Analogue Wayback, Ep. 11

Bumping your capstan.

I started out editing in an era of wrestling edits out of quad VTRs, so I tend to have less concern when there’s an issue with some plug-in. Not that it can’t be a problem, but it’s just one more indication of how far the industry has come.

In the 70s and 80s, the minimum configuration of an online edit bay involved three VTRs, a switcher, audio mixer, and the edit controller. Two VTRs were for playback and the third was what you edited onto. You needed both players to make a dissolve. If there was only one camera reel, then before starting the session, the editor would often make a complete copy (dub) of that camera reel. Once copied, you now had the A-Roll (camera original) and a B-Roll Dub to work from. You could roll A and B together and make a dissolve in a single pass, laying down clip 1 and clip 2 with the dissolve in-between. If it was a series of dissolves, then this required matched-frame edits in order to dissolve from the end of clip 2 to clip 3, then the same from clip 3 to clip 4, and so on.

To be completely seamless, the matched-frame edits had to be perfect. There’s the rub. In simple terms, NTSC and PAL are systems where the color signal rides on top of the black-and-white signal. This involves a colorburst signal and a sync pulse. NTSC follows a cadence of 4 fields (2 interlaced frames) in which the phase of the signal repeats every other frame. This cadence is known as the color frame sequence. When you play back a recording and the VTR first achieves servo-lock, it can lock up usually in one of two phase conditions as it syncs with the house sync generator. This slightly affects the horizontal position of the picture. 

If you record clip 1 and the VTR locks in one horizontal position, then when you make the matched-frame edit onto the end of clip 1, the VTR has to lock up again in that same position. If not, then there will be a slight, but noticeable, horizontal shift at the edit point. It’s a 50/50 probability of where the deck locks up. Some of the Ampex decks featured a bit more control, but the RCA TR-600 models that we were using tended to be sloppier. If you got an H-shift at the edit, you simply repeated the edit (sometimes several times) until it was right.

The facility hired a sharp young chief engineer who took it upon himself to create a viable workaround, since RCA was never going to fix it. His first step was to add an LED onto the front of one of the circuit boards as an indicator. This was visible to the editor when the VTR panels were open. This indicator could be monitored through the glass that separated the edit suite from the VTRs. Polarity condition 1, LED on. Condition 2, LED off. His next step was to add a remote switch for each player VTR next to the edit console. The editor could trigger it to “bump” the capstan control. This would cause the VTR to unlock and quickly relock its playback.

If the LED was on when recording the first part of the clip, then on the second edit the VTR would need to lock with the LED on, as well. If so, you’d achieve a successful matched-flame edit without any H-shift. Quad VTRs would lock up in anywhere from under one up to ten seconds or longer. The editor would monitor the LED status and could control the preroll length, which was generally five seconds for the TR-600s. During a matched-frame edit, if the condition was wrong, hit the switch and hope that the deck would lock up correctly before the end of the preroll. Otherwise lengthen the preroll time. This process worked better than expected and quickly became second nature.

At the risk of moving into the “kids, get off my lawn” territory, young editors clearly don’t know the fun they are missing with today’s modern nonlinear edit systems!

©2022 Oliver Peters

Advertisement