How much video storage do surveillance cameras need?

The five variables

Surveillance storage is driven by camera count, resolution, codec, frame rate, and retention period. Change any one and the total moves. The goal of sizing is to hold enough usable evidence for the retention policy without paying for capacity you never read.

A working estimate

A single 1080p H.264 camera at 15 fps consumes roughly 5–10 GB per day. The same camera on H.265 roughly halves that. A simple model:

• Total TB = (cameras x per-camera daily GB x retention days) / 1000
• Then adjust for recording mode: continuous recording uses the full figure; motion- or event-based recording typically cuts volume 40–70%.

So 100 cameras at 7 GB/day for 30 days continuous is about 21 TB; switch to H.265 and motion recording and the same estate can fall to a fraction of that.

What pushes it up — and down

• Up: 4K and higher resolutions, higher frame rates, longer retention, continuous recording, audio.
• Down: H.265 and smart codecs, dynamic frame rate, region-of-interest encoding, motion/event recording, and edge retention that keeps only flagged footage in the cloud.

Plan for retention policy first

Retention is usually the biggest single multiplier and is often set by regulation or incident-investigation windows. Decide the policy, then choose codec and recording mode to meet it economically. Tiered retention — keep everything for a short window, keep flagged events longer — is the common compromise.

How VMukti helps size it

VMukti Cloud VMS supports H.264 and H.265, smart and motion-based recording, edge retention of 7–30 days, and elastic cloud storage, so capacity scales with need rather than being capitalised up front. Proprietary compression reduces bandwidth and storage by up to 96% versus naive encoding, which is part of how the platform processes more than 1 billion camera feeds annually across 900+ deployments while keeping storage costs predictable.