Field note · 6 min
H.264 vs H.265 storage: plan the saving, do not assume it
H.265 can deliver similar visual quality at a lower bitrate than H.264, but “half the storage” is not a law. The realized saving depends on both encoders, the quality target, the scene, and whether every component handles the chosen profile correctly.
The codec does not write terabytes by itself
A codec defines tools for compressing video; an encoder decides how to use them. A newer codec has more efficient tools, but encoder quality, implementation limits, latency settings, and target quality influence the result. Comparing an aggressively compressed H.265 stream with a high-quality H.264 stream is not an apples-to-apples capacity comparison.
For planning, the output bitrate is what the network and storage see. If a 4 Mb/s H.264 profile is replaced by a verified 2.6 Mb/s H.265 profile with acceptable image quality, use 2.6—not a generic percentage—in the worksheet.
Compatibility can erase a paper saving
The camera, NVR/VMS, mobile client, browser path, export player, analytics pipeline, and any cloud relay must support the codec/profile in the required workflow. Unsupported analytics or transcoding can add compute cost, reduce viewing quality, or force another stream.
A capacity decision should therefore include a short interoperability test: live view, multiview, playback, export, incident sharing, analytics, and firmware behavior. Storage saved on paper is not valuable if an operational workflow breaks.
Test the difficult scenes
Codec savings can vary with scene complexity. Test daytime and night, motion and quiet, weather, digital noise, and critical target movement. Hold resolution, frame rate, and the usable image objective as constant as practical.
Record the average and high-percentile bitrate over a representative window. Use the average for storage, a conservative traffic case for links and recorder limits, and the same retention/reserve method for both codecs.
A practical decision rule
Choose H.265 when the end-to-end workflow supports it and a controlled comparison shows worthwhile saving without losing required detail or reliability. Keep H.264 when compatibility or processing constraints dominate. In mixed estates, group cameras by actual profile rather than forcing one estate-wide percentage.
Revisit the comparison when firmware, camera model, or image settings change. Codec is one variable in a system, not a permanent storage multiplier.
Sources and further reading
The article distinguishes arithmetic from professional design judgment. Primary and technical sources:
ONVIF Profile T overview ↗
H.265 support and interoperability context.
ITU-T H.265 recommendation ↗
Primary codec standard reference.