HEVC vs H.264 explained for security and IT leaders: storage and bandwidth savings, device and browser support, licensing, where AV1 fits, and what to choose.
A codec is the method a camera uses to shrink video so it can travel over a network and fit on a disk. The choice sounds technical, but it has very practical effects: how much storage you buy, how much bandwidth each site needs and whether every screen in the building can play the video.
As 4K cameras become normal and retention periods grow, codec choice has become a real budget decision. H.264 is still the universal baseline, but it is showing its age. HEVC, also called H.265, offers a large saving in exchange for some compatibility trade offs.
According to the Bitmovin annual Video Developer Report, HEVC adoption climbed from roughly 18% of respondents in 2018 to over 50% by 2024, while AV1 rose from under 5% to more than 15%. H.264 is still used by 83% of respondents. It is not going anywhere, but it is no longer the only default.
H.264, also called Advanced Video Coding (AVC), is the codec that made modern online and surveillance video possible. It was standardised in 2003 by the ITU-T and ISO/IEC, and it struck a practical balance between good compression and easy playback.
Its strength is that everything plays it. Phones, laptops, televisions, browsers and video walls all decode H.264 with dedicated hardware. According to MPEG LA public licensing data, more than 1,100 patent holders joined the AVC pool after 2004, making it one of the most widely licensed technology standards ever.
Its weakness shows at high resolutions. At 4K and above, H.264 needs a lot of data to keep the picture clean, which is where HEVC steps in.
Pros: plays on practically every device and every modern browser, including Firefox; simple, well understood licensing; quick to compress, even in software; a mature set of tools and hardware support.
Cons: uses considerably more data than HEVC for the same quality; not designed for 8K; shows blocky artefacts sooner when bandwidth is tight.
HEVC (High Efficiency Video Coding) is the successor to H.264. It was standardised in 2013 as ITU-T H.265 by the Joint Collaborative Team on Video Coding (JCT-VC). Its headline achievement is the same visual quality at roughly half the bitrate.
Iain Richardson, Professor at Robert Gordon University and author of "The H.264 Advanced Video Compression Standard" (Wiley), has written that each new codec generation typically cuts bitrate by 40 to 50% in exchange for roughly ten times more compression work. The industry has accepted that bargain for more than two decades.
At surveillance scale, that saving compounds across every camera and every day of retention. The Moscow State University Graphics and Media Lab codec comparison reports have consistently found reductions in the 40 to 50% range on real content.
Pros: roughly half the storage and bandwidth of H.264 at the same quality; built for 4K and 8K; cleaner pictures when bandwidth is tight; supported by most modern cameras, computers and phones; HDR and wide colour support built in.
Cons: needs more computing power to compress; licensing split across several patent pools; no native playback in Firefox; some older cameras and recorders do not support it.
The core difference is simple. For the same picture quality, HEVC needs roughly half the data of H.264. That means roughly half the disk space for the same retention, or twice the retention on the same disks.
The gain grows with resolution. It is smaller at 1080p and largest at 4K and above, which is exactly where modern surveillance estates are heading.
When bandwidth is limited, HEVC keeps textures, faces and fine detail clean at levels where H.264 starts to show blocks and banding. Research published on the Netflix Technology Blog, through its VMAF quality measure, has consistently shown HEVC scoring higher than H.264 at low bitrates.
Anne Aaron, former Director of Video Algorithms at Netflix, has described VMAF as a measure built to reflect what viewers actually see rather than raw pixel maths. Netflix reports VMAF differences of 5 to 15 points between HEVC and H.264 at the same low bitrate, most visible in dark and fast moving scenes. For security footage shot at night, that matters.
HEVC asks more of the hardware. Benchmarks from the open source x264 and x265 projects put software HEVC compression at roughly 5 to 10 times the processing cost of H.264 at comparable quality.
In surveillance this matters less than it sounds. Cameras compress video with dedicated chips, so the camera does the heavy work once. The video platform mostly stores and forwards what the camera sends.
Almost every device made in the last decade and more decodes H.264 in hardware. HEVC hardware support arrived from about 2015 across graphics chips, phone processors and computers, and is now widespread.
Older screens and workstations may lack HEVC hardware, which forces slower software playback. Check your control room machines before standardising on HEVC everywhere.
H.264 plays in Chrome, Firefox, Safari and Edge with no extra software. HEVC support is patchier: Safari on Apple devices, Edge on Windows with a hardware decoder, Chrome where the operating system allows it, and not natively in Firefox.
For a browser based video platform, the answer is to let the platform handle it. It should pass HEVC straight through where the browser can play it and convert it only where the browser cannot.
H.264 licensing is relatively simple, run mainly through one pool. HEVC licensing is split across several pools, Via LA (formerly MPEG LA), Access Advance and Velos Media, plus independent patent holders.
The Alliance for Open Media cited this complexity as a founding reason for creating the royalty free AV1. For most surveillance buyers, though, licensing is already covered in the cameras and software they buy.
Choose HEVC when you record 4K or higher; when bandwidth is limited, such as cellular, satellite or remote site links; when long retention makes storage your biggest cost; or when you want cleaner night footage at a given bitrate.
Choose H.264 when you need every device and browser to play the video without fuss; when cameras or screens are older; or when 1080p and below is enough for the job.
Many sites use both: HEVC on the main stream for recording, and a lighter sub stream for small live tiles.
AV1 was finalised in 2018 by the Alliance for Open Media, a group including Google, Netflix, Amazon, Meta, Microsoft, Apple, NVIDIA, Intel and Arm. It is royalty free and compresses even more efficiently than HEVC.
Matt Frost, then Head of Strategy and Partnerships for Chrome Media at Google, positioned AV1 as the next codec for the open web. Netflix has published analyses of its AV1 rollout reporting bitrate savings of up to 50% against its H.264 baseline. YouTube announced in 2018 that AV1 would become its default for popular content, and Meta has reported quality improvements of 20% or more at equal bitrates across Facebook and Instagram video.
For streaming films, AV1 is the likely long term successor. For security cameras, the picture is different: cameras record in H.264 and H.265, and those two formats will carry surveillance for years to come.
Strengths: the most efficient of the three codecs; royalty free; supported by major browsers; growing hardware support in new computers and phones.
Weaknesses: very demanding to compress in software; live compression at high quality needs dedicated hardware; rarely offered by security cameras today.
In surveillance, codec choice carries real weight. HEVC cuts bandwidth and storage across large camera estates without losing image clarity.
Industry surveys by IHS Markit and Omdia have reported that over 75% of IP cameras shipped in 2024 support HEVC natively, up from under 40% in 2020. Most new cameras can record in HEVC today.
The standard pattern is two streams per camera: a full quality main stream, often in HEVC, for recording, and a lighter sub stream for small live tiles. The video platform picks the right stream for each screen.
Storage is usually the biggest running cost of a surveillance estate, and it grows with every camera and every extra day of retention. Halving the data per camera roughly halves that cost, or doubles retention for the same spend.
According to Seagate surveillance storage whitepapers, the average enterprise video surveillance project now keeps 60 to 90 days of footage, against 30 days a decade ago. That longer retention would be far harder to afford without HEVC.
For a central control room pulling live video from many remote sites, halving the data per camera halves the load on every link. Over MPLS, SD-WAN or cellular backhaul, that becomes a direct saving on monthly network bills.
It also makes remote sites practical that would otherwise struggle to send high resolution video at all.
Visylix is an automated AI video surveillance platform from Aptibit Technologies. Its native streaming engine handles H.264 and H.265 (HEVC) cameras side by side, so you can mix old and new cameras in one estate.
Video passes through untouched wherever the viewer can play it. Where a browser cannot decode H.265, Visylix converts it to H.264 for live view, so every operator sees every camera. Each tile picks the camera's main or sub stream to suit its size and the network, and recording always keeps the full quality stream.
Live view is sub second over WebRTC, down to under 200 ms, across 13+ protocols. The 22 AI analytics built in house, from intrusion and weapon detection to number plate recognition and PPE, work on H.264 and H.265 cameras alike.
Everything runs on your own hardware: on premise, at the edge, in a hybrid mix, fully air gapped or in your own cloud account, shipped as a Docker image. Starter and Pro include unlimited cameras on the VMS licence from ₹4,999 or $49 per month, with AI added per camera per month on the cameras you choose.
HEVC and H.264 each have a clear job. H.264 gives universal playback. HEVC gives roughly half the storage and bandwidth, which matters most at 4K and on long retention.
For enterprise surveillance the practical answer is both: HEVC for recording to cut storage, and a platform that makes sure every operator can watch every camera live, in any browser. AV1 is a strong future format for streaming, but security cameras record in H.264 and H.265 today.
Visylix handles both natively, with sub second live view, 22 AI analytics and complete deployment on your own hardware.
For storage and bandwidth, yes. HEVC was designed to deliver the same picture quality as H.264 at roughly half the data. That makes it the better choice for high resolution cameras, long retention and sites with limited bandwidth. H.264 still wins on universal playback, since every device and browser supports it.
Yes. HEVC (High Efficiency Video Coding) and H.265 are two names for the same standard. H.265 is the ITU-T name and HEVC is the ISO/IEC MPEG name. Both describe the specification finalised in 2013.
Not everywhere. Safari plays HEVC on Apple devices, Edge plays it on Windows with a hardware decoder, and Chrome plays it where the operating system provides support. Firefox still lacks native HEVC playback. A good video platform handles this for you, so operators see live video whatever browser they use.
No. HEVC patents are licensed through several separate pools, including Via LA (formerly MPEG LA) and Access Advance, plus companies outside any pool. H.264 is also patented, but its licensing is simpler. For most surveillance buyers, licensing is handled inside the cameras and software they buy.
H.265 squeezes video into roughly half the space of H.264 at the same quality, but needs much more computing power to compress. It works on larger blocks of the picture and has more ways to predict detail, which is why its gains grow at 4K. H.264 remains the format every device can play.
HEVC needs noticeably more processing than H.264, especially to compress video in software. AV1 needs more still. Modern cameras and computers include dedicated chips for these formats, which closes most of the gap.
If your cameras and video platform support it, yes. The storage and bandwidth savings are substantial, especially at 4K. A common setup records the full quality main stream in HEVC while a lighter sub stream feeds smaller live tiles. Visylix handles both H.264 and H.265 cameras natively.
For streaming films and online video, it is already gaining ground, because AV1 is royalty free and compresses very efficiently. For security cameras, HEVC and H.264 remain the formats cameras record in, and they will stay dominant in surveillance for years.
macOS and iOS play HEVC natively. Windows 11 plays it out of the box, while Windows 10 may need the HEVC Video Extensions from the Microsoft Store. VLC plays HEVC on most systems without extra installs.
Visylix supports H.264 and H.265 (HEVC) cameras natively. Video passes through untouched wherever the viewer can play it, and Visylix converts H.265 to H.264 for live view only where a browser cannot decode H.265. Live view is sub second over WebRTC, and everything runs on your own hardware.