Y2K Cam ▸ Blog ▸ What VHS does

What a VHS Tape Actually Does to Footage

Technical · July 22, 2026 · 6 min read

The short version: VHS degradation isn't one effect, it's five stacked ones — heavily reduced color resolution, a soft luma channel, head switching noise at the bottom of the frame, random dropouts, and compounding generation loss. Most "VHS filters" only reproduce the last visible layer (grain and scanlines) and skip the ones that actually make tape look like tape, which is why so much of it reads as a sticker over a modern image.

If you want the look to be convincing, it helps to know what the format was doing. Here's the signal path, layer by layer, and which layers matter most.

1. The color channel is crushed — far more than the brightness

This is the big one and it's the one most filters miss. VHS records luma (brightness) and chroma (color) separately, and chroma gets dramatically less bandwidth. In practical terms color resolution is a small fraction of brightness resolution.

What you see: color that bleeds past edges. A red jacket smears a couple of pixels into the wall behind it. Saturated reds are the worst offenders — they wobble, they trail, they refuse to stay inside the object. Meanwhile the edge of that same jacket, in brightness terms, is still reasonably defined.

That mismatch — soft color riding on comparatively sharper brightness — is the most identifiable VHS characteristic there is. A filter that just blurs and adds noise to the whole frame equally will never look right, because tape never blurred the whole frame equally.

2. Horizontal detail is limited, vertical detail isn't

Tape bandwidth limits how fast the signal can change across a scanline, which caps horizontal detail. Vertical resolution is set by the line count, which is unaffected. So VHS softness is directional: it smears sideways more than it smears up and down.

This is why real tape footage of a picket fence looks different from real tape footage of venetian blinds. A uniform Gaussian blur can't produce that asymmetry.

3. Head switching noise — the torn strip at the bottom

A VCR reads the tape with rotating heads, and the handoff between them happens just below the visible frame. On a slightly misaligned deck, or on any capture that includes the full frame, you get that unmistakable band of horizontal tearing and static at the very bottom of the picture.

Most people never consciously noticed it and everyone recognizes it instantly. It's a strong, cheap authenticity signal — but only if it's confined to the bottom few percent of the frame. Sprayed across the whole image, it reads as generic glitch, not tape.

4. Dropouts: brief white or black horizontal streaks

Tape is a physical medium. Dust, a worn oxide patch, a crease — and the head momentarily reads nothing. You get a short bright horizontal dash, one or two lines tall, lasting a frame or two.

The key property is that dropouts are sparse and random. A few per minute on a decent tape, constant on a bad one. Filters that add them at a steady rhythm break the illusion, because the whole character of a dropout is that it's an accident.

5. Generation loss — why the third copy looks the way it does

Every analog copy loses signal. The tape everyone's nostalgic for usually isn't a first-generation recording — it's a dub of a dub of a broadcast, and each pass compounded the color bleed, softened the luma further, and added noise.

That's worth knowing because "more VHS" doesn't mean "more grain." It means more of everything above, applied again. Pushing intensity should deepen the color smear and the softness, not just crank a noise layer.

What about the audio?

Easy to forget and disproportionately effective. Linear VHS audio is a narrow-band mono track with audible hiss, limited high end, and a slight instability in pitch from tape speed variation. Even a light version of that — roll off the highs, add a low hiss floor, sum to mono — does an enormous amount of work on a clip that already looks right.

So what can a filter actually reproduce?

Most of it, honestly — if it's built as a signal chain rather than an overlay. Chroma subsampling, directional horizontal softening, head switching confined to the frame bottom, sparse randomized dropouts, and mono band-limited audio are all reproducible in real time on a modern phone. That's the approach Y2K Cam takes, which is why the look holds up on footage that isn't already grainy.

What a filter can't give you is the shooting. A perfectly processed tape image of a steady, well-composed, tightly-edited clip still looks modern, because no one shot like that in 1998. That half is on you — we wrote a whole camcorder shot list about it.

Getting the look right, in order