Hubbry Logo
VP8VP8Main
Open search
VP8
Community hub
VP8
logo
8 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Contribute something
VP8
VP8
VP8
View on Wikipedia
from Wikipedia
VP8
VP8 logo
Internet media typevideo/VP8
Developed byOn2 Technologies, Google
Initial releaseSeptember 13, 2008
Type of formatVideo coding format
Contained byWebM, Matroska
Extended fromVP7
Extended toVP9
StandardRFC 6386
Open format?Yes (specification under CC-by)[1]
Free format?See § History

VP8 is an open and royalty-free video compression format released by On2 Technologies in 2008.

Initially released as a proprietary successor to On2's previous VP7 format, VP8 was released as an open and royalty-free format in May 2010 after Google acquired On2 Technologies. Google provided an irrevocable patent promise on its patents for implementing the VP8 format, and released a specification of the format under the Creative Commons Attribution 3.0 license.[1] That same year, Google also released libvpx, the reference implementation of VP8, under the revised BSD license.[2]

Opera, Firefox, Chrome, Pale Moon, and Chromium support playing VP8 video in HTML video tag.[3] Internet Explorer officially supports VP8 if the user has the DirectShow filter installed.[4][5] According to Google, VP8 is mainly used in connection with WebRTC and as a format for short looped animations, as a replacement for the Graphics Interchange Format (GIF).[6]

VP8 can be multiplexed into the Matroska-based container format WebM along with Vorbis and Opus audio. The image format WebP is based on VP8's intra-frame coding. VP8's direct successor, VP9, and the royalty-free AV1 codec from the Alliance for Open Media are based on VP8.[7]

Features

[edit]

VP8 only supports progressive scan video signals with 4:2:0 chroma subsampling and 8 bits per sample. In its first public version, On2's VP8 implementation supports multi-core processors with up to 64 cores simultaneously. At least in the implementation (from August 2011), VP8 is comparatively badly adapted to high resolutions (HD). With only three reference frame buffers needed, VP8 enables decoder implementations with a relatively small memory footprint. The format features a pure intra mode, i.e. using only independently coded frames without temporal prediction, to enable random access in applications like video editing.

Technology

[edit]

VP8 is a traditional block-based transform coding format. It has much in common with H.264, e.g. some prediction modes.[8] At the time of first presentation of VP8, according to On2 the in-loop filter[9] and the Golden Frames[10] were among the novelties of this iteration. The first definition of such a filter is already found in the H.263 standard, though, and Golden Frames were already in use in VP5[11] and VP7.[12]

The discrete cosine transform (DCT) on 4×4 blocks and the Walsh–Hadamard transform (WHT) serve as basic frequency transforms. A maximum of three frames can be referenced for temporal prediction: the last Golden Frame (may be an intra frame), alternate reference frame, and the directly preceding frame. The so-called alternate reference frames (altref) can serve as reference-only frames for displaying them can be deactivated. In this case, the encoder can fill them with arbitrary useful image data, even from future frames, and thereby serve the same purpose as the b-frames of the MPEG formats.[13] Similar macroblocks can be assigned to one of up to four (even spatially disjoint) segments and thereby share parameters like the reference frame used, quantizer step size, or filter settings. VP8 offers two different adjustable deblocking filters that are integrated into the codec loops (in-loop filtering). Many coding tools use probabilities that are calculated continuously from recent context, starting at each intra frames. Macro blocks can comprise 4×4, 8×8, or 16×16 samples. Motion vectors have quarter-pixel precision.

History

[edit]

VP8 was first released by On2 Technologies on September 13, 2008, as On2 TrueMotion VP8, replacing its predecessor, VP7.[14][15]

After Google acquired On2 in February 2010,[16] calls for Google to release the VP8 source code were made. Most notably, the Free Software Foundation issued an open letter on March 12, 2010, asking Google to gradually replace the usage of Adobe Flash Player and H.264 on YouTube with a mixture of HTML5 and a freed VP8.[17]

Word of an impending open-source release announcement got out on April 12, 2010.[18] On May 19, at its Google I/O conference, Google released the VP8 codec software under a BSD-like license and the VP8 bitstream format specification under an irrevocable free patent license.[19][20][21] This made VP8 the second product from On2 Technologies to be opened, following their donation of the VP3 codec in 2002 to the Xiph.Org Foundation,[22] from which they derived the Theora codec.

In February 2011, MPEG LA invited patent holders to identify patents that may be essential to VP8 in order to form a joint VP8 patent pool. As a result, in March the United States Department of Justice (DoJ) started an investigation into MPEG LA for its role in possibly attempting to stifle competition.[23][24] In July 2011, MPEG LA announced that 12 patent holders had responded to its call to form a VP8 patent pool, without revealing the patents in question,[25] and despite On2 having gone to great lengths to avoid such patents.[26]

In November 2011, the Internet Engineering Task Force published the informational RFC 6386, VP8 Data Format and Decoding Guide.

In March 2013, MPEG LA announced that it had dropped its effort to form a VP8 patent pool after reaching an agreement with Google to license the patents that it alleges "may be essential" for VP8 implementation, and granted Google the right to sub-license these patents to any third-party user of VP8 or VP9.[27][28] This deal has cleared the way for possible MPEG standardisation as its royalty-free internet video codec, after Google submitted VP8 to the MPEG committee in January 2013.[29]

In March 2013, Nokia asserted a patent claim against HTC and Google for the use of VP8 in Android in a German court;[30] however, on August 5, 2013, the webm project announced that the German court has ruled that VP8 does not infringe Nokia's patent.[31]

Nokia has made an official intellectual property rights (IPR) declaration to the IETF with respect to the VP8 Data Format and Decoding Guide listing 64 granted patents and 22 pending patent applications.[32]

Implementations

[edit]

libvpx

[edit]
Main article: libvpx

The reference implementation of a VP8 (and VP9) codec is found in the programming library libvpx which is released as free software. It has a mode for one-pass and two-pass encoding, respectively, while the one-pass mode is known as being broken and not offering effective control over the target bitrate.[33][failed verification][34][failed verification]

Currently, libvpx is primary software library capable of encoding VP8 video streams,[35] but at least one independent implementation exists in ffvp8enc.

Encoding

[edit]

A Video for Windows wrapper of the VP8 codec based on the Google VP8 library (FourCC: VP80) is available.[36]

The WebM Project hardware team in Finland released an RTL hardware encoder for VP8 that is available at no cost for semiconductor manufacturers.[37][38]

The Nvidia Tegra mobile chipsets have full VP8 hardware encoding and decoding (since Tegra 4).[39]

Nexus 5 could use hardware encoding[40].

Decoding

[edit]

libvpx is capable of decoding VP8 video streams.[41]

On July 23, 2010, Fiona Glaser, Ronald Bultje, and David Conrad of the FFmpeg Team announced the ffvp8 decoder. Through testing, they determined that ffvp8 was faster than Google's own libvpx decoder.[42] The WebM Project hardware team released an RTL hardware decoder for VP8, that is releasable to semiconductor companies at zero cost.[38][43] TATVIK Technologies announced a VP8 decoder that is optimized for the ARM Cortex-A8 processor.[44] Marvell's ARMADA 1500-mini chipset has VP8 SD and HD hardware decoding support (used in Chromecast).[45] Intel has full VP8 decoding support built into their Bay Trail chipsets.[46] Intel Broadwell also adds VP8 hardware decoding support.[47]

Operating system support

[edit]
VP8 support by different operating systems
Microsoft Windows macOS BSD / Linux Android OS iOS
Codec support Yes Yes Yes Yes Yes
Container support On Windows 10 Anniversary Update (1607):
WebM (.webm is not recognised; requires pseudo extension)
Matroska (.mkv)
On Windows 10 October 2018 Update (1809):
WebM (.webm is recognised officially) 		WebM (.webm)
- Introduced in macOS 11.3 		WebM (.webm)
Matroska (.mkv) 		WebM (.webm)
Matroska (.mkv) 		WebM (.webm)
- Introduced in iOS 17.4
Notes On Windows 10:
- On Anniversary Update (1607), limited support is available in Microsoft Edge (via MSE only) and Universal Windows Platform apps.
- On April 2018 Update (1803) with Web Media Extensions preinstalled, Microsoft Edge (EdgeHTML 17) supports VP8 videos embedded in <video> tags.
- On October 2018 Update (1809), VP9 Video Extensions is preinstalled. It enables encoding of VP8 and VP9 content on devices that don't have a hardware-based video encoder.[48] 		- - Support introduced in Android 2.3.3+
- Streamable in Android 4.0+
[edit]

WebM

[edit]
Main article: WebM

Also on May 19, 2010, the WebM Project was launched, featuring contributions from "Mozilla,[49] Opera,[50][51] Google[52] and more than forty other publishers, software and hardware vendors" in a major effort to use VP8 as the video format for HTML5.[53] In the WebM container format, the VP8 video is used with Vorbis or Opus audio.[54][55] Internet Explorer 9 will support VP8 video playback if the proper codec is installed.[5] Android is WebM-enabled from version 2.3 - Gingerbread.[56] Since Android 4.0, VP8 could be read inside mkv[57] and WebM could be streamed.[58] Adobe also announced that the Flash Player will support VP8 playback in a future release.[59]

WebP

[edit]
Main article: WebP

On September 30, 2010, Google announced WebP, their new image format, on the Chromium blog.[60] WebP is based on VP8's intra-frame coding and uses a container based on Resource Interchange File Format (RIFF).

Comparison with H.264

[edit]

While H.264/MPEG-4 AVC contains patented technology and requires licenses from patent holders and limited royalties for hardware, Google has irrevocably released the VP8 patents it owns under a royalty-free public license.[19][61]

According to a comparison of VP8 (encoded with the initial release of libvpx) and H.264 conducted by StreamingMedia, it was concluded that "H.264 may have a slight quality advantage, but it's not commercially relevant" and that "Even watching side-by-side (which no viewer ever does), very few viewers could tell the difference". They also stated that "H.264 has an implementation advantage, not a technology advantage."[62]

Google claims that VP8 offers the "highest quality real-time video delivery"[63] and Libvpx includes a mode where the maximum CPU resources possible will be used while still keeping the encoding speed almost exactly equivalent to the playback speed (realtime), keeping the quality as high as possible without lag. On the other hand, a review conducted by streamingmedia.com in May 2010 concluded that H.264 offers slightly better quality than VP8.[64]

In September 2010 Fiona Glaser, a developer of the x264 encoder, gave several points of criticism for VP8, claiming that its specification was incomplete, and the performance of the encoder's deblocking filter was inferior to x264 in some areas.[65] In its specification, VP8 should be a bit better than H.264 Baseline Profile and Microsoft's VC-1. Encoding is somewhere between Xvid and VC-1. Decoding is slower than FFmpeg's H.264, but this aspect can hardly be improved due to the similarities to H.264. Compression-wise, VP8 offers better performance than Theora and Dirac. According to Glaser, the VP8 interface lacks features and is buggy, and the specification is not fully defined and could be considered incomplete. Much of the VP8 code is copy-pasted C code, and since the source constitutes the actual specification, any bugs will also be defined as something that has to be implemented to be in compliance.

In 2010, it was announced that the WebM audio/video format would be based on a profile of the Matroska container format together with VP8 video and Vorbis audio.[55]

See also

[edit]

References

[edit]

Further reading

[edit]
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

VP8

View on Grokipedia
from Grokipedia
VP8 is an open-source video compression format and video codec developed originally by On2 Technologies and released by Google in May 2010 as the core component of the royalty-free WebM multimedia project. It employs block-based hybrid coding techniques, including intra- and inter-frame prediction, motion compensation, and discrete cosine transform (DCT) or Walsh-Hadamard transform (WHT) for residue encoding, to achieve efficient compression of progressive-scan video in YUV 4:2:0 color format with 8 bits per channel. Designed primarily for web delivery, VP8 supports resolutions up to 16383×16383 pixels and is optimized for low-complexity decoding suitable for real-time streaming and playback on diverse hardware platforms. The codec's development stemmed from On2's earlier VP6 and VP7 formats, with acquiring On2 in 2010 to open-source VP8 under a BSD-like , aiming to provide a high-quality alternative to proprietary standards like H.264 without licensing fees. Released alongside the container, which pairs VP8 video with or Opus audio, it was intended to foster an open web media ecosystem, with initial support from browser vendors including , Mozilla Firefox, and . The format was later documented in RFC 6386 by the (IETF) in November 2011, serving as an informational specification for its bitstream structure and decoding process. Key technical innovations in VP8 include quarter-pixel motion compensation using 6-tap filters for improved accuracy, adaptive loop filtering to reduce blocking artifacts, and support for three reference frames (last, golden, and alternate) to enhance prediction efficiency. It features macroblock-based segmentation into 16×16 luma blocks with 4×4 subblocks, entropy coding via boolean and token trees, and up to eight parallel partitions for faster decoding. Performance benchmarks indicate VP8 decoding is approximately 30% faster than H.264 High Profile on contemporary processors like i7, while offering comparable compression efficiency at similar quality levels (30-45 dB PSNR). These attributes have positioned VP8 as a foundational for web video, influencing successors like and , though adoption has been tempered by favoring H.264 and HEVC in some contexts.

History

Development and Initial Release

On2 Technologies, a pioneer in video compression software, built upon its earlier successes with the VPx family of codecs, which evolved significantly for web-based applications. The company's VP6 codec, introduced in 2005, became a cornerstone for Flash Video delivery, enabling efficient streaming within Adobe Flash Player starting from version 8, due to its balance of compression and compatibility with limited bandwidth environments. VP7 followed as an enhancement, offering improved quality, though its adoption remained more niche compared to VP6, primarily in professional encoding tools and select embedded applications. These predecessors addressed key limitations in earlier codecs like VP3, focusing on higher fidelity for internet video while maintaining low computational demands. In response to growing demands for superior web video compression amid rising online streaming, On2 announced on September 13, 2008, positioning it as a successor to VP7 with claims of surpassing H.264, , and in quality and performance. The initial release occurred simultaneously, introducing On2 TrueMotion as a royalty-bearing licensed available to developers and platform providers. goals centered on delivering high-quality compression optimized for web video, targeting peak signal-to-noise ratios (PSNR) of 30-45 dB under bandwidth constraints typical of delivery. VP8 supported resolutions up to 16383×16383 pixels, , depth, and to align with prevalent web video formats and diverse hardware capabilities. Technically, it drew inspiration from block-based hybrid coding paradigms akin to H.264, employing motion-compensated prediction and (DCT) for intra- and inter-frame compression, but differentiated through a custom, simpler scheme using boolean arithmetic to enhance efficiency without the complexity of context-adaptive models. Under its early proprietary model, saw limited but targeted adoption, with announcing support for in an upcoming release of Flash Player. This licensing approach allowed On2 to monetize the technology while fostering initial ecosystem development for high-definition web content. Subsequently, following Google's acquisition of On2, was open-sourced in , broadening its accessibility.

Google Acquisition and Open-Sourcing

In August 2009, Google announced its agreement to acquire On2 Technologies, the developer of the VP8 video codec, for approximately $106.5 million in an all-stock transaction aimed at advancing high-quality video compression for the web. The acquisition was completed on February 19, 2010, following shareholder approval, with the final value reaching $124.6 million. This move integrated On2's video expertise, including VP8—which built on the company's earlier proprietary codecs like VP7—into Google's ecosystem to support broader online video adoption. On May 19, 2010, open-sourced the codec under a BSD-style license as part of the newly launched Project, which bundled for video compression with the audio codec in a royalty-free format designed for web video. The Project aimed to provide an open alternative to proprietary video standards, enabling free use and distribution without licensing fees. The initiative quickly garnered support from key industry players, including and , which integrated support into their browsers shortly after the announcement, while committed to adding compatibility in an upcoming Flash release. further reinforced 's accessibility by offering a royalty-free patent cross-license to essential patents, distinguishing it from H.264, which requires royalty payments to patent pools like . This commitment helped mitigate potential encumbrances and promoted as a viable for web interoperability.

Technical Specifications

Encoding Mechanism

VP8 employs a block-based hybrid coding structure to compress video data, dividing each frame into 16x16 macroblocks for luma components and corresponding 8x8 blocks for chroma, which are further subdivided into 4x4 blocks for processing residuals and prediction. This approach allows for efficient handling of spatial and temporal redundancies by processing macroblocks in raster-scan order, with optional segmentation into up to four regions for adaptive parameter tuning. Intra-frame prediction in VP8 leverages spatial correlations within a frame, using modes such as DC (average of neighbors), vertical, horizontal, and true-motion (TM) for 16x16 luma and 8x8 chroma blocks, while 4x4 luma blocks support ten directional modes including DC, vertical, horizontal, and diagonal variants. Inter-frame exploits temporal redundancies by referencing one of three frames—last, golden, or alternate—with achieving up to 1/4-pixel accuracy for luma and 1/8-pixel for chroma via six-tap filters, and flexible partitioning via split motion vectors for sub-blocks. After , residual errors are computed and transformed using a 4x4 integer (DCT) for individual 4x4 blocks, with a 4x4 Walsh-Hadamard transform applied to the DC coefficients of luma blocks in larger intra modes to capture low-frequency energy efficiently. These transform coefficients undergo scalar quantization with 128 levels adjustable across six bands (Y AC/DC, Y2 DC/AC, UV DC/AC), enabling coarse-to-fine control over compression quality. Entropy coding in VP8 utilizes a custom boolean arithmetic coder, which encodes symbols like coefficients, motion vectors, and modes using adaptive probability tables updated per frame, with a 12-token for DCT coefficients (e.g., zero runs, end-of-block, and category values) to achieve probability-based compression without the context-adaptive binary (CABAC) of H.264. The coder supports up to eight token partitions for parallelism, encoding coefficients in a that prioritizes significant values. To mitigate blocking artifacts, an in-loop is applied post-reconstruction, classifying edges as interior or boundary types and adjusting filter strengths (0-63 levels) based on mode, quantization index, and sharpness parameters, with simple (2-tap) or normal (6-tap) modes for different edge variances. Rate control in VP8 implementations supports single-pass encoding for real-time scenarios, targeting constant bitrate (CBR) or constant quality (CQ) via fixed quantization parameters, and two-pass modes for offline processing that analyze the video first to allocate bits optimally for (VBR) or targeted quality, improving efficiency over single-pass. These mechanisms, inspired by prior On2 codecs like VP7, enable VP8 to balance and visual fidelity across diverse content.

Decoding Process and Key Features

The VP8 decoding begins with , where the decoder extracts frame headers, data, motion vectors, and quantized transform coefficients using a boolean decoder based on a range coder. This stage identifies key elements such as frame type (intra or inter), segmentation parameters, and reference frame indices, enabling the decoder to uncompressed header chunks (10 bytes for key frames, 3 bytes for inter frames) before delving into compressed partitions. Following parsing, inverse quantization is applied to the decoded coefficients, multiplying them by segment-specific dequantization factors derived from quantization indices (7-bit for base values, adjustable by 16-bit signed deltas). These factors vary by color plane (Y, U, V, or Y2 for luma in key frames) and coefficient position (DC or AC), using predefined lookup tables to restore approximate transform domain values. The process then proceeds to the inverse discrete cosine transform (IDCT), employing a 4x4 integer IDCT for Y, U, and V subblocks with fixed-point approximations of trigonometric constants, while a Walsh-Hadamard transform (WHT) handles the Y2 subblock in key frames. Motion compensation follows for inter frames, utilizing motion vectors with quarter-pixel precision to predict macroblocks from reference frames (last, golden, or alternate reference). Sub-pixel employs a 6-tap filter kernel for luma and 4-tap bicubic for chroma, ensuring smooth reconstruction without bidirectional prediction in the baseline profile, which relies solely on P-frames and forward-referencing mechanisms. Finally, frame reconstruction adds the inverse-transformed residue to the motion-compensated prediction, clips values to the 0-255 range, and applies a loop filter (2-tap simple or 6-tap normal) across block boundaries to mitigate artifacts, yielding the output frame in order. VP8 decoding supports frame rates up to 60 fps through its profile levels, with the baseline profile optimized for web playback and defining constraints on resolution and processing speed via maximum s per second. For instance, Level 3.2 supports resolutions up to 2048×1152 and accommodates at 30 fps within its processing constraints, while higher levels like 3.3 support increased rates or resolutions. Alpha channel support for transparency is available in extensions, such as those integrated into the container, where a separate VP8-encoded alpha plane is decoded alongside the primary video using compatible formats like YUVA 4:2:0. Spatial and temporal modes enhance adaptability, with temporal layers achieved through hierarchical frame usage (e.g., base and enhancement layers via golden frame updates) and spatial via post-decoding factors like 5/4 or 2x, or segmentation for region-specific processing. resilience is bolstered by frame-level through periodic key frames for and recovery, alongside golden and alternate frames that facilitate partial reconstruction after ; in cases of , the decoder may employ concealed frame copying from prior valid references. A distinctive aspect of VP8 decoding is the absence of B-frames and bi-directional in the baseline, emphasizing single-directional P-frame reliance augmented by golden frame referencing for long-term temporal , which updates selectable past frames to improve efficiency without backward dependencies.

Implementations and Support

Reference Software Implementation

The reference software implementation for VP8 is , an open-source library developed by and released on May 19, 2010, under a BSD license. This library serves as the official SDK for VP8, providing both encoding and decoding capabilities essential for integrating VP8 into applications and tools. It includes the vp8.h header file, which defines the core API for VP8 controls shared between the encoder and decoder, enabling developers to access frame data structures and codec interfaces programmatically. libvpx incorporates command-line tools for practical usage, including vpxenc for encoding bitstreams and vpxdec for decoding them. The vpxenc encoder supports key options such as --cpu-used, which adjusts the speed-quality (values from 0 for slowest/highest quality to 5 for fastest/lowest quality), and deadline modes including "realtime" for low-latency live encoding and "good" for balanced offline encoding. These parameters allow users to tailor performance for specific scenarios, such as web streaming or archival compression. Meanwhile, vpxdec facilitates playback of streams and leverages multi-threaded decoding on multi-core systems, utilizing frame-based threading to distribute workload across CPU cores for improved efficiency during video rendering. Since its inception, has been integrated into major multimedia frameworks, notably FFmpeg, which added VP8 support in version 0.6 in 2010 using as the backend for both encoding and decoding operations. Subsequent FFmpeg updates post-2020 have incorporated enhancements, such as improved SIMD optimizations and threading improvements from releases like version 1.13 in 2023. As of 2025, is actively maintained under the WebM Project by in collaboration with the , with version 1.15.2 released in July 2025 featuring security fixes (including CVE-2025-5283 for VP8) and ongoing speed optimizations for both VP8 and VP9 processing.

Hardware and Platform Support

VP8 has been implemented on various platforms, primarily through SIMD instructions and dedicated video processing units. On ARM-based mobile devices, such as those running Android, VP8 decoding and encoding leverage intrinsics, an extension to the ARMv7 and ARMv8 architectures that enables efficient vector processing for video tasks. This support became available with Android 4.0 () in 2011, allowing real-time VP8 processing on single-core processors and later models. Intel's Quick Sync Video provides partial hardware acceleration for VP8, focusing on decoding (from 5th Generation Core processors) and encoding (from the Skylake 6th Generation Core onward), including support in integrated graphics for up to 4K resolutions. NVIDIA's NVDEC for decoding has supported VP8 since the Pascal architecture in 2016, enabling hardware-accelerated processing on , , and Tesla GPUs across Windows and ; NVENC does not support VP8 encoding. These implementations often fall back to software decoding using the library when hardware is unavailable or insufficient. At the operating system level, VP8 is natively integrated into Android since version 4.0 in 2011, supporting playback and streaming within the media framework. On Windows, integration occurs through via third-party plugins, such as the WebM Media Foundation Components, which enable VP8 playback in applications like . In Linux environments, VP8 is handled natively through the multimedia framework, which includes dedicated elements like vp8dec for decoding and vp8enc for encoding. Browser support for has been robust since its early adoption. Google introduced VP8 support in version 6 in 2010 as part of WebM integration. Mozilla Firefox followed in March 2011 with partial WebM/VP8 capabilities, expanding to full support in subsequent releases. supports VP8 via WebM containers natively, while Apple's offers partial support starting from version 12.1 in 2019, limited primarily to WebRTC scenarios rather than general HTML5 video. As of 2025, remains widespread for legacy web video content and certain streaming applications but is declining in favor of newer codecs like , which offers superior compression efficiency and growing hardware support. It continues as a default option in some implementations across browsers, ensuring compatibility for real-time communication where adoption is still maturing. A key challenge for deployment is its limited dedicated () support compared to H.264, which has been embedded in hardware decoders since the early across a broader range of devices. This results in greater reliance on software fallbacks for on older or budget hardware, potentially increasing power consumption and processing demands.

Associated Standards

WebM Container Format

WebM was introduced by on May 19, 2010, as an output of the open-source WebM Project, aimed at delivering a high-quality, audiovisual media format optimized for web delivery. This initiative sought to address the limitations of proprietary formats by providing an accessible alternative for embedding multimedia content directly in web pages. serves as the default in WebM files. The container format is a profile of the multimedia container, employing the . file extension for its files. It encapsulates video streams alongside or Opus audio streams, structured using the Extensible Binary Meta Language (EBML) for the header and metadata organization. EBML enables a hierarchical, binary representation of data, facilitating efficient parsing and extension while maintaining compatibility with 's core framework. Central to WebM's structure is the Segment element, which houses the primary audiovisual data clusters and supports seeking within the file. The Cues element provides an index of timestamps and byte positions, primarily referencing video keyframes to enable quick navigation and playback control. Additionally, the Attachments element allows embedding of supplementary files, such as subtitles or fonts, enhancing multimedia presentation without external dependencies. The format accommodates a maximum of 2^{63} bytes, leveraging EBML's variable-length integer encoding for large-scale media. The underlying container format, on which is based, has been under IETF standardization since 2016 and was published as RFC 9559 in 2024. The container is distributed royalty-free under the Attribution 3.0 Unported license, ensuring broad accessibility for developers and users. It is predominantly utilized with the <video> element to support seamless, patent-unencumbered streaming of web video content across browsers.

WebP Image Format

announced the WebP image format on September 30, 2010, adapting the intra-frame prediction techniques from the to enable efficient compression for still images. This format was designed to provide superior compression compared to established standards like and , supporting both lossy and lossless modes while maintaining compatibility with web infrastructure. In its lossy mode, WebP employs keyframe encoding, utilizing 4x4 (DCT) blocks for compressing RGB image data, which allows for tunable quality levels similar to but with enhanced efficiency. Transparency support is integrated via a separate alpha plane, encoded using a lossless VP8L and stored in an 'ALPH' chunk, enabling semi-transparent pixels without significantly increasing file size. The lossless mode builds on VP8's foundations with additional optimizations, including spatial modes that estimate values from neighboring samples (such as left, top, or diagonal predictors), a color cache to reuse frequent colors via hashing, and with canonical Huffman trees for compact representation of literals, LZ77 matches, and cache indices. These enhancements result in lossless files that are approximately 26% smaller than equivalent PNGs, preserving exact while reducing bandwidth needs. WebP files use a RIFF-based container with the .webp extension, starting with a 'RIFF' header followed by a 'WEBP' form type and various chunks: 'VP8 ' or 'VP8L' for core image data, 'VP8X' for extended features like alpha and animation flags, 'ICCP' for ICC color profiles, and 'EXIF' for metadata such as camera settings. This structure ensures modularity, allowing optional elements without affecting basic decoding. By 2025, WebP has achieved widespread adoption, with support in major browsers including Chrome since version 32 (2014), Opera since version 19 (2014), Edge since version 18 (2018), and Firefox since version 65 (2019), alongside full integration in Safari from version 16 (2022). It is extensively used in Google services such as Search, Maps, and YouTube for thumbnails and static assets to optimize loading times. Despite its widespread browser support, WebP has encountered compatibility challenges in non-browser applications. Some photo editing software, such as older versions of Adobe Photoshop, requires additional plugins for native support, and operating systems like Windows may exhibit issues when editing WebP file properties or setting them as wallpapers, often necessitating conversions to formats like PNG or JPEG. Users have reported frustrations in these scenarios, with workarounds including the use of the Windows Photos app, online conversion tools, or scripts for batch processing. However, advocates emphasize WebP's efficiency benefits, such as reduced file sizes and improved web performance, arguing that these advantages outweigh limitations in web-centric use cases. The format was standardized by the IETF as RFC 9649 in November 2024, encompassing both lossy (building on RFC 6386 for VP8) and lossless modes, affirming its role as a web-native image standard.

Performance and Comparisons

Comparison with H.264/AVC

VP8 and H.264/AVC exhibit comparable compression efficiency in terms of PSNR at equivalent bitrates when compared to H.264's baseline profile, though VP8 typically results in 20-30% larger file sizes for the same quality compared to H.264 High Profile due to its simpler mechanism. This disparity arises because H.264 employs more advanced context-adaptive (CABAC) or variable-length coding (CAVLC), which provide better compression ratios than VP8's non-adaptive bool coder. In terms of encoding speed, is generally 5-25 times slower than H.264 on software implementations, primarily because it lacks support for B-frames, which H.264 uses to achieve up to 20% better compression efficiency through bi-directional prediction. H.264 benefits from extensive across devices, further widening the performance gap in real-time applications. A key practical difference lies in licensing: is fully under an irrevocable patent license granted by in 2013, allowing unrestricted use without fees. In contrast, H.264 requires licensing through the , which imposes royalties for commercial distributions, though free for non-commercial . Both codecs support intra (I) and predicted (P) frames with block-based at quarter-pixel precision, enabling efficient inter-frame prediction. However, offers more intra-prediction modes (10 for 4x4 luma blocks versus H.264's 9) but H.264 provides additional advanced options overall, while 's bool coder is simpler but less efficient. Additionally, lacks native support for , limiting its applicability to content, whereas H.264 handles both progressive and interlaced formats. In real-world deployment as of 2025, H.264 remains dominant in broadcast and television applications due to its mature ecosystem and hardware support, while VP8 finds primary use in web-based archiving and streaming via the container, particularly where royalty-free alternatives are prioritized. Benchmarks from the MSU Comparison indicate VP8 performs about 20-30% worse in quality at low bitrates compared to H.264, making it less ideal for bandwidth-constrained scenarios.

Relation to Successor Codecs

VP9, released in 2013 by as the direct successor to VP8, builds upon its predecessor's core architecture while introducing enhancements for improved compression . Key advancements include support for larger block sizes up to 64x64 pixels, enabling more precise partitioning and compared to VP8's maximum of 16x16 macroblocks. Additionally, VP9 employs advanced intra- and inter-prediction modes with multiple reference frames, effectively emulating bidirectional prediction without traditional B-frames, which contributes to its overall performance gains. These modifications allow VP9 to achieve 30-50% better compression over VP8 at equivalent levels, reducing bitrate requirements for the same video . AV1, finalized in 2018 through the collaborative efforts of the Alliance for Open Media (AOM)—a consortium including Google, Netflix, and other major stakeholders—serves as a conceptual and technical evolution from VP8 via VP9. It incorporates and refines VP8's foundational tools, such as block-based hybrid coding, while adding sophisticated features like extended transform types, compound prediction modes, and advanced loop filtering to boost efficiency. Notably, AV1 introduces film grain synthesis, a parametric model that removes grain during encoding and regenerates it at the decoder to preserve artistic texture without bloating the bitstream, enhancing perceptual quality for cinematic content. This results in approximately 30% greater compression efficiency over VP9, translating to smaller file sizes or higher quality at the same bitrate. By 2025, VP8 persists primarily in legacy applications, such as older files and baseline implementations for broad compatibility in real-time communication. In contrast, and have become the preferred choices for high-resolution streaming, with platforms like deploying for 1440p content and for 4K and above to optimize bandwidth and quality. Mozilla's shift to defaulting in calls exemplifies this transition, prioritizing efficiency for modern 4K video delivery. The patent landscape established for has been extended to its successors through AOM's irrevocable commitments, where participating members pledge not to assert essential patents against implementers, ensuring and remain accessible without licensing fees. This open policy, formalized in AOM's agreements, has facilitated widespread adoption by avoiding the royalty burdens associated with codecs. In terms of performance, AV1's encoding complexity is substantially higher, often 5-10 times slower than and up to 10 times slower than due to its advanced tools, though hardware accelerations are mitigating this gap in 2025. Decoding speeds for AV1 remain comparable to and on modern hardware, benefiting from parallel processing features. Meanwhile, narrows the efficiency gap with H.264 High Profile, offering competitive quality at lower bitrates for many use cases.

References

  1. https://research.google.com/pubs/archive/37073.pdf
  2. https://datatracker.ietf.org/doc/html/rfc6386
  3. https://www.eetimes.com/the-on2-vp6-codec-how-it-works/
  4. https://www.streamingmedia.com/Glossary/Terms/VP7
  5. https://www.streamingmedia.com/Articles/Editorial/Featured-Articles/WebM-Not-Ready-for-Prime-Time-67538.aspx
  6. https://www.loc.gov/preservation/digital/formats/fdd/fdd000578.shtml
  7. https://www.streamingmediablog.com/2010/05/adobe-announces-flash-player-support-for-googles-vp8-video-codec.html
  8. http://googlepress.blogspot.com/2009/08/google-to-acquire-on2-technologies_05.html
  9. https://www.reuters.com/article/idUS3097266726/
  10. https://www.streamingmedia.com/Articles/News/Online-Video-News/On2-Shareholders-Approve-Google-Acquisition-65764.aspx
  11. https://www.osnews.com/story/23322/breaking-google-opens-vp8-codec-enables-it-on-youtube/
  12. https://techcrunch.com/2010/05/19/webm-google-h-264/
  13. https://betanews.com/2010/05/19/google-open-sources-web-video-codec-mozilla-opera-adobe-sign-on/
  14. https://www.webmproject.org/cross-license/vp8/agreement/
  15. https://arstechnica.com/information-technology/2013/03/google-mpeg-la-agree-to-royalty-free-terms-for-vp8-video-codec/
  16. https://www.webmproject.org/docs/encoder-parameters/
  17. https://datatracker.ietf.org/doc/html/rfc6386#section-7
  18. https://datatracker.ietf.org/doc/html/rfc6386#section-9
  19. https://datatracker.ietf.org/doc/html/rfc6386#section-11
  20. https://datatracker.ietf.org/doc/html/rfc6386#section-9.6
  21. https://datatracker.ietf.org/doc/html/rfc6386#section-14.1
  22. https://datatracker.ietf.org/doc/html/rfc6386#section-14.2
  23. https://datatracker.ietf.org/doc/html/rfc6386#section-14.3
  24. https://datatracker.ietf.org/doc/html/rfc6386#section-16
  25. https://datatracker.ietf.org/doc/html/rfc6386#section-18
  26. https://datatracker.ietf.org/doc/html/rfc6386#section-14.5
  27. https://datatracker.ietf.org/doc/html/rfc6386#section-15
  28. https://datatracker.ietf.org/doc/html/rfc6386#section-9.1
  29. https://datatracker.ietf.org/doc/html/rfc6386#section-5
  30. https://wiki.webmproject.org/alpha-channel
  31. https://datatracker.ietf.org/doc/html/rfc6386#section-9.3
  32. https://datatracker.ietf.org/doc/html/rfc6386#section-9.7
  33. https://datatracker.ietf.org/doc/html/rfc6386#section-16.2
  34. https://datatracker.ietf.org/doc/html/rfc6386#section-3
  35. https://multimedia.cx/eggs/vp8-and-ffmpeg/
  36. https://www.phoronix.com/news/libvpx-1.13-Released
  37. https://security.gentoo.org/glsa/202509-07
  38. https://developer.android.com/ndk/guides/cpu-arm-neon
  39. https://www.osnews.com/story/24502/webm-project-releases-new-version-of-vp8-codec/
  40. https://developer.android.com/about/versions/android-4.0-highlights
  41. https://www.intel.com/content/www/us/en/docs/onevpl/developer-reference-media-intel-hardware/1-1/overview.html
  42. https://developer.nvidia.com/video-codec-sdk
  43. https://docs.nvidia.com/video-technologies/video-codec-sdk/12.1/nvdec-video-decoder-api-prog-guide/index.html
  44. https://learn.microsoft.com/en-us/windows/win32/medfound/supported-media-formats-in-media-foundation
  45. https://gstreamer.freedesktop.org/documentation/vpx/vp8dec.html
  46. https://gstreamer.freedesktop.org/documentation/vpx/vp8enc.html
  47. https://blog.chromium.org/2011/01/html-video-codec-support-in-chrome.html
  48. https://developer.mozilla.org/en-US/docs/Web/Media/Guides/Formats/Video_codecs
  49. https://www.red5.net/blog/av1-vs-vp9-vs-vp8-comparison-for-live-streaming/
  50. https://developer.mozilla.org/en-US/docs/Web/Media/Guides/Formats/WebRTC_codecs
  51. http://blog.webmproject.org/2010/05/introducing-webm-open-web-media-project.html
  52. https://www.webmproject.org/docs/container/
  53. https://www.rfc-editor.org/rfc/rfc8794.html
  54. https://www.rfc-editor.org/rfc/rfc9559.html
  55. https://www.loc.gov/preservation/digital/formats/fdd/fdd000518.shtml
  56. https://datatracker.ietf.org/doc/rfc9559/
  57. https://developers.google.com/speed/webp
  58. https://developers.google.com/speed/webp/docs/riff_container
  59. https://developers.google.com/speed/webp/docs/webp_lossless_bitstream_specification
  60. https://caniuse.com/webp
  61. https://www.img2go.com/blog/why-webp-sparks-so-much-debate
  62. https://cloudinary.com/guides/front-end-development/webp-format-technology-pros-cons-and-alternatives
  63. https://learn.microsoft.com/en-us/answers/questions/4156426/windows-photos-app-support-for-webp-files
  64. http://compression.ru/video/codec_comparison/h264_2010/vp8_vs_h264.html
  65. https://deeplearning.eng.wayne.edu/_resources/pdf/detailed-comparitive.pdf
  66. https://tmpgenc.pegasys-inc.com/en/product/tvmw8_spec.html
  67. https://www.dacast.com/blog/best-video-codec/
  68. https://www.videoproc.com/resource/best-video-codec.htm
  69. https://edgeone.ai/learning/vp8-vs-vp9
  70. https://videoconvert.minitool.com/video-converter/vp8-vs-vp9.html
  71. https://www.fastpix.io/blog/vp8-vs-vp9-video-codec-differences
  72. https://aomediacodec.github.io/av1-spec/av1-spec.pdf
  73. https://netflixtechblog.com/av1-scale-film-grain-synthesis-the-awakening-ee09cfdff40b
  74. https://www.gumlet.com/learn/av1-vs-vp9/
  75. https://www.streamingmediaglobal.com/Articles/Editorial/Featured-Articles/The-State-of-the-Video-Codec-Market-2025-168627.aspx
  76. https://streaminglearningcenter.com/codecs/which-codec-does-youtube-use-part-deux.html
  77. https://www.sisvel.com/insights/vp9-av1-q-and-a/
  78. http://aomedia.org/license/
  79. https://www.videoproc.com/resource/vp9-vs-av1.htm
  80. https://imagekit.io/blog/av1-vs-vp9/
Add your contribution
Related Hubs
Contribute something
User Avatar
No comments yet.