Glossary - Amazon Interactive Video Service


Also see the AWS glossary. In the table below, LL stands for IVS low-latency streaming; RT, IVS real-time streaming.

Term Description LL RT Chat
AAC Advanced Audio Coding. AAC is an audio coding standard for lossy digital audio compression. Designed to be the successor of the MP3 format, AAC generally achieves higher sound quality than MP3 at the same bitrate. AAC has been standardized by ISO and IEC as part of the MPEG-2 and MPEG-4 specifications.
Adaptive bitrate streaming Adaptive Bitrate (ABR) streaming allows the IVS player to switch to a lower bitrate when connection quality suffers, and to switch back to a higher bitrate when connection quality improves.
Adaptive streaming See Layered encoding with simulcast.
Administrative user An AWS user with administrative access to resources and services available in an AWS account. See Terminology in AWS Setup User Guide.
ARN Amazon Resource Name, a unique identifier for an AWS resource. Specific ARN formats depend on the resource type. For ARN formats used by IVS resources, see in Service Authorization Reference.
Aspect ratio Describes the ratio of frame width to frame height. For example, 16:9 is the aspect ratio that corresponds to the Full HD or 1080p resolution.
Audio mode A preset or custom audio configuration optimized for different types of mobile device users and the equipment that they use. See IVS Broadcast SDK: Mobile Audio Modes (Real-Time Streaming).
AVC, H.264, MPEG-4 Part 10 Advanced Video Coding, also referred to as H.264 or MPEG-4 Part 10, a video compression standard for lossy digital video compression.
Background replacement A type of camera filter that enables live-stream creators to change their backgrounds. See Background Replacement in IVS Broadcast SDK: Third-Party Camera Filters (Real-Time Streaming).
Bitrate A streaming metric for the number of bits transmitted or received per second.
Broadcast, broadcaster Other terms for stream, streamer.
Buffering A condition that occurs when the playback device is unable to download the content before the content is supposed to be played. Buffering can manifest in several ways: content may randomly stop and start (also known as stuttering), content may stop for long periods of time (also known as freezing), or the IVS player may pause playback.
Byte-range playlist

A more granular playlist than the standard HLS playlist. The standard HLS playlist is made up of 10-second media files. With a byte-range playlist, the segment duration is the same as the keyframe interval configured for the stream.

Byte-range playlist is available only for the broadcasts that were auto-recorded to an S3 bucket. It is created in addition to the HLS playlist. See Byte-Range Playlists in Auto-Record to Amazon S3 (Low-Latency Streaming).

CBR Constant Bitrate, a rate-control method for encoders that maintains a consistent bitrate throughout the entire playback of a video, regardless of what is happening during the broadcast. Lulls in the action may be padded to achieve the desired bitrate, and peaks may be quantized by adjusting the quality of encoding to match the target bitrate. We strongly recommend using CBR instead of VBR.
CDN Content Delivery Network or Content Distribution Network, a geographically distributed solution that optimizes delivery of content such as streaming video by bringing it closer to where users are located.
Channel An IVS resource that stores configuration for streaming, including an ingest server, a stream key, a playback URL, and recording options. Streamers use the stream key associated with a channel to start a broadcast. All metrics and events generated during a broadcast are associated with a channel resource.
Channel type Determines the allowable resolution and frame rate for the channel. See Channel Types in the IVS Low-Latency Streaming API Reference.
Chat logging An advanced option that can be enabled by associating a logging configuration with a chat room.
Chat room An IVS resource that stores configuration for a chat session, including optional features such as Message Review Handler and Chat Logging. See Step 2: Create a Chat Room in Getting Started with IVS Chat.
Client-side composition

Uses a host device to mix audio and video streams from stage participants and then sends them as a composite stream to an IVS channel. This allows more control over the look of the composition at the cost of higher utilization of client resources and a higher risk of a stage or a host issue impacting the viewers.

Also see server-side composition.

CloudFront A CDN service provided by Amazon.
CloudTrail An AWS service for collecting, monitoring, analyzing, and retaining events and account activity from AWS and external sources. See Logging IVS API Calls with AWS CloudTrail.
CloudWatch An AWS service for monitoring applications, responding to performance changes, optimizing resource use, and providing insights into operational health. You can use CloudWatch to monitor IVS metrics; see Monitoring IVS Real-Time Streaming and Monitoring IVS Low-Latency Streaming.
Composition The process of combining audio and video streams from multiple sources into a single stream.
Composition pipeline A sequence of processing steps required to combine multiple streams and encode the resulting stream.
Compression Encoding of information using fewer bits than the original representation. Any particular compression is either lossless or lossy. Lossless compression reduces bits by identifying and eliminating statistical redundancy. No information is lost in lossless compression. Lossy compression reduces bits by removing unnecessary or less important information.
Control plane Stores information about IVS resources such as channels, stages, or chat rooms and provides interfaces for creating and managing these resources. It is regional (based on AWS regions).
CORS Cross-Origin Resource Sharing, an AWS feature that allows client web applications that are loaded in one domain to interact with resources such as S3 buckets in a different domain. Access can be configured based on headers, HTTP methods, and origin domains. See Using cross-origin resource sharing (CORS) - Amazon Simple Storage Service in Amazon Simple Storage Service User Guide.
Custom image source An interface provided by the IVS Broadcast SDK that allows an application to provide its own image input instead of being limited to the preset cameras.
Data plane The infrastructure that carries data from ingest to egress. It operates based on the configuration managed in the control plane and is not restricted to an AWS region.
Encoder, encoding The process of converting video and audio content into a digital format, suitable for streaming. Encoding can be hardware or software based.
Event An automatic notification published by IVS to the AmazonEventBridge monitoring service. An event represents a state or health change of a streaming resource such as a stage or a composition pipeline. See Using Amazon EventBridge with IVS Low-Latency Streaming and Using Amazon EventBridge with IVS Real-Time Streaming.
FFmpeg A free and open-source software project consisting of a suite of libraries and programs for handling video and audio files and streams. FFmpeg provides a cross-platform solution to record, convert and stream audio and video.
Fragmented stream

Created when a broadcast disconnects and then reconnects within the interval specified in the channel’s recording configuration. The resulting multiple streams are considered a single broadcast and merged together into a single recorded stream. See Merge Fragmented Streams in Auto-Record to Amazon S3 (Low-Latency Streaming).

Frame rate A streaming metric for the number of video frames transmitted or received per second.
HLS HTTP Live Streaming (HLS), an HTTP-based adaptive bitrate streaming communications protocol used to deliver IVS streams to viewers.
HLS playlist A list of media segments that make up a stream. Standard HLS playlists are made up of 10-second media files. HLS also supports more granular byte-range playlists.
Host A real-time event participant who sends video and/or audio to the stage.
IAM Identity and Access Management, an AWS service that allows users to securely manage identities and access to AWS services and resources, including IVS.
Ingest IVS process for receiving video streams from a host or broadcaster for processing or delivery to viewers or other participants.
Ingest server

Receives video streams and delivers them to a transcoding system, where streams are transmuxed or transcoded into HLS for delivery to viewers.

Ingest servers are specific IVS components that receive streams for channels, along with an ingestion protocol (RTMP, RTMPS). See the information on creating a channel in Getting Started with IVS Low-Latency Streaming.

Interlaced video Transmits and displays only odd or even lines of subsequent frames to create perceived doubling of frame rate without consuming extra bandwidth. We do not recommend using interlaced video due to the video quality concerns.
JSON JavaScript Object Notation, an open-standard file format that uses human-readable text to transmit data objects consisting of attribute-value pairs and array data types or other serializable values.
Keyframe, delta frame, keyframe interval The keyframe (also referred to as intra-coded or i-frame) is a full frame of the image in a video. Subsequent frames, the delta frames (also referred to as predicted or p-frames), only contain the information that has changed. Keyframes will appear multiple times within a stream, depending on the keyframe interval defined in the encoder.
Lambda An AWS service for running code (referred to as Lambda functions) without provisioning any server infrastructure. Lambda functions can run in response to events and invocation requests, or based on a schedule. For example, IVS Chat uses Lambda functions to enable message review for a chat room.
Latency, glass-to-glass latency

A delay in data transfer. IVS defines latency ranges as:

  • Low latency: under 3 sec

  • Real-time latency: under 300 ms

Glass-to-glass latency refers to the delay from when a camera captures a live stream to when the stream appears on a viewer’s screen.

Layered encoding with simulcast Enables simultaneous encoding and publishing of multiple video streams with different quality levels. See Adaptive Streaming: Layered Encoding with Simulcast in Real-Time Streaming Optimizations.
Message review handler Enables IVS Chat customers to automatically review/filter user chat messages before they are delivered to the chat room. It is enabled by associating a Lambda function with a chat room. See Creating a Lambda Function in Chat Message Review Handler.
Mixer A feature of the IVS Mobile Broadcast SDKs that takes multiple audio and video sources and generates a single output. It supports management of on-screen video and audio elements representing sources such as cameras, microphones, screen captures, and audio and video generated by the application. The output can then be streamed to IVS. See Configuring a Broadcast Session for Mixing in IVS Broadcast SDK: Mixer Guide (Low-Latency Streaming).
Multi-host streaming

Combines streams from multiple hosts into a single stream. This can be accomplished using either client-side or server-side composition.

Multi-host streaming enables scenarios such as inviting viewers onto a stage for Q&A, competitions between hosts, video chat, and hosts conversing with each other in front of a large audience.

Multivariant playlist An index of all the variant streams available for a broadcast.
OAC Origin Access Control, a mechanism for restricting access to an S3 bucket, so that content such as a recorded stream can be served only through CloudFront CDN.
OBS Open Broadcaster Software, free and open source software for video recording and live streaming. OBS offers an alternative (to the IVS broadcast SDK) for desktop publishing. More sophisticated streamers familiar with OBS may prefer it for its advanced production features, such as scene transitions, audio mixing, and overlay graphics.
Participant A real-time user connected to a stage as a host or viewer.
Participant token Authenticates a real-time event participant when they join a stage. A participant token also controls whether a participant can send video to the stage.
Playback token, playback key pair

An authorization mechanism that allows customers to restrict video playback on private channels. Playback tokens are generated from a playback key pair.

A playback key pair is the public-private pair of keys used to sign and validate the viewer authorization token for playback. See Create or Import a Playback Key in Setting up Private Channels and see the Playback Key Pair endpoints in the IVS Low-Latency API Reference.

Playback URL Identifies the address a viewer uses to start playback for a specific channel. This address can be used globally. IVS automatically selects the best location on the IVS global content delivery network for delivering the video to each viewer. See the information on creating a channel in Getting Started with IVS Low-Latency Streaming.
Private channel Allows customers to restrict access to their streams using an authorization mechanism based on playback tokens. See Workflow for Private Channels in Setting up Private Channels.
Progressive video Transmits and displays all lines of each frame in sequence. We recommend using progressive video during all stages of a broadcast.

The maximum numbers of IVS service resources or operations for your AWS account. That is, these limits are per AWS account, unless noted otherwise. All quotas are enforced per region. See Amazon Interactive Video Service endpoints and quotas in AWS General Reference Guide.


Provide access to AWS services that physically reside in a specific geographic area. Regions provide fault tolerance, stability, and resilience, and can also reduce latency. With Regions, you can create redundant resources that remain available and unaffected by a regional outage.

Most AWS service requests are associated with a particular geographic region. The resources that you create in one region do not exist in any other region unless you explicitly use a replication feature offered by an AWS service. For example, Amazon S3 supports cross-region replication. Some services, such as IAM, do not have cross-regional resources.

Resolution Describes the number of pixels in a single video frame, for example, Full HD or 1080p defines a frame with 1920x1080 pixels.
Root user The owner of an AWS account. The root user has complete access to all AWS services and resources in the AWS account.
RTMP, RTMPS Real-Time Messaging Protocol, an industry standard for transmitting audio, video, and data over a network. RTMPS is the secure version of RTMP, running over a Transport Layer Security (TLS/SSL) connection.
S3 bucket A collection of objects stored in Amazon S3. Many policies, including access and replication, are defined at the bucket level and apply to all objects in the bucket. For example, an IVS broadcast is stored as multiple objects in an S3 bucket.
SDK Software Development Kit, a collection of libraries for the developers building applications with IVS.
Selfie segmentation Enables replacing the background in a live stream, using a client-specific solution that accepts a camera image as input and returns a mask that provides a confidence score for each pixel of the image, indicating whether it is in the foreground or the background. See Background Replacement in IVS Broadcast SDK: Third-Party Camera Filters (Real-Time Streaming).
Semantic versioning A version format in the form of Major.Minor.Patch. Bug fixes not affecting the API increment the patch version, backward compatible API additions/changes increment the minor version, and backward incompatible API changes increment the major version.
Server-side composition

Uses an IVS server to mix audio and video from stage participants and then sends this mixed video to an IVS channel to reach a larger audience or to store it in an S3 bucket. Server-side composition reduces client load, improves resilience of the broadcast, and enables more efficient use of bandwidth.

Also see client-side composition.

Service quotas An AWS service that helps you manage your quotas for many AWS services from one location. Along with looking up the quota values, you can also request a quota increase from the Service Quotas console.
Service-linked role A unique type of IAM role that is linked directly to an AWS service. Service-linked roles are automatically created by IVS and include all the permissions that the service requires to call other AWS services on your behalf, for example, to access an S3 bucket. See Using Service-Linked Roles for IVS in IVS Security.
Stage An IVS resource that represents a virtual space where real-time event participants can exchange video in real time. See Create a Stage in Getting Started with IVS Real-Time Streaming.
Stage session Begins when the first participant joins a stage and ends a few minutes after the last participant stops publishing to the stage. A long-lived stage may have multiple sessions over its lifetime.
Stream Data representing video or audio content being sent continuously from a source to a destination.
Stream key An identifier assigned by IVS when you create a channel; it is used to authorize streaming to the channel. Treat the stream key like a secret, since anyone with it can stream to the channel. See Getting Started with IVS Low-Latency Streaming.
Stream starvation

A delay or halt in stream delivery to IVS. It occurs when IVS does not receive the expected amount of bits that the encoding device advertised it would send over a certain timeframe. An occurrence of stream starvation results in a stream starvation event.

From a viewer's perspective, stream starvation may appear as video that lags, buffers, or freezes. Stream starvation can be brief (less than 5 seconds) or long (several minutes), depending on the specific situation that resulted in stream starvation. See What is Stream Starvation in Troubleshooting FAQ.

Streamer A person or a device sending a video or audio stream to IVS.
Subscriber A real-time event participant who receives video and/or audio of the hosts. See What is IVS Real-Time Streaming.
Tag A metadata label that you assign to an AWS resource. Tags can help you identify and organize your AWS resources. On the IVS documentation landing page, see “Tagging” in any of the IVS API documentation (for real-time streaming, low-latency streaming, or chat).
Third-party camera filters Software components that can be integrated with the IVS Broadcast SDK to allow an application to process images before providing them to the Broadcast SDK as a custom image source. A third-party camera filter may process images from the camera, apply a filter effect, etc.
Thumbnail A reduced-size image taken from a stream. By default, thumbnails are generated every 60 seconds, but a shorter interval can be configured. Thumbnail resolution depends on the channel type. See Recording Contents in Auto-Record to Amazon S3 (Low-Latency Streaming).
Timed metadata

Metadata tied to specific timestamps within a stream. It can be added programmatically using the IVS API and becomes associated with specific frames. This ensures that all viewers receive the metadata at the same point relative to the stream.

Timed metadata can be used to trigger actions on the client such as updating team statistics during a sporting event. See Embedding Metadata within a Video Stream.

Transcoding Converts video and audio from one format to another. An incoming stream may be transcoded to a different format at multiple bitrates and resolutions, to support a range of playback devices and network conditions.
Transmuxing A simple repackaging of an ingested stream to IVS, with no re-encoding of the video stream. “Transmux” is short for transcode multiplexing, a process that changes the format of an audio and/or video file while keeping some or all of the original streams. Transmuxing converts to a different container format without changing the file contents. Distinguished from transcoding.
Variant streams

A set of encodings of the same broadcast in several distinct quality levels. Each variant stream is encoded as a separate HLS playlist. An index of the available variant streams is referred to as a multivariant playlist.

After the IVS player receives a multivariant playlist from IVS, it can then choose between the variant streams during playback, changing back and forth seamlessly as network conditions change.

VBR Variable Bitrate, a rate-control method for encoders that uses a dynamic bitrate that changes throughout playback, depending on the level of detail needed. We strongly recommend against using VBR due to video-quality concerns; use CBR instead.

A unique viewing session which is actively downloading or playing video. Views are the basis for the concurrent views quota.

A view starts when a viewing session begins video playback. A view ends when a viewing session stops video playback. Playback is the sole indicator of viewership; engagement heuristics such as audio levels, browser tab focus, and video quality are not considered. When counting views, IVS does not consider the legitimacy of individual viewers or try to deduplicate localized viewership, such as multiple video players on a single machine. See Other Quotas in Service Quotas (Low-Latency Streaming).

Viewer A person receiving a stream from IVS.

Web Real-Time Communication, an open-source project providing web browsers and mobile applications with real-time communication. It allows audio and video communication to work inside web pages by allowing direct peer-to-peer communication, eliminating the need to install plugins or download native apps.

The technologies behind WebRTC are implemented as an open web standard and are available as regular JavaScript APIs in all major browsers or as libraries for native clients, like Android and iOS.


WebRTC-HTTP Ingestion Protocol, an HTTP based protocol that allows WebRTC based ingestion of content into streaming services and/or CDNs. WHIP is an IETF draft developed to standardize WebRTC ingestion.

WHIP enables compatibility with software like OBS, offering an alternative (to the IVS broadcast SDK) for desktop publishing. More sophisticated streamers familiar with OBS may prefer it for its advanced production features, such as scene transitions, audio mixing, and overlay graphics

WHIP is also beneficial in situations where using the IVS broadcast SDK isn't feasible or preferred. For example, in setups involving hardware encoders, the IVS broadcast SDK might not be an option. However, if the encoder supports WHIP, you can still publish directly from the encoder to IVS.

See OBS and WHIP Support.

WSS WebSocket Secure, a protocol for establishing WebSockets over an encrypted TLS connection. It is being used for connecting to IVS Chat endpoints. See Step 4: Send and Receive Your First Message in Getting Started with IVS Chat.