by having our computers do the required complex math on the fly. But the equipment only works correctly if it is set up and operated correctly. At some point in the process of scoring this particular film, my synchronizer changed its default clock value to 30 fps instead of 29.97 non-drop, and suddenly all my hits were in the wrong place, and the MIDI tracks were not agreeing with the audio. That one took a couple of hours to find. (And when we could not get the first transfer to work, I wondered whether it was my fault that they were having problems.)
In Europe, video frame rates are nice, whole numbers. Europeans do not understand why Americans would make their video run 0.1 percent off or why they spend so much time and energy trying to deal with it.
The version that most people know is this: In the early days of television, when everything was shades of gray, the standard frame rate in the United States (and Canada) was 30 fps. Because each television frame consists of two “fields” of alternating lines, the field rate was 60 Hz.
When the color standard was adopted in 1953 by the NTSC, it was discovered that the new color signals were susceptible to interference from 60 Hz AC fields, and if there was any difference between the line frequency and the scan rate of the received signal, it would show up as a visual “beating.” But if you altered the field rate enough, the beating would speed up to the point that it was no longer visible.
AMS Neve Logic 2 digital mixing console and AudioFile hard disk editor in Chicago Recording Company’s new Studio 55A.
Unfortunately, this version is simply not true. Line-frequency beating never was a problem. And if it were, the cure is worse than the illness: At a field rate of 59.94 Hz, if there really were beating caused by AC-line leakage, you would see a bar roll across the screen about every 17 seconds. Not pretty.
The real reason is this: The lower frame/field rate is designed to prevent potential visual beating between the chrominance subcarrier in the broadcast signal and the audio subcarrier. Why is this a problem, and how does changing the frame rate help?
Well, for various reasons, the color subcarrier frequency in the television signal needs to be modulated onto the picture carrier at 455/2 times the horizontal line frequency. At 30 fps, and with 525 horizontal scan lines per frame, this frequency is 15.750 kHz, which means the color subcarrier would have to be at 3.583125 MHz. The sound subcarrier (as established on the first monochrome TV systems) is at 4.5 MHz.
If the two subcarriers were to interfere and beat against each other, the 916.875 kHz difference might be visible — and in fact, according to one report presented to the NTSC, it was visible in some monochrome sets at the time, under some conditions. Since backwards compatibility was a major consideration for the NTSC, this was a problem.
If the difference signal happened to be an odd multiple of one-half the scan rate, this beating would be reduced. If the frame rate were dropped 0.10001 percent, the scanning frequency would be 15.734264 kHz, the chrominance subcarrier would be 3.579545 MHz, and the beat product (if there was one) would be 920.455 kHz, which is very close to the 117th multiple of half the scan rate.
But a close look at the technical documents and the committee proceedings around this point seems to show that the problem never really existed.
An engineer who was there at the beginning, Rollie Zavada of Eastman Kodak, diplomatically calls the decision to change the frame rate “debatable.” Other sources say that the first generation of color sets, and the black-and-white sets that were made by the time the color standard was adopted, had good enough filters on the audio section that leakage between the subcarriers was simply not an issue. The decision meant expensive alterations to transmission equipment, as the AC line could no longer be used as a frequency reference for sync, according to video engineer Tim Stoffel.
While it is depressing and frustrating to realize that changing the frame rate to an irrational number probably was not necessary, what is sadder still is to realize that apparently we are never going to get away from it — although we had the chance to change the frame rate with the advent of digital television and HDTV but did not.
The original HDTV standards all specified a frame rate of 30 fps. Progressive, interleaved, 1080, 720, whatever variation you looked at, there was no mention of 29.97 anywhere in the proposals. But the HDTV programs now in production and going out over the air are running at — you guessed it — 29.97 frames.
The FCC mandate for HDTV is incredibly vague, and has over the years been increasingly dictated by the broadcasters themselves, which means networks and stations have been free to do just about anything they want with it. And dropping the frame rate is something that came easily.
Paul D. Lehrman is Insider Audio columnist for Mix magazine, and Web editor for Mix and related sites.