VHS video is read from diagonal tape tracks by rotating heads that alternate as each television field is recovered. The handoff happens near the field boundary, where one head is leaving useful tape contact, and the next is taking over.
On an old television, you often never saw the messy part. CRT sets commonly overscanned the picture, pushing the outer edges beyond the visible screen area, while modern capture hardware can preserve more of the raster. A thin torn strip at the bottom can therefore appear in a digital transfer even when the tape has not suddenly developed a new fault.
Head-switching noise is easy to confuse with tracking damage because both can bend or scramble horizontal bands. The difference is timing. Normal switching trouble stays tied to the bottom of each field, while mistracking can wander, broaden, or cut through more useful picture area.
A helical-scan head-switching system describes this selection directly, with alternate heads chosen according to drum rotation. In NTSC operation, the documented system switches between the relevant heads every 1/60 second, matching the field rhythm rather than some random burst of interference.
The outgoing head cannot keep reading useful information after it leaves the recorded track, and the incoming head needs to reach the right part of its next track. Switching at the wrong moment would expose a much uglier gap, so VCR designers place the handoff close to the vertical interval where television systems already expect non-picture timing information.
Real machines still leave a visible switching region. A few lines can become skewed, torn, gray, or horizontally displaced around the switch point because the recovered signal is not perfectly continuous during the transition. The artifact repeats in a fixed vertical region because the head change follows field timing.
Full-raster capture changes the deal. A capture device may digitize lines that a living-room television never showed, turning the old head-switching region into a clearly visible strip at the bottom of the file. Cleaning the heads will not automatically remove it because normal switching noise is not simply dirt on the tape.
Some playback decks handled the region more deliberately. Switching-point masking could replace troublesome bottom lines with black, while other workflows simply crop or mask the affected edge after capture. Neither method proves the underlying handoff vanished. One hides the region electrically and the other removes it from the finished frame.
A time-base corrector solves a different class of problem. It can stabilize line timing and reduce horizontal waviness caused by an unstable playback signal, but normal head-switching noise is tied to the rotating-head transition itself. Treating every bottom-edge disturbance as a time-base error can send you after the wrong repair.
Recording conditions can matter too. If the original machine placed its switching point badly, another VCR cannot necessarily reconstruct picture information that was never joined cleanly in the first place. Playback hardware may mask the symptom better, but the location and severity can already be baked into the recorded timing.
This distinction matters when recreating the effect in software. Generic static, random horizontal tearing, and tracking bars may look suitably damaged, yet none of them reproduce a disturbance locked to a field boundary. In Thenatan's rotating-head VHS treatment, head switching belongs beside other tape faults as its own behavior because its timing comes from a different mechanical event.
A convincing switch artifact should therefore stay spatially disciplined. Keep it near the lower field boundary, let its displacement and noise vary within that region, and avoid making it roam freely through the frame. Once the band starts behaving like random glitch art, the rotating drum has stopped being the thing that explains it.
On an old television, you often never saw the messy part. CRT sets commonly overscanned the picture, pushing the outer edges beyond the visible screen area, while modern capture hardware can preserve more of the raster. A thin torn strip at the bottom can therefore appear in a digital transfer even when the tape has not suddenly developed a new fault.
Head-switching noise is easy to confuse with tracking damage because both can bend or scramble horizontal bands. The difference is timing. Normal switching trouble stays tied to the bottom of each field, while mistracking can wander, broaden, or cut through more useful picture area.
Two rotating heads have to trade the signal cleanly
A VHS drum carries video heads across the tape on a helical path, producing the familiar diagonal recorded tracks instead of straight horizontal ones. During playback, electronics select whichever head is correctly crossing the tape, then hand the signal to the other head as the drum turns.A helical-scan head-switching system describes this selection directly, with alternate heads chosen according to drum rotation. In NTSC operation, the documented system switches between the relevant heads every 1/60 second, matching the field rhythm rather than some random burst of interference.
The outgoing head cannot keep reading useful information after it leaves the recorded track, and the incoming head needs to reach the right part of its next track. Switching at the wrong moment would expose a much uglier gap, so VCR designers place the handoff close to the vertical interval where television systems already expect non-picture timing information.
Real machines still leave a visible switching region. A few lines can become skewed, torn, gray, or horizontally displaced around the switch point because the recovered signal is not perfectly continuous during the transition. The artifact repeats in a fixed vertical region because the head change follows field timing.
Overscan hid a defect that digital capture exposes
Traditional television viewing hid much of the problem. Manufacturers expected the edges of the transmitted picture to fall outside the safely visible area on many CRT sets, so several untidy parts of analog video could live near those borders without bothering viewers.Full-raster capture changes the deal. A capture device may digitize lines that a living-room television never showed, turning the old head-switching region into a clearly visible strip at the bottom of the file. Cleaning the heads will not automatically remove it because normal switching noise is not simply dirt on the tape.
Some playback decks handled the region more deliberately. Switching-point masking could replace troublesome bottom lines with black, while other workflows simply crop or mask the affected edge after capture. Neither method proves the underlying handoff vanished. One hides the region electrically and the other removes it from the finished frame.
A time-base corrector solves a different class of problem. It can stabilize line timing and reduce horizontal waviness caused by an unstable playback signal, but normal head-switching noise is tied to the rotating-head transition itself. Treating every bottom-edge disturbance as a time-base error can send you after the wrong repair.
Severe switching noise can still signal a real problem
Normal does not mean every amount is normal. A narrow, stable band at the bottom fits ordinary VHS behavior, while a switching region that suddenly grows, jumps vertically, or invades more active picture can point toward servo alignment, switching-point adjustment, or deck-specific playback trouble.Recording conditions can matter too. If the original machine placed its switching point badly, another VCR cannot necessarily reconstruct picture information that was never joined cleanly in the first place. Playback hardware may mask the symptom better, but the location and severity can already be baked into the recorded timing.
This distinction matters when recreating the effect in software. Generic static, random horizontal tearing, and tracking bars may look suitably damaged, yet none of them reproduce a disturbance locked to a field boundary. In Thenatan's rotating-head VHS treatment, head switching belongs beside other tape faults as its own behavior because its timing comes from a different mechanical event.
A convincing switch artifact should therefore stay spatially disciplined. Keep it near the lower field boundary, let its displacement and noise vary within that region, and avoid making it roam freely through the frame. Once the band starts behaving like random glitch art, the rotating drum has stopped being the thing that explains it.
