railbird/react-native-vision-camera
10
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

try: Improvements from WWDC 2021 1:1 workshop (#197)

Browse Source 
* perf: Automatically determine Pixel Format depending on active format. (More efficient video recording 🚀)
* perf: Skip `AVAssetWriter` transform by directly correctly orienting the Video Output connection
* feat: Support camera flipping while recording
* feat: Run frame processor on separate queue, avoids stutters in video recordigns
* feat: Automatically drop late frame processor frames
This commit is contained in:
Marc Rousavy
2021-06-11 21:06:19 +02:00
committed by GitHub
parent 26cf21ff5f
commit 9c579c65aa
14 changed files with 103 additions and 74 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
ios/CameraQueues.swift
View File

@@ -17,13 +17,20 @@ public class CameraQueues: NSObject {
autoreleaseFrequency: .inherit,
target: nil)
/// The serial execution queue for output processing of videos as well as frame processors.
/// The serial execution queue for output processing of videos for recording.
@objc public static let videoQueue = DispatchQueue(label: "mrousavy/VisionCamera.video",
qos: .userInteractive,
attributes: [],
autoreleaseFrequency: .inherit,
target: nil)
/// The serial execution queue for output processing of videos for frame processing.
@objc public static let frameProcessorQueue = DispatchQueue(label: "mrousavy/VisionCamera.frame-processor",
qos: .userInteractive,
attributes: [],
autoreleaseFrequency: .inherit,
target: nil)
/// The serial execution queue for output processing of audio buffers.
@objc public static let audioQueue = DispatchQueue(label: "mrousavy/VisionCamera.audio",
qos: .userInteractive,