H.264 vs H.265 vs H.265+ for CCTV: Which Codec Saves the Most Storage?

When comparing CCTV cameras, most buyers focus on resolution — megapixels, 4K, wide angle. But codec choice often has a bigger impact on your total cost of ownership than resolution, because it directly determines how much hard drive space you need in your NVR. A 4MP H.265+ camera can store footage at roughly one-third of the HDD space consumed by the same camera on H.264 — while delivering the same visible image quality. This guide explains what each codec actually does and what the storage difference looks like in concrete numbers. Use our CCTV HDD Storage Calculator to model your own setup with different codec choices.

What codecs actually do

A codec (coder-decoder) compresses raw video data before writing it to storage and decompresses it on playback. The compression algorithm determines how aggressively temporal redundancy (the similarity between consecutive frames) and spatial redundancy (repeated pixel patterns within a frame) are removed. A more efficient codec produces a smaller file at the same perceptible quality — or better quality at the same file size. For CCTV, the practical result is fewer and smaller hard drives in your NVR.

H.264 (AVC): the baseline everyone has

H.264 / AVC (Advanced Video Coding) became the standard codec for IP cameras around 2008 and is still supported by virtually every camera and NVR on the market. It's the reference point. For a 2MP (1080p Full HD) camera recording at medium quality and 20-25fps, H.264 typically uses around 4 Mbps. A 4MP camera runs around 8 Mbps. An 8MP (4K) camera sits at 16 Mbps. These numbers translate directly to storage: 4 Mbps = 0.5 MB/s = 43.2 GB/day per camera.

H.265 (HEVC): half the storage, same quality

H.265 / HEVC (High Efficiency Video Coding) was standardised in 2013 and processes larger block sizes, which makes it dramatically better at eliminating spatial redundancy — especially in uniform areas of a frame like walls, ceilings, and sky. The result is roughly 50% bitrate reduction at the same quality level versus H.264. On our calculator's reference bitrate table, the numbers look like this:

• 2MP: H.264 = 4.0 Mbps → H.265 = 2.0 Mbps (50% saving)
• 4MP: H.264 = 8.0 Mbps → H.265 = 4.0 Mbps (50% saving)
• 8MP (4K): H.264 = 16.0 Mbps → H.265 = 8.0 Mbps (50% saving)

For a 16-camera H.265 system at 4MP and 30-day retention, you need roughly half the HDD capacity of the same setup on H.264. Over a typical 3–5 year NVR lifespan, that difference can mean one fewer hard drive bay, or the ability to double retention without buying extra drives.

H.264+ and H.265+: smart codecs from Hikvision and Dahua

H.264+ and H.265+ are proprietary extensions developed by Hikvision and Dahua (sometimes marketed by other brands under different names like "Smart Codec"). They add background modelling on top of the base codec: the encoder identifies static parts of the scene — walls, ceilings, floor areas without movement — and compresses them even more aggressively between keyframes, reserving higher quality only for areas with actual motion. In practice this reduces bitrate by a further 30–50% compared to standard H.265 in typical CCTV scenes (which are mostly static):

• 2MP: H.265 = 2.0 Mbps → H.265+ = ~1.4 Mbps (30% additional saving)
• 4MP: H.265 = 4.0 Mbps → H.265+ = ~2.6 Mbps (35% additional saving)
• 8MP (4K): H.265 = 8.0 Mbps → H.265+ = ~5.5 Mbps (31% additional saving)

The tradeoff: H.265+ requires both the camera and the NVR to support the same proprietary extension. You can't use a Hikvision H.265+ camera with a generic third-party NVR and expect the compression benefit — the NVR will fall back to treating it as standard H.265 at the camera's configured bitrate, but you lose the encoder's flexibility. This is one reason we spec cameras and NVRs together rather than mixing brands arbitrarily.

MJPEG: still used in specific situations

MJPEG stores each frame as a separate JPEG image rather than using inter-frame compression. It uses roughly 3.5× more storage than H.264 at the same quality — a 2MP MJPEG stream at medium quality runs around 14 Mbps. The only reason to use MJPEG today is forensic-grade single-frame extraction, where you need to access any individual frame without decoding a GOP (group of pictures). Some older video analytics systems or legal evidence workflows specify MJPEG for this reason. For almost every commercial CCTV application in Mumbai, H.265 or H.265+ is the right choice and MJPEG should be avoided.

Practical recommendation

• If you're installing new cameras in 2026: use H.265 as the minimum — it's supported by all current-generation NVRs from Hanwha, Honeywell, Matrix, Hikvision, and Dahua
• If you're staying within a single brand ecosystem (e.g., all Hikvision or all Dahua): enable H.265+ and size storage at the H.265+ bitrate, not H.265
• For Axis cameras: use Axis Zipstream (their proprietary equivalent of smart codec) and enter the actual bitrate from the camera's web UI using the manual override in the calculator
• For Hanwha Wisenet cameras: use H.265, and use manual bitrate override from the camera spec sheet for accurate sizing — Hanwha doesn't use a proprietary compression extension
• For retrofits of older H.264 NVR infrastructure: run H.264 on legacy channels, H.265 on new IP camera channels if your NVR supports mixed recording

To see the actual storage difference for your specific camera count, retention requirement, and recording mode — try it in the CCTV HDD Storage Calculator. Add two camera groups with identical settings and switch one to H.264 and one to H.265 to compare directly.