tencent cloud

Video on Demand

Audio and video Enhancement

ダウンロード
フォーカスモード
フォントサイズ
最終更新日: 2026-06-09 16:14:12

Introduction to Audio & Video Enhancement

Overview

Audio and video enhancement functions rely on Tencent Cloud's audio and video AI processing models and extensive accumulation of business data to provide professional-grade audio and video enhancement solutions. This feature supports distributed real-time image quality enhancement, including video deblocking, noise reduction, color enhancement, detail enhancement, face enhancement, SDR2HDR, large model enhancement, etc. It can significantly improve audio and video quality and is widely used in scenarios such as OTT, e-commerce, and sports events, effectively achieving the dual-dimensional improvement of QoE and QoS and creating significant business value.

Technical Advantages

All-Scenario AI Enhancement Algorithms. We offer industry-leading AI enhancement algorithms customized for diverse scenarios, including gaming, User-Generated Content (UGC), PGC high-definition film and television, online education, live showrooms, e-commerce, and legacy footage, comprehensively improving both audio and video quality.
Comprehensive Audio Enhancement. Supporting voice noise reduction, audio separation, sound quality enhancement, and volume equalization, this feature significantly boosts audio clarity and quality to meet the demands for high-quality audio across all types of scenarios.

Video Enhancement Capabilities

Ability
Description
Basic Image Quality Enhancement
Large Model Enhancement
Based on the Diffusion large model, it utilizes powerful AI generation capabilities to significantly improve video quality restoration. The results far exceed conventional methods, making it particularly suitable for repairing old videos.
Comprehensive Enhancement
Through AI's comprehensive analysis capabilities, it automatically balances the texture content in the frame, enhancing key details while removing compression artifacts and "jaggies," thereby improving the overall subjective perception of the entire image.
Artifact Removal Enhancement
Artifact removal technology analyzes coding information to intelligently remove aliasing, repair jaggies, blurriness, or unnatural colors, restoring clarity and naturalness to improve overall video quality.
Extended Enhancement Capabilities
Intelligent Frame Interpolation
Once enabled, if the set interpolation frame rate is higher than the source file's frame rate, the system will analyze motion between adjacent frames and intelligently generate intermediate frames to provide users with a smoother, silkier visual experience.
Super-Resolution
Super-resolution can identify the content and contours of a video to reconstruct high-definition details and local features, converting low-resolution video into high-resolution video. It is suitable for scenarios such as old film restoration.
HDR
Supports HDR10 and HLG, enabling a wider color gamut and displaying more color details to provide higher-quality video content.
Low-Light Enhancement
Due to environmental conditions or camera hardware limitations, footage shot in certain scenarios may lack brightness and contrast, resulting in dark images or missing details. Enabling low-light enhancement can significantly improve detail and contrast in dark areas, enhancing subjective visual quality.
Color Enhancement
Aimed at improving the color performance of the video to make the image closer to true-to-life colors and enhancing them to a certain extent to satisfy human visual preferences. It adjusts color saturation, contrast, and brightness to repair color distortion caused by capture equipment or storage issues, thereby improving the overall visual effect of the video.
Video Noise Reduction
Random noise can be introduced by the camera and environment during filming. The video noise reduction service can eliminate random noise in the frame while maintaining details without loss.
Scratch Removal
Scratch removal can repair damaged content such as scratches and "snow" spots in the video.

Audio Enhancement Capabilities

Ability
Description
Audio Noise Reduction
Uses intelligent algorithms to identify and eliminate background noise while preserving and enhancing vocals or primary audio signals, significantly improving audio clarity and the auditory experience.
Audio Separation
Separates vocals from background sounds or singing voices from accompaniment in audio/video files, facilitating further post-production processing.
Volume Equalization
Intelligently identifies and adjusts volume to prevent issues such as audio being too loud, too quiet, or sudden volume spikes, providing a better listening experience.
Audio Beautification
Intelligently enhances audio by removing ambient noise and suppressing unnatural sibilance or harsh, piercing sounds to improve overall audio quality.

Application Scenarios

Audio and video enhancement is applicable to business scenarios such as UGC/PGC video quality improvement, game live stream recording, old film restoration, and low-resolution super-resolution enhancement.
Scenario
Description
UGC/PGC Quality Improvement
Through face enhancement technology, while eliminating overall facial blur and compression artifacts, it further reconstructs key facial features—including eyes, mouth, ears, skin, and even hair strands—by adding details and textures, significantly increasing facial detail and realism.
Game Live Stream Recording
Transcoding is typically performed before recording a live stream to resolve file exceptions caused by stream interruptions. However, the compression during transcoding can lead to distortion and blur. This template primarily focuses on deblocking/decompression repair to restore image details and improve visual effects.
Old Film Restoration
Due to the technical limitations at the time of filming, some old movies may contain a large number of artifacts and scratches, resulting in poor image quality. Utilizing the repair and enhancement capabilities of "Image Quality Rebirth," old films are restored to give them a "new lease on life."
Low-Resolution Super-Resolution
Due to filming conditions or storage costs, some archived videos are stored at low resolutions. When playback is required on high-end display devices, simply transcoding low-resolution video makes it appear blurrier. By combining Cloud VOD "Image Quality Rebirth" Super-Resolution with low-quality repair and key detail enhancement, high image quality is ensured after the super-resolution process.

How to Use

1. Console: For detailed instructions, please refer to Audio & Video Enhancement.
2. Development Guide: For a comprehensive development guide, please refer to Audio & Video Enhancement.
2.1 Initiating Tasks: To initiate an audio/video enhancement task, please refer to Initiating Tasks.
2.2 Obtaining Results: To obtain the results of an enhancement task, please refer to Querying Task Details and Pulling Event Notifications.

Billing Details

Audio and video enhancement functions are implemented based on transcoding; specifically, enhancement processing involves overlaying enhancement parameters on top of the transcoding process. Therefore, when using enhancement features, you must simultaneously configure either Standard Transcoding or Top Speed Codec (TESC) parameters alongside the enhancement parameters. Both transcoding and enhancement fees will be charged.
1. Enhancement billing methods, please refer to Audio and Video Enhancement Billing.
2. For transcoding billing methods, please refer to Standard Transcoding Billing or Top Speed Codec Billing.

ヘルプとサポート

この記事はお役に立ちましたか?

フィードバック