Versatile Video Coding (VVC / H.266)

As digital media consumption escalates toward ultra-high-definition 4K, 8K, and immersive 360-degree virtual reality, bandwidth demands are pushing existing compression standards to their limits. Versatile Video Coding (VVC), also formally designated as ITU-T H.266 and ISO/IEC 23090-3 (MPEG-I Part 3), represents the latest evolutionary leap in video compression technology. Developed jointly by the Joint Video Experts Team (JVET)—a collaboration between ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11—VVC was finalized in July 2020 with a single, clear objective: reduce data bitrate by approximately 50% relative to its predecessor, HEVC (H.265), while maintaining identical visual quality.

1. What is VVC (Versatile Video Coding)?

Versatile Video Coding is a block-based, hybrid video codec designed to address the full spectrum of video applications. Unlike older codecs optimized primarily for conventional broadcast and streaming, VVC was architected from the ground up to accommodate:

• High-frame-rate (HFR) video (up to 120 fps and beyond)
• High Dynamic Range (HDR) and wide color gamut (BT.2020)
• Screen content coding (computer-generated text, graphics, UI rendering)
• Omnidirectional $360^\circ$ video for AR/VR environments
• Adaptive resolution switching without stalling video playback
• Low-latency real-time interactive communications (gaming, video conferencing)

By achieving unprecedented compression ratios, VVC enables broadcasters, streaming providers, and mobile networks to deliver premium visual experiences over constrained network connections.

2. Key Characteristics of VVC / H.266

Core Architectural Advantages:

1. Bandwidth Economy: Delivers pristine 4K video at bitrates traditionally required for 1080p Full HD.
2. True Multipurpose Flexibility: Native profiles dedicated to gaming, screen sharing, immersive spherical video, and surveillance.
3. Optimized for Cloud & CDN Delivery: Drastically cuts distribution and storage overhead for major video streaming platforms.

3. In-Depth Technical Features of VVC

VVC incorporates sophisticated algorithmic improvements across every phase of the hybrid video coding framework:

A. Advanced Block Partitioning (QTMT Structure)

While HEVC relied on a Quad-Tree (QT) block structure with maximum Coding Tree Unit (CTU) sizes of $64 \times 64$, VVC introduces Quad-Tree with Multi-Type Tree (QTMT).

• Maximum CTU size is expanded to $128 \times 128$ pixels.
• Allows recursive splitting into binary trees (horizontal/vertical) and ternary trees (three-way partitioning).
• Enables the encoder to contour partition blocks closely around complex object boundaries, fine textures, and smooth gradients.

B. Enhanced Intra-Picture Prediction

For individual frames (I-frames), VVC vastly enriches spatial direction modeling:

• 67 Intra Prediction Modes: Expanded from HEVC’s 35 modes to capture precise edge orientations.
• Wide-Angle Prediction: Seamlessly models angles exceeding conventional 45-degree diagonals for non-square blocks.
• Cross-Component Linear Model (CCLM): Predicts chroma (color) samples directly from reconstructed luma (brightness) samples, cutting chroma transmission overhead.
• Multiple Reference Line (MRL): Uses non-adjacent boundary pixel lines to predict complex background areas.

C. Precision Inter-Picture Prediction & Motion Compensation

For temporally predicted frames (P-frames and B-frames), VVC introduces:

• Sub-block Temporal Motion Vector Prediction (SbTMVP): Subdivides a block into $8 \times 8$ sub-blocks, assigning unique motion vectors to each sub-region.
• Affine Motion Inter Prediction: Models complex real-world transformations such as zooming, rotation, stretching, and perspective tilt using 4-parameter or 6-parameter models.
• Bi-Directional Optical Flow (BDOF): Refines motion compensation at the individual sample level without adding signaling overhead.
• Decoder-Side Motion Vector Refinement (DMVR): Fine-tunes motion vectors directly on the decoder side, saving transmission bits.

D. Advanced In-Loop Filtering

To eliminate visual artifacts (blocking, ringing, and color blurring), VVC employs a 4-stage filtering cascade:

1. Deblocking Filter (DBF): Smooths block boundary discontinuities.
2. Sample Adaptive Offset (SAO): Corrects edge and band offsets.
3. Adaptive Loop Filter (ALF): Utilizes 25 diamond-shaped filter shapes to reconstruct high-frequency texture details.
4. Cross-Component ALF (CC-ALF): Refines chroma details using luma channel information.

E. Native Subpicture & Omnidirectional Support

VVC defines Subpictures as independently decodable rectangular regions within a video canvas. This allows VR headsets to stream only the high-resolution viewport the user is looking at while transmitting peripheral vision in lower quality (foveated streaming), saving up to 70% in VR streaming bandwidth.

4. How to Create and Encode VVC Files

Encoding VVC content requires software libraries that implement the H.266 standard. The most mature open-source encoder is VVenC, developed by Fraunhofer Heinrich Hertz Institute (HHI).

Method 1: Using Fraunhofer VVenC (Stand-alone CLI)

1. Download the precompiled VVenC binaries or build from source via GitHub (https://github.com/fraunhoferhhi/vvenc).
2. Run the command to encode an uncompressed YUV or raw video stream:

vvencapp --input input_yuv420p.yuv --size 1920x1080 --fps 60 --preset medium --qp 28 --output output.266

Method 2: Encoding with FFmpeg (with libvvenc enabled)

Modern custom builds of FFmpeg compiled with libvvenc support allow direct conversion from standard video files (.mp4, .mov, .mkv):

ffmpeg -i input.mov -c:v libvvenc -preset medium -b:v 2500k -tag:v vvc1 -c:a aac -b:a 128k output_vvc.mp4

Key Encoding Presets:

• faster / fast: Best for rapid prototyping and live testing.
• medium: Recommended balance of compression ratio and encoding time.
• slow: Maximum visual quality and highest data compression for archival and VOD distribution.

5. How to Open and Play VVC Video Files

Because VVC is a cutting-edge standard, media player software and hardware decoders are actively rolling out support. Here is how to play VVC content today:

1. Fraunhofer VVdec Player

Fraunhofer HHI provides VVdec, an ultra-fast, multithreaded open-source software decoder.

• Download vvdecapp from GitHub.
• Play or decode bitstreams:

vvdecapp -b output.266 -o decoded_output.yuv

2. MPV Media Player (Recommended GUI/CLI Player)

Recent builds of MPV compiled with libvvdec or the latest FFmpeg backend can play VVC video inside .mp4 or .mkv containers out of the box:

mpv output_vvc.mp4

3. VLC Media Player

• VLC is integrating VVC support into its version 4.0 releases.
• Download the VLC 4.0 Nightly Build, open the player, and drag-and-drop the .mp4 or .vvc video file directly into the playback window.

4. MPC-HC / MPC-BE with LAV Filters (Windows)

Windows users can play VVC files using MPC-HC (Media Player Classic Home Cinema) paired with the latest nightly builds of LAV Filters that incorporate VVC decoding support.

VVC vs. HEVC vs. AV1 Comparison

Frequently Asked Questions (FAQ)

Q1: What is the main advantage of the VVC (H.266) video format over HEVC?
A1: VVC delivers approximately 50% data savings over HEVC while preserving the exact same visual quality.

Q2: Is VVC / H.266 a free, open-source video format?
A2: No, VVC is an ISO/ITU royalty-bearing standard licensed through patent pools like Access Advance, though open-source encoders/decoders exist.

Q3: What video container formats support VVC video bitstreams?
A3: VVC bitstreams are primarily packaged inside standard .mp4, .mkv (Matroska), and .ts (MPEG Transport Stream) container files.

Q4: Can VVC handle screen sharing and video game streaming effectively?
A4: Yes, VVC includes specialized Screen Content Coding (SCC) tools that compress computer interfaces, text, and synthetic graphics with exceptional sharpness.

Q5: When will mobile phones and smart TVs support native hardware VVC decoding?
A5: Major silicon vendors and smart TV manufacturers began incorporating hardware VVC decoders in premium chipsets starting in 2023–2024, with mainstream adoption expanding rapidly through 2026.

Summary & Future Outlook

Versatile Video Coding (VVC / H.266) sets the gold standard for compression efficiency and structural versatility. By cutting network bandwidth demands in half for 4K and unlocking realistic delivery channels for 8K, 16K, foveated virtual reality, and cloud gaming, VVC stands ready to power the next decade of digital visual experiences across the globe.

References

• Versatile Video Coding (VVC): The Next Generation Video Compression Standard