Short-Form Video Codecs Guide: H.264 (AVC) vs. HEVC (H.265) vs. AV1

1. The Fundamental Compression Problem in Short-Form Video

Every single second of uncompressed 1080p60 video in 8-bit RGB color requires approximately 3 gigabits of bandwidth. If platforms like TikTok and Douyin streamed uncompressed video, a standard 60-second clip would consume over 20 gigabytes of data—exhausting monthly cellular limits in seconds and collapsing global transit networks.

To solve this, short-form video networks depend on lossy inter-frame video compression algorithms. These algorithms eliminate temporal redundancy (differences between sequential frames) and spatial redundancy (predictable patterns within individual frames) through motion vector estimation and discrete cosine transforms (DCT).

However, the specific codec chosen to compress the stream directly determines three critical factors:

1. Compression Efficiency: How much visual fidelity is preserved at a given bitrate.
2. Hardware Decoding Support: Whether older phones, laptops, and smart TVs can play the video smoothly without overheating or lagging.
3. Editing Software Interoperability: Whether software like Adobe Premiere, DaVinci Resolve, Final Cut Pro, or CapCut can import the MP4 file natively.

2. Codec Comparison: AVC vs. HEVC vs. AV1

3. Why Downloaded Videos Sometimes Show a Black Screen

A frequent issue reported by creators downloading Douyin videos without watermark is that the downloaded MP4 file plays sound properly, but displays a black screen in Windows Media Player, QuickTime, or older video editors.

This is almost always caused by an HEVC codec mismatch:

• Windows 10/11 Without HEVC Video Extensions: Microsoft removed native HEVC hardware decoding from default Windows installations to avoid patent licensing costs. When an uncompressed Douyin video encoded in H.265/HEVC is opened in Windows, the default player can decode the AAC audio track but lacks the decoder DLL to unpack the HEVC video frames.
• Legacy Non-Studio Video Editors: Older versions of editing suites (such as Premiere Pro CS6 or older non-studio DaVinci Resolve) lack native HEVC decoders, resulting in an "unsupported codec" prompt or offline media alert.

4. The Anatomy of an MP4 Container: The Moov Atom

MP4 is not a video codec; it is an ISO Base Media File Format (ISO/IEC 14496-12) container. An MP4 file consists of discrete structured chunks known as "atoms" or "boxes":

• mdat (Media Data Atom): Contains the actual raw compressed video and audio packet samples.
• moov (Movie Atom): Contains the metadata index, frame offsets, timestamps, and duration pointers necessary to decode the mdat content.

Fast-Start Streaming vs. Local File Layout

In traditional file writing, the moov atom is placed at the end of the MP4 file after all video frames have been encoded. However, in web streaming, this creates a critical problem: a browser cannot begin playing the video until it downloads the entire file to read the moov atom at the tail end!

the platform's edge transcoders run a qt-faststart optimization pass that shifts the moov atom to the very beginning of the file (before the mdat chunk). This enables progressive pseudo-streaming: the browser reads the first few kilobytes, acquires the timing index, and begins video playback instantly while the rest of the media is buffered over the network.

5. Chroma Subsampling and Color Space Considerations

Nearly all short-form videos are encoded in YUV 4:2:0 chroma subsampling at 8-bit depth using the ITU-R Recommendation BT.709 color matrix.

In YUV 4:2:0:

• The luma channel (Y - brightness) is captured at full native resolution (1080 × 1920).
• The chroma channels (U/V - blue and red color differences) are downsampled by 50% in both horizontal and vertical axes (540 × 960).

Because human vision is significantly more sensitive to variations in luminance than chrominance, this subsampling cuts color data by 75% with virtually zero perceptible loss in everyday mobile viewing.

6. Transcoding Recommendations for Editors

If you are downloading clean video clips for fair-use commentary, documentary research, or educational analysis, consider these transcoding best practices: