What a codec does
Raw video is enormous. A camera capturing high-resolution frames many times a second would fill a drive and saturate a network almost immediately. A codec compresses the stream by removing information the viewer is unlikely to notice.
Two ideas drive most compression. The first is spatial, which means simplifying areas within a single frame, such as a smooth sky or a plain wall. The second is temporal, which means storing only what changes between frames instead of repeating the whole picture. A security scene, where most of the frame is static most of the time, compresses especially well.
Codecs are lossy. Some detail is discarded permanently, and the more aggressive the settings, the more is lost. That is the central tension in every choice below.
H.264 in brief
H.264, also called AVC, is the long-standing workhorse. Nearly every camera, recorder, player and browser handles it. It decodes easily on modest hardware, which makes it a safe default.
Its weakness is efficiency. At high resolutions, such as 4K or the 8MP cameras now common, an H.264 stream can still consume a lot of storage and bandwidth. That matters when you record many cameras for weeks or when your connection is limited.
H.265 in brief
H.265, also called HEVC, was designed to deliver similar visual quality at a lower bitrate, particularly at higher resolutions. In practice that means smaller files for the same retention period, or longer retention on the same drives.
The costs are on the processing side. Encoding and decoding H.265 takes more computing power, so older recorders, low-end viewing devices and some web browsers may struggle or not support it at all. Licensing and platform support have also been less uniform than for H.264.
Many cameras offer smart or enhanced variants that lower bitrate further by compressing static areas more heavily and giving moving areas more detail. These can help, but they behave differently from camera to camera and should be tested.
Side-by-side comparison
| Factor | H.264 | H.265 |
|---|---|---|
| Efficiency | Good | Generally better at the same quality |
| Storage needs | Higher for the same retention | Lower for the same retention |
| Bandwidth needs | Higher | Lower |
| Processing load | Light | Heavier |
| Compatibility | Very broad | Good but less universal |
| Best fit | Mixed or older equipment | Higher-resolution cameras and constrained links |
No exact percentage applies to every site. Savings depend on the scene, motion, lighting, frame rate and the encoder, so measure with your own cameras before sizing storage.
Why it matters on cellular links
Cellular connections are the clearest case where codec choice pays off. Data on a cellular link is limited, shared and variable, and sites that rely on it cannot stream every camera in full quality around the clock.
A more efficient codec helps in several ways.
- A live view can run at usable quality over a smaller connection.
- Event clips upload faster, so a reviewer sees the incident sooner.
- Data allowances stretch further.
The codec is only one lever. Resolution, frame rate and the choice between continuous streaming and event-based sending matter at least as much. Our article on event-based versus live video monitoring describes how sites balance those. For the connection itself, see cellular security cameras on 4G and 5G.
Storage and retention
On the recording side, a lower bitrate means more days of footage per drive. That makes retention easier to meet without buying more storage. But retention should be set by need first, such as contract, insurance and risk, and then storage sized to match. Our guide on how long security cameras should keep footage walks through that decision.
Keep in mind that retention math changes when cameras record continuously versus only on events, and when different cameras use different settings. Estimate per camera, not as one average.
Quality, resolution and evidence
Compression and resolution interact. A high-resolution camera with a heavily compressed stream can look worse than a lower-resolution camera with lighter compression, because the codec removes the fine detail that resolution was supposed to provide. Our security camera resolution guide explains how pixels, field of view and distance relate.
For plates, faces and other details you may need to prove later, protect quality in the critical zone. Avoid the most aggressive bitrate settings there, and keep original recordings when footage becomes evidence. If footage may be used in legal or insurance matters, requirements vary by state and situation, so confirm with counsel.
Bitrate modes and frame rate
Codec choice is only part of the picture. Cameras let you choose between constant bitrate, which holds data use steady and can starve busy scenes of detail, and variable bitrate, which spends more data when the scene is busy and less when it is quiet. Variable settings usually save the most storage on a mostly static site, but they make data use harder to predict, which matters on a capped cellular plan.
Frame rate is another lever. Lowering it reduces data but can hide quick events, so reserve higher rates for areas where fast movement matters, such as vehicle lanes.
Compatibility checklist
Before switching a site to H.265, check each link in the chain.
- Does the recorder or cloud service accept and store H.265?
- Can the viewing software on your team's computers and phones play it smoothly?
- Do your analytics or third-party integrations accept it? Open protocols such as those in our ONVIF and RTSP explainer often support multiple codecs, but confirm each one.
- Can the camera produce a second, more compatible stream when needed?
Many deployments run H.264 for compatibility on one stream and H.265 for recording or remote links on another. If anything in the chain cannot handle H.265, the savings disappear.
How VDS fits
Remote sites live with limited connections, which makes efficient video handling important. VDS mobile surveillance units run on dual-SIM 4G/5G cellular and carry a 5MP PTZ with 25x optical zoom plus three 8MP wide-angle cameras. AI detection for people, vehicles, loitering, objects and license plates flags what matters, and a 24/7 US-based security operations center verifies events, so people, yours or your security partner's, handle verified incidents without anyone watching raw streams.
To talk through connection and video needs at a particular site, reach out to our team.