`- (void)setWaterMark { GPUImageAlphaBlendFilter *blendFilter = [[GPUImageAlphaBlendFilter alloc] init]; blendFilter.mix = 1.0; NSDate *startTime = [NSDate date];

UILabel *timeLabel = [[UILabel alloc] initWithFrame:CGRectMake(0.0, 0.0, 240.0f, 320.0f)];
timeLabel.font = [UIFont systemFontOfSize:17.0f];
timeLabel.text = @"Time: 0.0 s";
timeLabel.textAlignment = NSTextAlignmentCenter;
timeLabel.backgroundColor = [UIColor clearColor];
timeLabel.textColor = [UIColor whiteColor];

GPUImageUIElement *uiElementInput = [[GPUImageUIElement alloc] initWithView:timeLabel];

[_filter addTarget:blendFilter];
[uiElementInput addTarget:blendFilter];

[blendFilter addTarget:_gpuImageView];

__unsafe_unretained GPUImageUIElement *weakUIElementInput = uiElementInput;
__weak typeof(self) _self = self;
[_filter setFrameProcessingCompletionBlock:^(GPUImageOutput * filter, CMTime frameTime){
    timeLabel.text = [NSString stringWithFormat:@"Time: %f s", -[startTime timeIntervalSinceNow]];
    [weakUIElementInput update];
    [_self processVideo:filter];
}];
[_videoCamera addTarget:_gpuImageView];

}` 这是我添加水印的代码 可是在[weakUIElementInput update] 这里崩溃

同时,如果服务器未打开,点击开始直播后,再次点击播报会崩溃到LFStreamRtmpSockert.m

- (void)_stop{
    if(self.delegate && [self.delegate respondsToSelector:@selector(socketStatus:status:)]){
        [self.delegate socketStatus:self status:LFLiveStop];
    }
    if(_rtmp != NULL){
        PILI_RTMP_Close(_rtmp, &_error);
        PILI_RTMP_Free(_rtmp);
        _rtmp = NULL;
    }
} 

Great library! I wonder can you make it more flexible by allowing users to specify the video input, for example, making LFVideoCapture an protocol and then let LFLiveSession accept an LFVideoCapture implementation. The implementation can call - (void)captureOutput:(nullable LFVideoCapture*)capture pixelBuffer:(nullable CVImageBufferRef)pixelBuffer; to specify the video input.

Thanks!

class PublishLiveViewController: UIViewController {
    
    @IBOutlet weak var livePathInput: UITextField!
    
    lazy var session: LFLiveSession = {
        let audioConfiguration = LFLiveAudioConfiguration.default()
        let videoConfiguration = LFLiveVideoConfiguration.defaultConfiguration(for: .high3)
        let session = LFLiveSession(audioConfiguration: audioConfiguration, videoConfiguration: videoConfiguration)
        session?.captureDevicePosition = .back
        session?.running = true
        session?.delegate = self
        session?.preView = self.view
        return session!
    }()

    
    
    override func viewDidLoad() {
        super.viewDidLoad()

        
    }
    
    override func viewDidDisappear(_ animated: Bool) {
        super.viewDidDisappear(animated)
        self.stopLive()
    }
    func startLive() -> Void {
        let stream = LFLiveStreamInfo()
        stream.url = livePathInput.text;
        livePathInput.isHidden = true
        session.preView = view
        session.startLive(stream)
    }
    
    func stopLive() -> Void {
        session.preView = nil
        livePathInput.isHidden = false
        session.stopLive()
    }
    
    @IBAction func publishLive(_ sender: UIButton) {
        
        startLive()
    }
    
    @IBAction func stopPublishLive(_ sender: UIButton) {
        
        stopLive()
    }
    
    @IBAction func switchCamera(_ sender: UIButton) {
        
        session.captureDevicePosition = session.captureDevicePosition == .back ? .front : .back;
    }

}

Crash code LFAudioCapture.m

- (void)dealloc {
    [[NSNotificationCenter defaultCenter] removeObserver:self];

     dispatch_async(self.taskQueue, ^{
        if (self.componetInstance) { // Crash code
            self.isRunning = NO;
            AudioOutputUnitStop(self.componetInstance);
            AudioComponentInstanceDispose(self.componetInstance);
            self.componetInstance = nil;
            self.component = nil;
        }
   });
}

如题，当我项目中pod同时引入 ReactiveCocoa 和 LFLiveKit 的时候，LFLiveKit 就会编译失败

我的pod写法入下：

source 'https://github.com/CocoaPods/Specs.git' platform :ios, "8.0"

use_frameworks!

target 'test' do

pod 'LFLiveKit' pod 'ReactiveCocoa'

end

我测试发现正常使用的过程中，源视频画面变化后，播放端要过6秒，甚至8秒以上才会看到。也就是延迟达到6-8秒。什么原因导致的？有什么参数可以调整试试看吗？

我测试的环境， 源端使用 iOS8.4.1， LFLiveKit 2.6 流媒体服务: Nginx+rmtp-module 播放端: vlc播放器 上述都在同一路由器网络中，其中nginx和vlc在同一PC机上。 我在 XCode8.1的输出中看到上传速度在 100KB左右，看似网络带宽不成问题。

2017-02-27 13:34:58.599 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 113.2 KB/s 2017-02-27 13:34:59.621 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 103.3 KB/s 2017-02-27 13:35:00.678 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 107.8 KB/s 2017-02-27 13:35:01.687 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 86.5 KB/s 2017-02-27 13:35:02.708 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 56.7 KB/s 2017-02-27 13:35:03.206 EasyLiveVideo[1235:145634] Increase bitrate 969200 2017-02-27 13:35:03.730 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 62.7 KB/s 2017-02-27 13:35:04.764 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 36.9 KB/s 2017-02-27 13:35:05.768 EasyLiveVideo[1235:145634] debugInfo uploadSpeed: 60.1 KB/s

我使用的参数如下，

`

• (LFLiveSession*)session{ if(!_session){ @try {
  LFLiveAudioConfiguration *audioConfiguration = [LFLiveAudioConfiguration new]; audioConfiguration.numberOfChannels = 2; audioConfiguration.audioBitrate = LFLiveAudioBitRate_32Kbps; audioConfiguration.audioSampleRate = LFLiveAudioSampleRate_16000Hz;

        LFLiveVideoConfiguration *videoConfiguration = [LFLiveVideoConfiguration defaultConfigurationForQuality:LFLiveVideoQuality_High1];
        videoConfiguration.videoSize = CGSizeMake(960, 540);
        videoConfiguration.videoBitRate = 800*1024;
        videoConfiguration.videoMaxBitRate = 1200*1024;
        videoConfiguration.videoMinBitRate = 400*1024;
        videoConfiguration.videoFrameRate = 10;
        videoConfiguration.videoMaxFrameRate = 15;
        videoConfiguration.videoMinFrameRate = 5;
        videoConfiguration.videoMaxKeyframeInterval = 20;
        videoConfiguration.outputImageOrientation =[[UIApplication sharedApplication] statusBarOrientation];
        videoConfiguration.sessionPreset = LFCaptureSessionPreset540x960;
  
        _session = [[LFLiveSession alloc] initWithAudioConfiguration:audioConfiguration videoConfiguration:videoConfiguration];
        _session.adaptiveBitrate =YES;
        _session.reconnectInterval =8;
        _session.reconnectCount =6;
  
        _session.saveLocalVideo =YES;
        NSString *cachesDir = [NSSearchPathForDirectoriesInDomains(NSCachesDirectory, NSUserDomainMask, YES) objectAtIndex:0];
        NSString * pathToMovie =[cachesDir stringByAppendingPathComponent:[NSString stringWithFormat:@"%@.mp4", [ELVView dataAndTimeNow]]];
        unlink([pathToMovie UTF8String]); // If a file already exists, AVAssetWriter won't let you record new frames, so delete the old movie
        _session.saveLocalVideoPath =[NSURL fileURLWithPath:pathToMovie];
  
        _session.showDebugInfo =YES;
  
        // 设置代理
        _session.delegate = self;
        _session.running = YES;
        _session.preView = self;
        _session.captureDevicePosition = AVCaptureDevicePositionBack;
    } @catch (NSException *exception) {
  
    } @finally {
  
    }
  

  } return _session; } `

另，作者这个库非常好。感谢~

I am new to this. This is a question on iOS

I wonder if there is a way to access to image of each video frame. There is a UIImage *currentImage property in LFLiveSession, however, the LFLiveSession delegate callback is only triggered when live state change.

If I can access each video frame, I would like to overlay objects to the video frame. Is this possible?

Thanks

I want to crop and keep aspect ratio in swift 3.1.

let audioConfiguration = LFLiveAudioConfiguration.defaultConfiguration(for: .medium)
let videoConfiguration = LFLiveVideoConfiguration.defaultConfiguration(for: .medium3, outputImageOrientation: .portrait)
videoConfiguration?.videoSize = CGSize(width: 720, height: 360)
videoConfiguration?.videoSizeRespectingAspectRatio = false

let session = LFLiveSession(audioConfiguration: audioConfiguration, videoConfiguration: videoConfiguration)!
session.delegate = self
session.captureDevicePosition = .back
session.preView = self.cameraView

But, I can't to crop while keep aspect ratio.

I did not solve with under issue.

• https://github.com/LaiFengiOS/LFLiveKit/issues/86
• https://github.com/LaiFengiOS/LFLiveKit/issues/166

Please teach me 🙏

Hey,

I started testing LFLiveKit with the code example and I have this error: ERROR: PILI_RTMP_ReadPacket, failed to read PILI_RTMP packet header

What's PILI RTMP ?

I would like to stream to YouTube, I use this code to start streaming:

let stream = LFLiveStreamInfo()
stream.url = "rtmp://a.rtmp.youtube.com/live2"
stream.streamId = "xxxx-xxxx-xxxx-xxxx"
session.startLive(stream)

Configuration: Xcode 8 Swift 3 iOS 10

Thanks,

Remi.

When trying to Archive my project which uses LFLiveKit, I get the following error:

ld: bitcode bundle could not be generated because '/path-to-project/Pods/LFLiveKit/Vendor/pili-librtmp.framework/pili-librtmp(amf.o)' was built without full bitcode. All object files and libraries for bitcode must be generated from Xcode Archive or Install build for architecture armv7
clang: error: linker command failed with exit code 1 (use -v to see invocation)

It doesn't work whether I set Enable Bitcode to YES or NO.

在4G环境下采集推流控制台提示: handleRouteChange reason is The category of the session object changed. handleRouteChange reason is The output route was overridden by the app. 数据应该没有推出去!

Hello, I using this project to implement live streaming app. It's work great. But I need face detect function to add some specail filter for user. If I want to add AVCaptureMetadataOutput to support face detect, where is easy to add it? Or what suggest do you prefer??

When try to stream multiple rtmp urls app was crashing in "C rtmp" file line -> SSL_load_error_strings();

B2B(32583,0x16dfc7000) malloc: Non-aligned pointer 0x2823e0f00 being freed (2) B2B(32583,0x16dfc7000) malloc: *** set a breakpoint in malloc_error_break to debug

How to achieve this : Simultaneously record the output from the front and back cameras into a single movie file by using a multi-camera capture session. And stream this to RTMP. https://developer.apple.com/documentation/avfoundation/cameras_and_media_capture/avmulticampip_capturing_from_multiple_cameras