Why Do Videos Take Up So Much Space? (And Why Compression Works)

You know the feeling: you recorded four minutes of your kid’s recital and your phone is suddenly asking you to free up space during the recital. Meanwhile the same phone holds a thousand photos without breaking a sweat.

There’s a real answer to “why are videos so enormous?” — and understanding it explains everything else: why compression is possible, why a compressed video can look identical, and why the camera industry is quietly shipping you a problem with every new sensor.

The absurd math of raw video

Start with what video actually is: a sequence of still images, played fast enough to look like motion. A single 4K frame is 3840 × 2160 × 3 color bytes ≈ 24 MB of raw pixel data.

At 30 frames per second:

24 MB × 30 fps = 720 MB per second — roughly 43 GB per minute of raw footage.

A 10-second clip would fill a 500 GB drive. Every phone would hold ~60 seconds of video, ever.

That’s the world without compression. The fact that a minute of video is “only” 400 MB instead of 43,000 MB is entirely the work of the encoder — the reason video exists at all is compression. Your camera already compresses aggressively; a video compressor just continues the same job, with better knowledge of what you need.

The one formula that governs everything

Compressed video size follows one law:

file size ≈ bitrate × duration

Bitrate is the data budget per second, and your camera sets it generously — 50–100 Mbps for 4K60 — because:

• it can’t know what you’re about to record (confetti? a static wall?), so it budgets for the hardest case,
• and storage is your problem, not the camera’s.

This is the entire engine of the “storage full” experience: your camera spends like quality is free, and you inherit the bill. (Everything about choosing bitrates flows from this one formula.)

What the encoder throws away

How do you get from 720 MB/s of raw pixels to 10 MB/s of video? The encoder finds and discards three kinds of data your visual system doesn’t actually register:

1. Spatial redundancy — repeated pixels

A sky is a million nearly identical blue pixels. A wall, a road, a tablecloth — huge areas of every frame repeat the same values. The encoder stores the pattern once (“blue, continues this way”) instead of storing every pixel. Encoders describe regions, not dots.

2. Temporal redundancy — repeated frames

Between two consecutive frames at 30 fps (33 ms apart), almost nothing changes: the background sits still, only the subject moves. Instead of storing a whole new frame, the encoder stores a motion description: “everything same, this hand moved 4 pixels left.” Static scenes cost almost nothing to encode — the insight behind why screen recordings compress so well.

3. Perceptual irrelevance — detail you can’t see

Human vision has finite resolution in detail, contrast, and color. Fine texture in busy areas, subtle brightness differences in dark regions, slight color shifts between adjacent pixels — all below the threshold of perception. The encoder ranks every piece of information by visibility and starves the invisible first.

Everything “compression” does is choosing what to starve next. A camera encoder starves nothing (it has budget to burn). A good compressor starves exactly the imperceptible, and stops the moment visible quality would suffer.

Why your eye lets 90% go

The punchline: most of your camera’s bitrate budget purchases detail you cannot perceive.

• Your phone screen is ~1080p — a 4K video’s extra pixel rows never reach your eye (when 4K actually matters).
• At normal viewing distance, your eye resolves far less than screens reproduce.
• Dark scenes and fast motion hide detail loss almost completely.

So when a compressor re-encodes 4K footage down to 1080p at a tenth of the bitrate and you can’t spot the difference — that’s not a trick. The discarded 90% was invisible at purchase. You’re cancelling a subscription to detail you never saw.

This is also why compression quality is about the destination, not the source: footage watched on phones, shared to apps, and stored for memory-keeping can carry far less data than footage destined for a 4K TV or an editing timeline.

Why videos will keep getting bigger

Every camera generation raises bitrate to match its better sensor — 4K60 became 4K120, 8K arrived, and each multiplies the budget. The industry’s answer to storage is always “buy more.”

But compression improves alongside: HEVC does H.264’s job at 60–70% of the cost (comparison), AV1 promises more, and on-device compression means the bill is reclaimable after the fact. The practical answer for your phone is:

1. Record at settings matching your real use (1080p beats 4K for most people),
2. Compress finished videos on-device, before they pile up — the iPhone playbook / the Android playbook.

The takeaway

Videos are huge because raw video is impossibly huge — and your camera overcompensates, spending bitrate on detail your eye will never see. Compression works because your eye is lossy: it discards most of what arrives at your retina.

A tool like Kompress just runs the math in your favor: re-encode on-device, discard the imperceptible, keep the memories, hand the gigabytes back. First run free, nothing uploaded — App Store · Google Play.