H.265 HDMI Encoder Guide

Where HEVC earns its keep on an HDMI encoder, where it costs you, and how to check what your encoder is really sending

An H.265 HDMI encoder takes the same HDMI signal as any other encoder. HEVC changes only what leaves the network port: the same picture carried in fewer bits. That is a real saving on an uplink you pay for, and a real problem at the far end if the player cannot decode it.

This guide covers what H.265 buys at each resolution, when it costs you compatibility and delay, how to decide per destination rather than per preference, and which ORIVISION models encode H.265 today.

1. The gain is in the encode, not the input

The HDMI connector, the cable and the source resolution are unchanged. Everything that matters happens after the capture stage.

Same input, different output

The encoder captures HDMI, decodes it to a frame buffer and encodes it again. Selecting H.265 changes the third step only. Nothing upstream has to change, and no different cable or source device is needed.

It is a compute decision, not a specification on paper

H.265 uses larger coding units and a much wider set of prediction options than H.264, so it takes more silicon to encode in real time. That is precisely why low-cost single-chip encoders often stay on H.264 while models built for efficiency support both.

On most ORIVISION models it is a setting, not a purchase

The models listed further down encode H.265, H.264 and in most cases MJPEG. You select the format in the device configuration, which means a change of destination does not force a change of hardware.

2. What H.265 buys you

The HEVC standard was designed around roughly half the bitrate of H.264 at comparable subjective quality. Treat 30 to 50 percent as the realistic planning range and measure on your own footage.

Why there is no single honest percentage

The saving depends on content. A mostly static presentation benefits far more than high-motion sport with fine grain. Any vendor quoting one fixed number for all content is quoting a laboratory sequence, not your signal.

The saving shows up as the bitrate you settle on

Because the encoder exposes the full range, the efficiency gain is realised when you choose a lower bitrate for the same accepted quality. Published configurable ranges on current models run from 16 kbps up to 12, 16 or 20 Mbps depending on model, which comfortably covers 1080p and 4K contribution.

Arithmetic for planning, using round numbers. This is a calculation, not a measurement of any specific encoder.

Channels Per-channel bitrate Total video What that implies
1 8 Mbps 8 Mbps Comfortable on almost any uplink
4 8 Mbps 32 Mbps Fine on fibre or a business line; marginal on a domestic uplink
8 8 Mbps 64 Mbps Needs a dedicated or well-provisioned link
16 8 Mbps 128 Mbps Dedicated link; per-channel bitrate becomes the real lever
16 4 Mbps 64 Mbps Halving per-channel bitrate is where H.265 pays for itself

Then add headroom

If you carry the stream over SRT, leave spare capacity above the video bitrate for retransmission. See the SRT encoder guide for how that headroom is budgeted.

3. The compatibility catch

The hard part of H.265 is not producing it. It is the far end decoding it.

Playback support is uneven

Older phones, older set-top boxes and some conferencing endpoints do not decode HEVC at all. H.265 in browsers depends on the platform and on hardware decode support, so HLS carrying HEVC is not evenly available across devices.

Failure looks like nothing at all

An undecodable stream usually does not raise an error; it shows a black screen or an indefinite spinner. Test the complete chain with the actual player and the actual device before a live event, not on the day.

H.264 remains the safe last mile

If the destination is a general web audience, H.264 is still the option that plays everywhere. Keep H.265 for the parts of the chain you control: contribution links, recording and archive.

The fallback usually costs nothing

Because the encoding format is selectable on the device, switching back to H.264 is a configuration change rather than a new purchase. Rehearse the H.264 setting too, so that a fallback at the venue is a one-minute operation rather than a discovery.

For a deeper comparison of the two codecs, including where H.264 is still the right call, see H.264 vs H.265 in our engineering blog.

4. What HEVC adds in delay

H.265 is not a low-latency choice by nature, but on hardware encoders the penalty is smaller than the settings around it.

More compute means deeper buffering

HEVC encodes in larger blocks with more prediction options, so the pipeline holds more frames before it can emit them. On a dedicated hardware encoder this is measured in milliseconds, not seconds, but it is not zero.

GOP length and B-frames matter more

If you are chasing the lowest possible delay, the interval between keyframes and whether the encoder uses B-frames influence end-to-end delay more than the codec choice does. Fix those first.

Predictable beats minimal

For live production, a steady 300 ms is far easier to work with than a delay that swings between 100 and 600 ms. Constrain the encoder rather than pushing it to its limit, and measure the result at the player.

Ask for the number on your content

End-to-end delay depends on encoder settings, network conditions and player buffering. Any millisecond figure quoted without those three is a laboratory number, not a specification for your chain.

5. Deciding per destination, not per preference

The right codec is the one the far end can decode, at a bitrate the link can carry.

Decide by where the stream is going

  • Archive and internal distribution: H.265, because storage and bandwidth are recurring costs and you control the players.
  • Platform ingest (YouTube, Facebook, Twitch): check what the platform accepts and prefers for your resolution; RTMP/RTMPS ingest is the practical constraint here.
  • Web playback to a general audience: H.264 for the last mile, whatever the codec used on the contribution leg.
  • Surveillance and VMS integration: confirm HEVC decode support in the recording and management platform before committing, particularly for ONVIF and GB28181 integrations.
  • Bandwidth-limited uplink (4G, 5G, satellite): H.265, where the bitrate saving translates directly into a stable link.

6. Current H.265 HDMI encoder line-up

All models below encode H.265. Figures are quoted from each product page; where a page does not state a value it is left out rather than estimated.

Model Channels Max resolution Bitrate range Notes
ZY-EH1301 1 HDMI with loop-out 4K@30Hz / 1080p@60Hz 16 kbps – 20 Mbps ONVIF and GB28181; PoE optional
ZY-EH1401 1 HDMI with loop-out 4K@30Hz / 1080p@60Hz 16 kbit/s – 16 Mbit/s LCD shows IP address and status
ZY-EH1211 1 HDMI 1080p@30Hz 16 kbit/s – 16 Mbit/s Compact 125 × 72 × 25 mm
ZY-EH1304 4 HDMI 1.4 plus audio Up to 3840 × 2160 16 kbps – 12 Mbps Four streams per channel, 16 in total; HDCP 1.4
ZY-EHP1414 4 HDMI with 4 loop-out 4K@30Hz / 1080p@60Hz 16 kbit/s – 16 Mbit/s 1U rack chassis, per-channel display
ZY-EH1308 8 HDMI plus 8 audio 4K@30Hz / 1080p@60Hz 16 kbit/s – 16 Mbit/s Two RJ45 ports
ZY-EHV1401 1 HDMI and 1 VGA 1080p@60Hz 16 kbit/s – 20 Mbit/s Dual-interface input on one device
ZY-ES1401 1 SDI with loop-out 1080p@60Hz 16 kbps – 20 Mbps For SDI camera sources

FAQ

At comparable subjective quality, usually yes, and the HEVC standard was designed around roughly half the bitrate. The real saving varies with content and with encoder implementation. On low-motion content the gap is large; on high-motion, noisy footage it narrows considerably.
If you record the encoded stream rather than the uncompressed source, yes, and that saving repeats every day. If you record in a different format, the codec used for transmission has no effect on storage at all.
The encoding format is selected per device on the models listed here. We do not quote simultaneous dual-codec output unless it is stated on the product page. Tell us the mix of destinations you need and we will confirm on the specific model rather than generalise.
No. HDMI capture and the source device are unchanged. Only the video that leaves the network port is different, which is why H.265 is normally a configuration change rather than a hardware change.
Assume decode support first, not your encoder. Test the same stream in a known-good software decoder to separate a decode problem from a network problem, then compare that against the target device. The encoder setting is usually the last thing to suspect, not the first.

Related guides

Live Streaming Encoder Guide

From a single camera to a 16-channel rack: protocols, bitrate and interfaces explained before you choose a model

8 FAQs

SRT Encoder Guide

What SRT actually fixes, the one setting most installations get wrong, and which models ship with it

6 FAQs

4, 8 and 16-Channel HDMI Encoder Guide

Choose the channel count by counting streams, then size the network before you choose the chassis

5 FAQs

Need H.265 and H.264 side by side in one project?

Send us the source resolution, the destination platform and the uplink you have available, and our engineers will recommend the codec, bitrate and model rather than leaving you to guess at the trade-off.

HDMI video encoder models — compare channels, inputs and protocols across the range.

Talk to an engineer