Question

我已经从Github获取了AVAnimator库的最新代码。除了AutoPropertyRelease.h和AutoPropertyRelease.m之外，我还添加了文件夹AVAnimator和LZMASDK。

我使用以下代码播放带alpha通道的视频（从另一个AVAnimator应用示例中采用）：

 CGRect iPhoneExplosionRect = CGRectMake(0, -2, 640/2, 480/2);
CGRect iPadExplosionRect = CGRectMake(0, -5, 840, 630);

NSString *rgbResourceName = @"ExplosionAdjusted_rgb_CRF_30_24BPP.m4v";
NSString *alphaResourceName = @"ExplosionAdjusted_alpha_CRF_30_24BPP.m4v";

// Output filename

NSString *tmpFilename;
NSString *tmpPath;
tmpFilename = @"Explosion.mvid";
tmpPath = [AVFileUtil getTmpDirPath:tmpFilename];

// Set to TRUE to always decode from H.264

BOOL alwaysDecode = FALSE;

if (alwaysDecode && [AVFileUtil fileExists:tmpPath]) {
    BOOL worked = [[NSFileManager defaultManager] removeItemAtPath:tmpPath error:nil];
    NSAssert(worked, @"could not remove file %@", tmpPath);
}

// This loader will join 2 H.264 videos together into a single 32BPP .mvid

AVAssetJoinAlphaResourceLoader *resLoader = [AVAssetJoinAlphaResourceLoader aVAssetJoinAlphaResourceLoader];

resLoader.movieRGBFilename = rgbResourceName;
resLoader.movieAlphaFilename = alphaResourceName;
resLoader.outPath = tmpPath;
//resLoader.alwaysGenerateAdler = TRUE;

AVAnimatorMedia *media = [AVAnimatorMedia aVAnimatorMedia];
media.resourceLoader = resLoader;

self.expMedia = media;

// Frame decoder will read from generated .mvid file

AVMvidFrameDecoder *aVMvidFrameDecoder = [AVMvidFrameDecoder aVMvidFrameDecoder];
media.frameDecoder = aVMvidFrameDecoder;

// Create layer that video data will be directed into

CGRect expFrame;

if ([[UIDevice currentDevice] userInterfaceIdiom] == UIUserInterfaceIdiomPhone) {
    expFrame = iPhoneExplosionRect;
} else {
    expFrame = iPadExplosionRect;
}

CALayer *layer = [CALayer layer];
layer.frame = expFrame;
}

if (FALSE) {
    layer.backgroundColor = [UIColor orangeColor].CGColor;
}

[self.view.layer addSublayer:layer];

AVAnimatorLayer *animatorLayer = [AVAnimatorLayer aVAnimatorLayer:layer];

self.expAnimatorLayer = animatorLayer;

// Finally connect the media object to the layer so that rendering will be
// sent to the layer.

[animatorLayer attachMedia:media];

//media.animatorRepeatCount = 3;
//media.animatorRepeatCount = 30;
//media.animatorRepeatCount = INT_MAX;

[media prepareToAnimate];

一旦prepareToAnimate被命中，应用程序就会在断言中失败，如下所述

** Assertion failure in -[AVMvidFrameDecoder advanceToFrame:], /Users/user/Documents/.../AVAnimator/AVMvidFrameDecoder.m:844 

...*** Terminating app due to uncaught exception 'NSInternalInconsistencyException', reason: 'framebuffer num bytes'

我使用的架构（32位和64位）是否会导致它或其他什么？

我希望有人能让我走上正轨

AF

Answer 1

我研究了这种断言以及可能被复制的不同方式。我尝试不正确地设置arch标志，在可执行文件中只包含32位或仅64位代码，并在64位设备和模拟器上进行测试。我无法重现这个问题，目前尚不清楚它是如何发生的。为了收集更多信息，我添加了一个在分配帧缓冲区时以DEBUG或OPT模式编译的额外检查，是否可以从github中提取最新版本的AVAnimator或从patch here获取补丁。请确保您的代码版本中没有修改任何定义，并且您使用的是最新版本的Xcode。

AVAnimator断言失败 - [AVMvidFrameDecoder advanceToFrame：]＆＃39;帧缓冲区num字节＆＃39;

1 个答案: