What Is Transcoding, How It Works, Recent Trends & Innovations

What is transcoding?

Transcoding is taking encoded (or “compressed”) video or other digital content, decompressing it, and altering and re-compressing it. For example, a high-resolution video shot on a digital camera (HD, 4K, etc.) can be transcoded into a lower-resolution format for editing; in other words, smaller files that are faster and easier to manipulate in editing software. Or video for a live broadcast can be transcoded from its original format into differently formatted streams to be delivered to the largest number of viewers on the widest range of devices. Transcoding is the digital-to-digital conversion of one type of encoded data (video or audio) to another, often because the target device that will display the content requires a smaller file size. Think about watching a feature film on a smartphone, and you’ll get the idea.

Editor’s note: Updated the article to cover recent market trends and product evaluation criteria as of 2026.

Video transcoding market trends

Demand for video transcoding is rising with the streaming market

The need for video transcoding is growing alongside the broader video streaming market. Recent market estimates show the global video streaming market is valued at $811.37 billion and expected to grow to $3.39 trillion by 2034.

That growth creates more pressure on broadcasters, OTT platforms, enterprises, educators, and content creators to deliver video in multiple resolutions, bitrates, formats, and device-compatible versions. In practice, this means transcoding is no longer just a technical backend step; it is a core part of streaming scalability and viewer experience.

Video traffic continues to dominate internet usage

Video remains one of the largest drivers of internet traffic. Video was the largest downstream traffic category across fixed and mobile networks, accounting for 38% of traffic by application category. Video is the dominant traffic type on mobile networks, with video accounting for roughly three-quarters of mobile data traffic. 

This matters because viewers now expect smooth playback across phones, tablets, smart TVs, laptops, and unstable mobile connections. Transcoding supports adaptive bitrate streaming, which allows a platform to serve the right version of a video based on the viewer’s device, bandwidth, and screen size. Without transcoding, platforms would be more likely to experience buffering, playback errors, excessive data usage, or poor-quality streams.

New codecs are increasing the need for smarter transcoding

Codec adoption is another major trend. H.264/AVC remains widely supported, while HEVC/H.265, AV1, and emerging codecs are becoming more important for 4K, HDR, mobile, and bandwidth-efficient streaming. Netflix reported that AV1 powered about 30% of its streaming, showing how major platforms are using newer codecs to reduce bandwidth needs and improve video quality.

However, newer codecs also add complexity. Not every device supports every codec, and some formats require more computing power to encode or decode. This is where transcoding becomes critical: it allows platforms to maintain compatibility with older devices while also taking advantage of newer compression technologies for supported devices.

How does transcoding work?

Transcoding takes your video media (or possibly audio media), decodes it into an intermediate uncompressed format, and then re-encodes the content into its’ target format. As we also laid out in the last section, the process is likely to additionally include transrating and transsizing/image scaling.

 A transcoding solution could take a video file that’s already been compressed and encoded (we’ll call it “ExampleVideo.mov”) and reformat it into an MP4 file using the H.264 codec that would be more suitable for online streaming (let’s call that transcoded version of the video “ExampleVideo.mp4”).

 Depending on the video transcoder you decide to use, it could be anything from open-source software with a command-line interface to something with much more robust features and user interface. Transcoding can potentially happen via software on any PC or laptop, a dedicated media server, or a SaaS platform. One thing to keep in mind is that transcoding is, as they say, “computationally intensive.” In short, it benefits from substantial hardware and system resources, like generous amounts of system RAM, graphics acceleration, and higher-end CPUs.

In other words, don’t assume you’ll be able to easily transcode 4K video to a quality HLS or DASH stream with a Chromebook: typical ingest and transcoding of HD video for editing in traditional video production (corporate video, television, etc.), for example, can be known to tie up beefy desktop machines—sometimes for hours at a time.

Video transcoding vs encoding vs transmuxing

Encoding/compression is an important related concept. Compression is essential in preparing video for streaming; in earlier eras of video when audio and footage might come in on analog tape formats or even film, media would need to be digitized and compressed to be compatible with computer applications, like web pages or video editing workstations. This isn’t much of a concern now as today’s all-digital cameras can be set to internally compress footage into more manageable codecs like H.264. (Note: the word “footage” used to literally refer to a number of feet of film or tape!)

 However, digital cameras still may capture uncompressed RAW files, which are very large since they are data directly from the camera sensors with no loss of quality or alteration. This is often desirable to provide full detail that can be edited to a producer or content creator’s preference, but, due to the large file size, they will have to be compressed to a more manageable size for playback. Relating to transcoding: video pulled in by a digital camera may be immediately compressed by the camera or via a linked encoder, however as mentioned above, it still won’t be suitable for delivery to a wide online audience without transcoding!

 Transmuxing/rewrapping is when content that is compressed is repackaged into a different delivery format, but without making any changes (including any further compression) to that packaged content. This is less of an intensive process than transcoding, and also a fairly common procedure in content delivery. This only changes the way video and audio data packets are organized. For instance, you may have an H.264 video clip captured from a camera on your drive, and by transmuxing, or rewrapping its container, it can be made suitable for delivery over the web via HLS (which breaks the video into small MPEG-2-TS files, referred to as chunks, of varied bitrate—but notably won’t make changes to the base clip).

 Neither of these digital media tasks can properly be called transcoding. However, because they are related, they’re sometimes confused.

“Transcoding” as an umbrella term

Essentially, transcoding is a two-step process in which (encoded) data is decoded to an intermediate format and then encoded into a target format.

Three tasks might fall under the larger umbrella when someone refers to transcoding video content: 

• ”Standard” transcoding

This is meant in the most general sense of transcoding a video or stream, making changes to the video/audio itself. For instance, if you were streaming a digital conference to the web, you might be working with IP cameras in your conference space. The IP cameras most likely function on the RTSP protocol and will not create a video stream suitable for playback over the web, so transcoding software or service will be necessary to convert your content into an adaptive bitrate stream. The “transcoding” process might also include the two types of changes below.

• Transrating

Transrating is a more specific type of transcoding that is intended specifically to change the bitrate. So it’s the same video content, video format, and codec, and the alteration is the bitrate: you might want to bring an 8Mbps bitrate down to 3Mbps, making it possible for the media to fit into less storage space or be broadcast over a lower bandwidth connection.

• Transsizing

This is another specific type of transcoding that is used to resize a video frame (this may also be called “image scaling”), for example, bringing down 4K resolution to 1080p.

Why transcoding is essential for successful streaming

As you now know, transcoding is a key part of an adaptive streaming workflow and a step in preparing your content for a delivery protocol (such as current industry-standard HLS) that can reach the widest possible number of display devices.

 Media is constantly changing and evolving, with new devices, applications, and input sources being introduced regularly. While new cameras, new feeds, and innovations in webcasting improve the quality and variety of content being produced, they also inevitably create new challenges in content delivery. And if you want your viewers to always get the best quality viewing experience, adaptive streaming (at least at the time of this writing!) is the way to go. Transcoding is the essential intermediate step in moving your best content from the capture device to a high-quality streaming output such as HLS or DASH. These adaptive streaming formats are the best for reducing buffering and playback issues, delivering the sharpest possible picture continuously and with no interruptions.

Examples of transcoding

We pointed out above that transcoding is used heavily in traditional film and video production. Typical needs might be to “down-res” files captured with a camera to lower resolution, smaller size “proxy” files which have lossy video quality but can be quickly edited; or to transcode high quality finished video from an editing format like Apple ProRes to a delivery format like H.264. This is usually a process of “local transcoding” relying on video editing and compression software (DaVinci Resolve, Avid Media Composer, Adobe Media Encoder, etc.), which will transcode files on the user’s computer. The main downsides are possible limitations due to hardware, as we mentioned earlier, and the need to manage multiple files yourself.

 Let’s take a look at how transcoding might work in a streaming workflow. In this case, you would have a video asset (clip, livestream, etc.) from a capture device (video camera, IP camera, drone, or another device). The incoming video and audio data would likely be encoded in a file or streaming format by the camera and then sent off to an encoder … you might do this on a desktop, with a dedicated hardware encoder, or by sending it to a cloud server on the web to transcode. The encoder will then do the transcode, making sure that whatever the original data format is (as mentioned, some cameras shoot RAW, or .raw, files, other video cameras might be outputting files with the extension .mov or .mpeg), it will be recompressed into a format suitable for internet streaming (most likely H.264).  The encoder would also create multiple stream renditions that transrate/transsize to different bitrates and resolutions. The data is then sent to a media server (possibly, but not necessarily, the same server where the transcoder is located) to be packaged into an adaptive streaming format (like HLS) and served via HTTP over the “last mile” to viewers. As described, this will create multiple options for content delivery, from HD output on a smart TV or desktop to a suitable size for a tablet or smartphone screen.

Best practices for video transcoding

Choose the right codec and container

Codec selection directly affects compression efficiency, compatibility, and processing cost. H.264 is still the safest default because it works across almost all devices and browsers. However, newer codecs like H.265 (HEVC) and AV1 can reduce bandwidth by 30–50%, which matters at scale. The trade-off is higher encoding complexity and, in some cases, limited playback support.

Containers should align with how the video will be delivered. For example, MP4 works well for progressive playback and VOD, while MPEG-TS or CMAF is better suited for adaptive streaming protocols like HLS and DASH. Choosing the wrong combination can lead to playback failures or unnecessary processing steps later in the pipeline.

Use adaptive bitrate ladders

Adaptive bitrate (ABR) ladders are essential for delivering video across varying network conditions. Instead of a single output, you generate multiple renditions at different resolutions and bitrates. Players then switch between these renditions in real time based on available bandwidth and device capability.

A good ladder avoids redundancy. For example, having multiple renditions with similar bitrates but different resolutions can waste storage and processing. Each step should provide a meaningful quality improvement. Many workflows now use dynamic or per-title ladders, where encoding parameters are adjusted based on the complexity of each video rather than using a fixed template.

Optimize resolution and bitrate together

Resolution and bitrate are tightly coupled. Increasing resolution without increasing bitrate leads to visible artifacts, while using too much bitrate for a low resolution wastes bandwidth. The goal is to find the minimum bitrate that maintains acceptable visual quality for each resolution.

Content type matters here. Fast-moving sports footage requires higher bitrates than static content like presentations or interviews. Modern workflows often use content-aware encoding to automatically determine optimal bitrate-resolution pairs, improving efficiency without manual tuning.

Apply two-pass or content-aware encoding

Two-pass encoding analyzes the video in a first pass and allocates bits more effectively in the second pass. This improves overall quality, especially in scenes with varying complexity. It is commonly used in VOD workflows where encoding time is less critical.

Content-aware encoding goes further by analyzing scenes or frames and adjusting encoding parameters dynamically. For example, it may allocate more bits to high-motion segments and fewer to static ones. This reduces file size while maintaining consistent perceived quality, which is especially valuable for large-scale libraries.

Leverage hardware acceleration where appropriate

Hardware encoders significantly increase throughput and reduce encoding time. GPUs and dedicated ASICs can handle multiple streams in parallel, making them suitable for live streaming and high-volume processing environments.

However, hardware encoding may not always match the quality of software-based encoders at the same bitrate. The gap has narrowed, but it still exists in some scenarios. A common approach is to use hardware acceleration for live or time-sensitive workloads and software encoding for high-quality VOD outputs.

Set keyframe intervals and GOP structure correctly

Keyframes (I-frames) are reference points in a video stream. Their placement affects seekability, compression efficiency, and adaptive streaming behavior. Shorter keyframe intervals improve switching and reduce latency but increase file size.

Most streaming workflows use a fixed GOP (Group of Pictures) structure with keyframes every 2–4 seconds. This aligns with segment boundaries in HLS or DASH. Misaligned keyframes can cause playback issues, inefficient bitrate switching, or increased buffering.

Normalize audio and handle multiple tracks

Audio quality is often overlooked but plays a major role in user experience. Loudness should be normalized to standards such as EBU R128 or ATSC A/85 to avoid sudden volume changes between videos or segments.

Support for multiple audio tracks is also important, especially for global audiences. This includes different languages, commentary tracks, or surround sound formats. Efficient audio encoding ensures good quality without significantly increasing overall file size.

Validate output across devices and networks

Transcoded outputs must be tested in real-world conditions. This includes different devices (smartphones, TVs, desktops), operating systems, and browsers. Each environment may have unique decoding capabilities or limitations.

Network testing is equally important. Simulate low bandwidth, high latency, and packet loss scenarios to evaluate playback behavior. Metrics such as startup time, buffering frequency, and bitrate switching stability provide insight into the actual user experience.

Automate workflows and monitor quality

Manual transcoding does not scale. Automated pipelines handle ingest, transcoding, packaging, and delivery with minimal intervention. This reduces errors and ensures consistency across large volumes of content.

Quality control should be built into the workflow. Automated checks can detect issues like missing audio, incorrect resolution, or sync problems. Monitoring tools and metrics such as VMAF, PSNR, and rebuffer rate help maintain consistent output quality over time. 

Conclusion

Hopefully, this has shed some light on the questions of what transcoding is and what makes transcoding essential for video streaming, especially for those just getting started in the streaming field! Keep an eye on the Kaltura blog for more building blocks on your way to understanding video streaming and workflows.