A comprehensive technical research report covering Agora.io's SD-RTN architecture, Unity/Unreal Engine integration, high-frame-rate screen sharing, AI noise suppression, and cost analysis for building esports team video conferencing systems.
Agora.io Video Technology Architecture Deep Research Report: Implementation Plan for Esports Team Collaboration Systems
1. Executive Summary and Project Background
1.1 Research Background and Objectives
With the professionalization of the esports industry and the digital transformation of team management systems, real-time audio and video communication (RTC) has become the core infrastructure for team training, match reviews, and tactical command. This technical research report aims to respond to the technical requirements of "developing video conferencing functionality for gaming teams," providing exhaustive architectural analysis and feasibility verification of the video SDK and related real-time engagement (RTE) technologies provided by Agora.io.
The core objective of this research is to build a highly available, low-latency video conferencing system that can be deeply integrated into game clients (Unity/Unreal Engine). This system must not only meet basic video calling needs but also address the special challenges of gaming scenarios, including:
* **Extreme low-latency requirements**: In tactical command for real-time strategy (RTS) or first-person shooter (FPS) games, millisecond-level latency differences can cause tactical execution disconnection.
* **Non-invasive game performance**: As an auxiliary system, the video conferencing function must never compete for CPU/GPU resources from the game's core rendering loop and must ensure game frame rate (FPS) stability.
* **High-fidelity screen sharing**: Support for 60fps or higher frame rate game screen sharing to enable coaching staff to perform real-time frame-by-frame tactical analysis.
* **Complex audio environment handling**: Effective suppression of mechanical keyboard sounds and environmental noise while avoiding conflicts with in-game 3D sound effects (Ducking/Mixing).1.2 Technology Selection Overview
After in-depth analysis of Agora documentation and technical features, this report concludes that Agora.io's SD-RTN™ (Software-Defined Real-Time Network) architecture has significant advantages in handling cross-border and cross-regional weak network transmission. Compared to traditional WebRTC P2P solutions or CDN-based streaming solutions, Agora's proprietary network coverage and gaming-optimized SDKs (especially Unity and Unreal Engine plugins) can significantly reduce development costs and improve end-user experience.
This report will provide a comprehensive technical implementation guide of approximately 15,000 words from multiple dimensions including network architecture, core audio/video functionality, game engine integration practices, signaling system design, security architecture, and cost analysis.
2. Core Network Infrastructure: SD-RTN™ and Latency Architecture Analysis
2.1 Software-Defined Real-Time Network (SD-RTN™) Technical Principles
Traditional real-time communication often relies on the public internet's best-effort transmission mode, where packet routing paths are typically determined by BGP protocol, which tends to minimize hop counts rather than optimize latency or packet loss rates. For team members scattered across different physical locations (such as headquarters, dormitories, or even international tournament venues), this transmission method can easily cause "long-distance hairpin turn" routing, leading to high latency and jitter.
Agora's core competitiveness lies in its SD-RTN™ (Software-Defined Real-Time Network). This is not just a transmission protocol but a virtual overlay network of data centers covering more than 200 countries and regions worldwide.
2.1.1 Dynamic Routing and Intelligent Scheduling
SD-RTN™ acts as a "virtual private channel" on top of the public internet. When team members join a channel, the SDK doesn't connect directly to the other party's IP but connects to the nearest Agora edge node.
* **Mechanism Analysis**: Control nodes in the network monitor link quality across the entire network in real-time (including packet loss rate, RTT round-trip delay, jitter). Based on this real-time data, intelligent routing algorithms dynamically calculate the optimal path from sender to receiver.
* **Gaming Scenario Application**: Suppose a team coach is in Seoul, South Korea, while players are at a European training base. Traditional cross-continental internet transmission latency could be as high as 300-500ms and extremely unstable. SD-RTN™ can plan an optimized backbone transmission path, controlling end-to-end latency to around 200ms and significantly reducing stuttering caused by packet loss.2.1.2 Weak Network Resilience and Packet Loss Resistance
Esports team members' network environments are complex and variable (such as Wi-Fi signal interference, mobile network access). Agora has implemented highly customized UDP protocol optimization at the transport layer.
- Anti-packet loss technology: Through a hybrid mechanism of Forward Error Correction (FEC) and Automatic Repeat Request (ARQ), SD-RTN™ can maintain call continuity even with video packet loss rates up to 70% and audio packet loss rates up to 80%. Although image quality will degrade (blurry or blocky) under high packet loss, "voice command deliverability" is guaranteed, which is crucial for tactical command.
2.2 Latency Benchmarks and Product Selection
In Agora's product system, there are two main forms: Video Calling and Interactive Live Streaming. For gaming team meeting functionality, accurate product selection directly determines system success or failure.
| Feature Dimension | Video Calling | Interactive Live Streaming - Standard/Ultra-low Latency |
|---|---|---|
| Design Purpose | Emphasizes point-to-point or small group real-time bidirectional interaction, all users equal | Emphasizes "broadcaster-audience" model, suitable for large-scale distribution |
| End-to-End Latency | < 200ms - 400ms (global average) | Ultra-low: 400ms-800ms; Standard: 1500ms-2000ms |
| Role Management | Default "communication" mode, no on/off mic operations needed, speak anytime | Users must distinguish between Host and Audience, speaking requires signaling role switch |
| Applicable Scenarios | Team internal voice/video meetings, tactical discussions, reviews | Tournament external broadcasts, fan meetups, public lectures |
Technical Decision: For internal team collaboration, Video Calling product line must be selected, with channel scenario (Channel Profile) set to COMMUNICATION during SDK initialization.
* **Reason One**: Tactical discussions involve high-frequency bidirectional interaction. Players will interrupt coaches at any time to ask questions or sync information. Role Switching Latency and high audience-side latency in interactive live streaming mode are unacceptable.
* **Reason Two**: In COMMUNICATION mode, the audio engine enables aggressive modes of echo cancellation (AEC) and automatic gain control (AGC), prioritizing voice clarity over background music quality, which better suits meeting needs.3. Game Engine Integration Deep Guide: Unity
Unity is currently one of the most mainstream engines in game development. The biggest challenge in integrating Agora Video SDK into Unity projects lies in managing the interaction between Android/iOS native plugins and Unity managed code (C#), as well as handling render thread synchronization.
3.1 SDK Architecture and Lifecycle Management
Agora Unity SDK is a C# wrapper around the native C++ SDK. In Unity, all API calls (such as JoinChannel) are ultimately passed to the underlying .so (Android) or .framework (iOS) libraries through P/Invoke mechanism.
3.1.1 Initialization and Singleton Pattern
Throughout the game's entire lifecycle, the IRtcEngine instance should exist as a global singleton. Frequently creating and destroying engines will bring huge CPU overhead and memory churn.
* **Recommended Implementation**: Create an AgoraManager class that inherits from MonoBehaviour and is set to DontDestroyOnLoad.
* **Event Callbacks**: IRtcEngineEventHandler is the core for handling all asynchronous events (such as user joining, offline, network quality callbacks). In C#, this is typically implemented through the Delegate pattern. Note that callback functions are usually triggered on non-main threads (Background Thread), so when updating UI in callbacks (such as displaying prompt text), you **must** use UnityMainThreadDispatcher or a simple task queue to marshal operations to the main thread, otherwise Unity will crash or become unresponsive.3.2 Video Rendering Pipeline Optimization
Rendering video streams in Unity (especially multiple 1080p streams) is very resource-intensive. Agora provides the VideoSurface.cs component to simplify this process, but for performance-critical games, we need to understand its underlying mechanism and optimize.
3.2.1 Texture Update Mechanism
VideoSurface's working principle is: underlying SDK decodes YUV data -> converts to RGB -> updates to Unity's Texture2D -> applies to RawImage or MeshRenderer.
* **Performance Bottleneck**: Per-frame texture upload (CPU to GPU) is the main bottleneck. If 5 people have cameras on in a team meeting, there are 5 texture upload operations.
* **Optimization Strategies**:
* **Resolution Control**: In the in-game overlay interface, teammate camera views are usually small. Force the received stream resolution to 320x180 or lower (request low stream via SetRemoteVideoStreamType), only switching to high stream when viewing someone fullscreen.
* **Frame Rate Limiting**: For non-game camera streams, human eye fluency perception threshold is lower. Video encoding frame rate can be limited to 15fps or 24fps to reduce rendering pressure and preserve more GPU time slices for core game rendering logic.3.2.2 Mobile Thermal and Battery Control
Long video conferences cause mobile devices to heat up, triggering system-level Thermal Throttling and causing game frame drops.
* **Configuration Recommendations**:
* In SetVideoEncoderConfiguration, set degradationPreference to MAINTAIN_FRAMERATE. This means when CPU/network load is too high, the SDK will automatically reduce video clarity to maintain frame rate and prevent stuttering.
* For Unity projects, it's recommended to completely disable the video module via MuteLocalVideoStream(true) and EnableLocalVideo(false) during non-critical moments (such as voice-only communication), which can significantly reduce power consumption.3.3 Code Obfuscation and Packaging
When building Android versions, Unity typically enables ProGuard or R8 for code obfuscation. If Agora's core classes are obfuscated, JNI calls will fail, causing UnsatisfiedLinkError.
- Implementation Details: You must add the following rules in proguard-user.txt or proguard-rules.pro: -keep class io.agora.** { ; } -dontwarn io.agora.*
4. Game Engine Integration Deep Guide: Unreal Engine
Unreal Engine (UE) is known for its high-performance C++ architecture and is commonly used for large PC and console games. Agora Unreal SDK integration is more flexible but also more complex.
4.1 C++ vs Blueprints Decision
Agora Unreal SDK supports both C++ and Blueprints.
* **Blueprints**: Suitable for rapid prototyping and UI logic binding (such as button click events). For simple "click to join meeting" functionality, Blueprints are sufficient and highly efficient.
* **C++**: For team systems, core logic should be implemented in C++.
* **Reason**: C++ can directly access underlying memory and pointers, handling Raw Data (raw audio/video data) with much higher efficiency than Blueprint virtual machine. Additionally, C++ has advantages in handling complex multi-threading logic and game state synchronization, and is easier for version control and merging (Diff/Merge).4.2 Frame Rate Synchronization and Rendering Stutter Issues
Unreal Engine by default dynamically adjusts frame rate based on load, or locks at 60fps. Video SDK rendering often depends on engine Tick events.
* **Smooth Frame Rate Trap**: UE's bSmoothFrameRate feature sometimes conflicts with Agora's video capture frame rate, causing slight video jitter or delay.
* **Solutions**:
* In DefaultEngine.ini or via console commands, ensure bSmoothFrameRate is configured correctly, or unlock frame rate limits for the Level containing the video conferencing module.
* For laptop devices, NVIDIA's Battery Boost feature may force games to lock at 30fps when unplugged, severely affecting 60fps screen sharing experience. The application layer should detect power state and prompt users to adjust settings, or try using r.DontLimitOnBattery 1 command to bypass the limitation.4.3 Modular Integration
To keep projects clean, it's recommended to encapsulate Agora functionality as an independent UE Plugin or Module.
* **Dependency Configuration**: In Project.Build.cs, you must correctly add AgoraPlugin and AgoraBlueprintable to PrivateDependencyModuleNames.
* **Android Permissions**: UE's Android permission management requires explicitly adding android.permission.CAMERA, android.permission.RECORD_AUDIO, android.permission.INTERNET, etc. in Project Settings, otherwise the packaged APK will crash at runtime.5. Tactical-Level Screen Sharing: High Frame Rate and Dual Stream Architecture
For teams, screen sharing is not just "viewing PPTs" but requires high-fidelity transmission of 60fps game footage for review.
5.1 High Frame Rate Screen Sharing Configuration
Ordinary meeting software screen sharing is typically at 5-15fps, which is disastrous for FPS game reviews.
* **Parameter Configuration**: Using ScreenCaptureParameters structure, you must explicitly set:
* dimensions: 1920x1080 (or 2560x1440)
* frameRate: **60** (critical)
* bitrate: Recommend setting to 2-3x system default, or use STANDARD_BITRATE with high quality policy.
* **Bandwidth Consumption**: 1080p60fps video stream bitrate is typically 3Mbps - 6Mbps. The system must check uplink bandwidth before enabling; if bandwidth is insufficient, force resolution reduction rather than frame rate reduction, because for tactical analysis, fluidity (seeing action continuity) is usually more important than static clarity.5.2 Dual Stream Architecture Design
When coaches or commanders review, they typically need to display both "game footage" and "their own facial expressions/gestures" simultaneously. Agora SDK allows pushing two video streams simultaneously through a single instance on desktop (Windows/macOS), or through auxiliary processes on mobile.
* **Implementation Plan**:
* **Main Stream (Camera)**: Push camera capture data, using default UID (e.g., 1001).
* **Secondary Stream (Screen)**: Call StartScreenCaptureByDisplayId or StartScreenCaptureByWindowId. To distinguish these two streams on the receiving end, the common approach is to assign a specific UID rule to screen sharing streams (e.g., MainUID + 10000).
* **Receiver Processing**: In OnUserJoined callback, client checks UID. If UID > 10000, render to "large screen/tactical board" area; otherwise render to "avatar/camera" area. This logical distinction is crucial for user experience.6. Immersive Game Audio Architecture
Audio is the soul of game experience and also the difficulty of meeting systems. How to achieve clear voice communication without interfering with game sound effects?
6.1 Audio Scenario Selection
This is the most critical API setting. Most integration failures are due to selecting the wrong Audio Scenario.
* **Recommended Setting**: Must use AUDIO_SCENARIO_GAME_STREAMING.
* **Features**: This mode is specifically designed for games - it doesn't duck background game sound effect volume and supports high-fidelity audio capture.
* **Comparison**: Using AUDIO_SCENARIO_DEFAULT or CHATROOM on mobile may force switching to call volume (earpiece output), causing game sound effects to become thin with severe mono sensation.6.2 AI Noise Suppression
Team bases are filled with mechanical keyboard (blue/brown switch) clicking sounds and mouse clicks - high-frequency noise that severely interferes with voice communication.
* **Agora Solution**: Integrate agora-extension-ai-denoiser plugin.
* **Technical Principle**: This plugin uses deep learning models specifically targeting non-human voice stationary and transient noise (such as keyboard, applause, construction sounds) for spectral-level elimination while preserving human voice waveform characteristics.
* **Performance Tradeoff**: AI noise suppression consumes extra CPU resources. This is usually not a problem for PC, but on mobile, Balance mode is recommended over Aggressive mode to avoid competing with games for CPU resources.6.3 3D Spatial Audio
To enhance presence, spatial audio technology can be introduced so teammates' voices seem to come from specific directions in the game.
* **Implementation Logic**: Using ILocalSpatialAudioEngine, call UpdateSelfPosition (update local player coordinates and orientation) and UpdateRemotePosition (update teammate coordinates) each frame.
* **Algorithm**: The SDK calculates Head-Related Transfer Function (HRTF) based on coordinate differences, achieving sound localization. This is extremely valuable in "virtual training room" or RPG-type team social scenarios.7. Signaling and Collaboration: RTM SDK vs Chat SDK
Besides audio/video, team systems also need text chat, online invitations, global mute, and other control signals. Agora provides Signaling (formerly RTM) and Chat (formerly Hyphenate) solutions.
7.1 Deep Comparison and Selection
| Dimension | Signaling (RTM) SDK | Chat SDK | Team System Compatibility Analysis |
|---|---|---|---|
| Positioning | Low-latency signaling, state sync, instant messages | Complete IM solution | |
| Message Persistence | Primarily ephemeral. Beta supports limited history | Strong persistence. Supports roaming, fetching history from server | Teams need to review historical tactical commands, Chat wins |
| Rich Media Support | Only small files/text, requires self-handling upload/download | Native support for image, voice, video, file messages | Teams often need to send game screenshots/recordings, Chat wins |
| Offline Messages | Weak support | Native support for offline push and storage | Players can receive notifications while offline, Chat wins |
| Latency | Very low (< 100ms) | Low (200-400ms) | Video meeting instant interaction (like raising hands) uses RTM; regular chat uses Chat |
Architecture Recommendation: Adopt hybrid architecture.
* **Chat SDK**: For building team main chat groups, friend private chats, history storage, rich media (screenshots/demo files) sending. This is the team's "social hall."
* **Signaling (RTM) / Data Stream**: During video meetings, use Video SDK's built-in data stream or lightweight RTM to transmit extremely real-time control signals (like: captain forcing global mute, marking map coordinates). This ensures high synchronization between control signals and video streams.8. Security Architecture and Token Authentication
In professional esports, preventing DDoS attacks and "stream sniping" is a security red line.
8.1 Token Authentication Mechanism
Agora enforces Token authentication and prohibits storing App Certificate on clients.
- Generation Flow:
- Client sends POST /login request to business backend.
- Business backend verifies user identity (Session/JWT).
- Business backend calls Agora-provided buildTokenWithUid algorithm, combining AppID, AppCertificate, ChannelName, UID, Role to generate Token.
- Token returns to client, client uses this Token to join channel.
- Validity Management: Token validity is recommended to be 24 hours. Client must listen to onTokenPrivilegeWillExpire callback, requesting new Token from business backend before expiration (e.g., 30 seconds ahead) and updating via RenewToken to ensure long training sessions aren't interrupted.
8.2 Permission Control
- Role Differentiation: When generating Tokens on backend, user permissions can be specified. For example, only "coaches" and "captains" have Publisher permission (can push streams), while regular substitute players only have Subscriber permission (view only). This fundamentally prevents unauthorized personnel from disrupting meeting order.
9. Cost Model and Budget Analysis
Agora uses a "Pay-as-you-go" model. Accurate cost estimation is crucial for commercial operations.
9.1 Core Billing Elements
* **Video HD (720p and below)**: Approximately $3.99 / 1,000 minutes.
* **Video Full HD (1080p and above)**: Approximately $8.99 / 1,000 minutes. **Note**: If 1080p screen sharing is used, that user's time will be billed at Full HD rate.
* **Audio**: $0.99 / 1,000 minutes. If users turn off camera and keep voice only, system automatically downgrades to audio billing.
* **AI Noise Suppression Plugin**: Additional $0.59 / 1,000 minutes.9.2 Scenario-Based Cost Calculation
Assume a team has 1 coach and 5 starting players. Conducting a 2-hour (120-minute) review meeting.
* **Configuration**: Coach enables 1080p screen sharing + voice; 5 players enable 720p camera + voice; all enable AI noise suppression.
* **Coach Cost** (Full HD + AI): 120 minutes × ($8.99 + $0.59) / 1000 = **$1.15**.
* **Player Cost** (HD + AI): 5 people × 120 minutes × ($3.99 + $0.59) / 1000 = **$2.75**.
* **Total Single Meeting Cost**: **$3.90**.Business Insight: While unit prices seem low, costs grow linearly as team count and training duration increase. Chat SDK uses monthly active user (MAU) billing ($349/month starting, includes 5000 MAU), which may not be cost-effective for professional team apps with small user counts but high-frequency usage. Consider whether to use RTM only (billed by message volume/daily active users) to replace some functionality.
10. Competitive Analysis and Technical Barriers
10.1 Why Not Develop WebRTC In-House?
Many teams initially consider self-building based on open-source WebRTC (like Jitsi, Mediasoup).
- Disadvantages: Self-built WebRTC cannot solve cross-border transmission latency issues (lacking global private network like SD-RTN). In esports scenarios, network fluctuation causing stuttering can interrupt review rhythm. Additionally, maintaining large-scale SFU cluster operations costs are extremely high.
10.2 Agora vs. ZEGOCLOUD / Stream
* **Stream.io**: Very strong in Chat functionality and UI components (UIKit), suitable for quickly building social apps. But in underlying network optimization for real-time audio/video and game engine support depth, slightly inferior to Agora.
* **ZEGOCLOUD**: Provides similar one-stop service with highly competitive pricing. But in global node coverage density and vertical optimization for game audio (Spatial Audio, AI Noise Suppression), Agora has a deeper technical moat.11. Summary and Recommendations
In summary, building a gaming team video conferencing system using Agora.io is highly technically feasible and can significantly shorten Time-to-Market compared to self-developed solutions.
Core Recommendation Checklist:
- Architecture: Firmly choose Video Calling product line + COMMUNICATION channel property.
- Engine: Unity side strictly controls VideoSurface rendering overhead; Unreal side pay attention to C++ module dependencies and frame rate synchronization.
- Audio: Must configure AUDIO_SCENARIO_GAME_STREAMING and integrate AI Noise Suppression extension.
- Features: Screen sharing must support 60fps high frame rate mode, using UID offset method for dual stream identification.
- Signaling: Build hybrid signaling network combining Chat SDK (social) and RTM (real-time control).
By following the architectural specifications and best practices detailed in this report, development teams can build a collaborative communication system that meets the stringent requirements of professional esports while having good scalability.
Works Cited
- Interactive Live Streaming Overview | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/3.x/interactive-live-streaming/introduction/product-overview
- WebRTC vs Agora: Key Differences for Real-Time Communication Apps - Digittrix Infotech, accessed January 4, 2026, https://www.digittrix.com/blogs/webrtc-vs-agora-key-differences
- Agora RTC Video SDK - Fab, accessed January 4, 2026, https://www.fab.com/listings/bcd83cbd-e6cf-42b9-9088-854833a1a083
- Agora vs Stream: which should you choose in 2025? - Ably Realtime, accessed January 4, 2026, https://ably.com/compare/agora-vs-stream
- Agora's Software Defined Real-Time Network Delivers Real-Time Internet Advantages Over Content Delivery Network, accessed January 4, 2026, https://hello.agora.io/rs/096-LBH-766/images/Agora_WP_SD-RTN-Delivers-RealTime-Internet-Advantages.pdf
- Agora.io Brings Live Video and Broadcasting Solutions with Global Scalability and Ultra Low Latency to IBC 2018 - PR Newswire, accessed January 4, 2026, https://www.prnewswire.com/news-releases/agoraio-brings-live-video-and-broadcasting-solutions-with-global-scalability-and-ultra-low-latency-to-ibc-2018-300712746.html
- ConvoAI Device Kit R1 Architecture and specifications | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/convo-ai-device-kit/overview/architecture
- Video Calling Quickstart | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/video-calling/get-started/get-started-sdk
- AgoraIO/video-sdk-samples-unity - GitHub, accessed January 4, 2026, https://github.com/AgoraIO/video-sdk-samples-unity
- Agora Video SDK for Unity Quick Start Programming Guide, accessed January 4, 2026, https://www.agora.io/en/blog/agora-video-sdk-for-unity-quick-start-programming-guide/
- Broadcast Streaming Configure video encoding | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/broadcast-streaming/enhance-call-quality/configure-video-encoding
- Everything You Need to Know about Agora Video SDK v4.5, accessed January 4, 2026, https://www.agora.io/en/blog/everything-you-need-to-know-about-agora-video-sdk-v4-5/
- Agora-Unreal-SDK-Blueprint/GUIDE.md at master - GitHub, accessed January 4, 2026, https://github.com/AgoraIO-Community/Agora-Unreal-SDK-Blueprint/blob/master/GUIDE.md
- Unreal Engine Blueprints vs. C++: Performance and Ease of Use Compared - Program-Ace, accessed January 4, 2026, https://program-ace.com/blog/unreal-engine-blueprints-vs-c/
- Coding in Unreal Engine: Blueprint vs. C++ - Epic Games Developers, accessed January 4, 2026, https://dev.epicgames.com/documentation/en-us/unreal-engine/coding-in-unreal-engine-blueprint-vs-cplusplus
- Solo devs. Which do you prefer? Blueprints or C++? - Unreal Engine Forum, accessed January 4, 2026, https://forums.unrealengine.com/t/solo-devs-which-do-you-prefer-blueprints-or-c/1706645
- Stuck at 60 FPS with smooth rate disabled - Unreal Engine Forums, accessed January 4, 2026, https://forums.unrealengine.com/t/stuck-at-60-fps-with-smooth-rate-disabled/398803
- Interactive Live Streaming Screen sharing | Agora Docs, accessed January 4, 2026, https://docs-staging.agora.io/en/interactive-live-streaming/advanced-features/screen-sharing?platform=blueprint
- How to Differentiate Screen Sharing and Camera Video in Agora Whiteboard SDK, accessed January 4, 2026, https://agoraio.zendesk.com/hc/en-us/articles/37624047203220-How-to-Differentiate-Screen-Sharing-and-Camera-Video-in-Agora-Whiteboard-SDK
- Video Calling Set the Audio Profile | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/3.x/video-calling/basic-features/audio-profiles
- Joining a channel on Agora (Unity/iOS) makes in-game music almost inaudible, accessed January 4, 2026, https://stackoverflow.com/questions/71438700/joining-a-channel-on-agora-unity-ios-makes-in-game-music-almost-inaudible
- AI Noise Suppression - Agora - SoftwareOne, accessed January 4, 2026, https://platform.softwareone.com/product/ai-noise-suppression/PCP-0522-6438
- Interactive Live Streaming AI Noise Suppression | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/interactive-live-streaming/advanced-features/ai-noise-suppression
- Video Calling 3D Spatial Audio | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/video-calling/advanced-features/spatial-audio
- video-sdk-samples-unity/Assets/spatial-audio/spatialAudioManager.cs at main - GitHub, accessed January 4, 2026, https://github.com/AgoraIO/video-sdk-samples-unity/blob/main/Assets/spatial-audio/spatialAudioManager.cs
- Signaling Message history (Beta) | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/signaling/core-functionality/message-history
- Chat Message overview | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/agora-chat/reference/message-overview
- Chat and Signaling comparison - Agora Documentation, accessed January 4, 2026, https://docs.agora.io/en/agora-chat/reference/comparison-chat-signalling
- Signaling Secure authentication with tokens | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/signaling/get-started/authentication-workflow
- Generating Authentication Token for Agora Applications - DEV Community, accessed January 4, 2026, https://dev.to/zolomohan/generating-authentication-token-for-agora-applications-1nij
- Video Calling Authenticate Your Users with Tokens | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/3.x/video-calling/basic-features/token-server
- Video Calling Pricing | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/video-calling/overview/pricing
- Video Calling Legacy pricing | Agora Docs, accessed January 4, 2026, https://docs.agora.io/en/video-calling/overview/pricing-legacy
- Pricing - Agora, accessed January 4, 2026, https://www.agora.io/en/pricing/
- Agora vs Twilio: which should you choose in 2025? - Ably Realtime, accessed January 4, 2026, https://ably.com/compare/agora-vs-twilio
- Chat Pricing - Agora, accessed January 4, 2026, https://www.agora.io/en/pricing/chat/
- Agora vs Stream: Which One is Better for You? - ZEGOCLOUD, accessed January 4, 2026, https://www.zegocloud.com/blog/agora-vs-stream
コメント
コメント (0)