CCTV Hard Disk Recording Days
The Arithmetic, Shown in Full (2026)

The one formula, and why it is the only honest answer

Every recorder is doing the same simple thing: writing a stream of data to a disk until the disk is full, then writing over the oldest footage. So how many days you keep is not a setting you can turn up. It is arithmetic:

Days of footage = drive capacity (GB) × 0.9 ÷ (number of cameras × bitrate in Mbps × hours recorded per day × 0.45)

That 0.45 is just unit conversion, and it is worth seeing where it comes from rather than trusting it. One megabit per second is 1,000,000 bits every second. Divide by 8 to get bytes: 125,000 bytes per second. Multiply by 3,600 seconds to get an hour: 450 MB per hour, per megabit per second. In other words 1 Mbps = 0.45 GB per hour = 10.8 GB per day, per camera.

Anyone who tells you “this system records 30 days” without asking how many cameras, at what bitrate, and recording for how many hours a day, is telling you nothing. The whole calculation is above; the rest of this page is about getting each input right and what to do with the answer.

Step 1: find your recorder’s actual bitrate

This is the number people guess, and it is the only one that really matters. Do not use a figure from a web page, including this one. Your recorder knows its own setting and will show it to you.

On almost every DVR or NVR the path is similar: the recording or encoding settings page, per channel, where you will find a stream type (main stream and sub stream), a resolution, a frame rate, a bitrate type and a bitrate in Kbps or Mbps. The main stream is the one being recorded; the sub stream is the low-quality copy used for phone viewing and takes a small fraction of the space. If the bitrate is shown in Kbps, divide by 1,000: 4,096 Kbps is 4 Mbps (recorders use 1,024 to the megabit).

Two things to note while you are in there. If the bitrate type is constant, the camera writes that number all day and the arithmetic below is exact. If it is variable, the figure shown is a ceiling and the real average depends on how much movement the scene contains — a camera watching a quiet corridor will write far less than one pointed at a busy road with trees moving behind it. With variable bitrate, treat the calculation as a worst case, which is the right way round for planning.

Step 2: gigabytes per camera, per day

Take the bitrate you just read and find it here. These are conversions of the formula above, not claims about what any system is set to — substitute your own number:

If your bitrate is not in the table, multiply it by 10.8 for the daily figure. A camera set to 3 Mbps writes 32.4 GB a day; one set to 5 Mbps writes 54.0 GB.

Step 3: multiply by cameras, divide into your drive

Now multiply by the number of cameras recording, and divide the drive capacity by the result. This table does that for continuous 24-hour recording on a 2TB and a 4TB drive, using 2,000 GB and 4,000 GB as the drive-maker’s capacity and keeping 10 per cent back as usable-space reserve:

Read the bottom rows before you accept a 16-camera quote with a free drive in it. Sixteen cameras at 8 Mbps — not an unusual setting once somebody specifies 4K across the board — fill a 2TB drive in 1.3 days. That is not a fault. It is what the arithmetic says, and it is why the drive line on a quote deserves as much attention as the camera line.

Worked example: eight cameras on a 2TB drive

A typical Klang Valley shoplot: eight cameras, main stream set to 4 Mbps, recording continuously, on the 2TB drive that came with the package.

1. One camera at 4 Mbps writes 4 × 10.8 = 43.2 GB per day.
2. Eight cameras write 8 × 43.2 = 345.6 GB per day.
3. A 2TB drive holds about 2,000 GB. Keep 10 per cent back and you have 1,800 GB to promise, so 1,800 ÷ 345.6 = 5.2 days.
4. The same system on a 4TB drive: keep the same 10 per cent back and 3,600 ÷ 345.6 = 10.4 days.
5. The same system recording only 8 hours a day on a schedule: 115.2 GB per day, so a 2TB drive gives 15.6 days and a 4TB drive 31.2 days.

Every step there is arithmetic you can redo with your own bitrate and camera count in about a minute. Do it before you sign, not after the incident.

Motion recording and schedules: the multiplier you choose

The easiest way to buy more days without buying a bigger drive is to record less of the day — and the honest way to model it is as a fraction that you choose, not a percentage anyone can quote you.

If a camera records only 8 of the 24 hours, it writes one third of the data and you get three times the days. If it records 12 hours, half the data and twice the days. Substituting your own fraction into the formula is the whole method.

Two warnings about motion recording. It needs pre-record and post-record buffers set, or clips start after the event; and false triggers — rain, moving foliage, a spider on the lens at night, headlights — can turn “motion only” into “continuous with gaps”, which is the worst of both. If footage genuinely matters, continuous recording on a drive sized for it is the more reliable engineering choice.

H.264 and H.265: the same picture, fewer gigabytes

Compression is the other lever. H.265 is a later codec than H.264 and is designed to deliver comparable picture quality at a lower bitrate; how much lower depends on the scene and the encoder, so the honest way to use it is not to apply a percentage but to switch it on and read the new bitrate off the recorder — then rerun the arithmetic with that figure.

Three practical points. The camera and the recorder both have to support the codec, so it is a system decision rather than a checkbox. Older phones, browsers and third-party software sometimes struggle to play H.265 exports, which matters on the day you have to hand a clip to somebody else. And some manufacturers ship a further “smart” variant that lowers the bitrate on static parts of the scene — useful, but check what it does to detail on the part of the frame you actually care about before trusting it.

Resolution and frame rate: the two dials that move the bitrate

Bitrate is not chosen in a vacuum. It follows from what you have asked the camera to do:

• Resolution. More pixels per frame means more data per second at the same quality. This is why speccing 4K on every position is a storage decision as much as a picture decision — and why it belongs on the positions that need to read a plate or a face, not everywhere. IP versus analog CCTV covers where resolution is worth paying for.
• Frame rate. Halving the frames per second roughly halves the data. For most surveillance a moderate frame rate is entirely adequate; cinema-smooth motion is rarely what you need from a camera pointed at a gate.
• Scene content. With variable bitrate, a busy scene costs more than a still one. Trees, rain, traffic and flickering light all raise the real average, which is why a variable-bitrate calculation should be treated as a floor rather than a promise.
• Night mode. Infrared images with noise in them compress worse than clean daylight images, so the overnight hours can write more than the daytime ones.

What a “2TB” drive really holds

Two adjustments make the tables above a ceiling rather than a promise, and both are worth knowing before you are disappointed.

First, drive makers count a terabyte as 1,000,000,000,000 bytes while an operating system counts in powers of two, so a “2TB” drive is reported by most software as roughly 1.82TB. The tables here use the drive maker’s figure — 2,000 GB and 4,000 GB — because that is what is printed on the box, so if your recorder shows a smaller number, that is why.

Second, a recorder does not use every last byte: it reserves space for its own database and index, and it begins overwriting before the disk is literally full. So take the calculated figure as an upper bound and give yourself margin rather than sizing a drive to exactly the days you need.

The practical version of both points: if you need a fortnight of footage, do not buy the drive that calculates to exactly fourteen days.

Choosing the drive: surveillance-grade, not desktop

A CCTV recorder writes continuously, every second, for years — a duty cycle an ordinary desktop drive is not built for. Surveillance drives (the purple and hawk-branded families) are specified for continuous writing and for several drives sitting in one chassis. Malaysian marketplace prices for them look like this. These are retail shelf prices for a product, shown as floors because the listing spread is enormous, and they are not part of any installation cost:

One oddity in those rows is worth flagging rather than hiding: the 1TB floor sits above the 2TB floor, because 1TB stock is thin and the listings are strange. So do not draw a neat price-per-terabyte curve through these three numbers — check the current listings for both capacities before buying, and expect the larger drive to be the better value more often than not.

Other practical points: check the maximum drive capacity your recorder supports before buying a large disk, and check how many bays it has. Fitting a desktop drive to save money is a false economy that usually ends with a drive failure and no footage from the period you needed.

How many days do you actually need?

There is no rule we can hand you here, and any page that states one as a general legal requirement is overreaching. It is a business and engineering judgement, and the useful way to make it is to ask how long it realistically takes for somebody to notice that something happened:

• A break-in or an accident is noticed the same day. Days of footage would cover it.
• A missing item from a store room, a damaged car in a car park or a delivery that never arrived is often noticed a week or two later.
• Anything involving a holiday, a closed shop, a rental unit between tenants or an owner who lives elsewhere stretches to weeks.
• Anything that becomes a dispute takes longer still, because the request for footage arrives after the argument starts.

Set the target from the longest of those that applies to you, then size the drive from the arithmetic. And keep in mind the reverse point: footage you never need is footage you are storing about people, so retaining far more than you have a reason to is not a free choice either. Sizing deliberately is better than sizing by accident in both directions.

When one drive is not the answer

• A second drive. Many recorders take more than one. Two 4TB drives are usually simpler and cheaper than an exotic single drive, provided the recorder supports the capacity.
• Redundancy. Some recorders can mirror to a second disk so a drive failure does not lose the footage. It costs a drive and buys resilience, not extra days — understand which of the two you actually want.
• Network storage. Larger sites record to a NAS. This is a project rather than an accessory, and it needs a network designed for it.
• Export the clips that matter. The cheapest retention strategy of all: when something happens, export the clip that day and keep it somewhere else. No drive size protects footage you never took out of the recorder before it was overwritten.
• Cloud. Practical for one or two cameras, expensive and bandwidth-hungry for a whole system, and it uses the upload side of your broadband, which is the smaller number.

What the quote must state about storage

• The drive capacity, in terabytes, and whether it is a surveillance-grade drive.
• The main-stream bitrate and resolution being set for each camera type.
• The recording mode: continuous, scheduled, or motion-triggered — and the schedule if there is one.
• The recording days those three produce, stated as a number on the quote. This is the line to insist on.
• Whether the recorder has a spare drive bay, and the maximum capacity it accepts.
• Who is responsible if the drive fails inside the warranty period, and whether the labour to swap it is included.

The rest of the lines a comparable CCTV quote should carry are on what CCTV installation costs in the Klang Valley, and the items most often left off a cheap package are on the “free installation” page.

Checking it on your own system, afterwards

You do not have to trust anybody’s calculation, including this one. Once the system has been running long enough to have overwritten once, open the playback calendar and find the earliest date that still has footage. The gap between that date and today is your real retention, measured rather than modelled.

Make it a habit: check it monthly, along with two other things worth thirty seconds each — that every camera is still online, and that the picture on each is not obscured by a web, dust or condensation. Cameras fail quietly, and a system nobody checks is a system that will turn out to have been broken since the last storm.

If the measured retention is much shorter than the calculation, the usual causes are a higher real bitrate than the setting suggested, cameras recording continuously when you thought they were on a schedule, or a sub stream also being recorded. All three are visible in the recorder’s own settings.

Why we quote after a site survey

Storage is one of the few parts of a CCTV job that can be settled on paper before anyone visits — but only once the camera count, the positions and the resolution each position needs are settled, and those come out of the survey. ClickBina does not publish a package price for CCTV, because the cabling route decides the job; what we will do is tell you what recording days your proposed system gives before you agree to it, and size the drive to the answer you actually want.

Send photographs of the positions you want covered and where the recorder will live, and we will come back with the specification and the arithmetic behind it. See the rest of our electrical and plumbing services, or compare the two system types first on IP versus analog CCTV.