AVAssetWriter creating video from images with Ken Burns effect& - ios

Question

I am trying to generate video from images with AVAssetWriter and add Ken Burns effect to each image, this is my code for creating video from images array.
The questions is how to implement Ken Burns effect for each image for 5 seconds
- (void) writeImagesAsMovie:(NSArray *)array toPath:(NSString*)path
{
UIImage *first = array[0];
CGSize frameSize = first.size;
NSError *error = nil;
AVAssetWriter *videoWriter = [[AVAssetWriter alloc] initWithURL:
[NSURL fileURLWithPath:path] fileType:AVFileTypeQuickTimeMovie
error:&error];
if(error) {
NSLog(#"error creating AssetWriter: %#",[error description]);
}
NSDictionary *videoSettings = [NSDictionary dictionaryWithObjectsAndKeys:
AVVideoCodecH264, AVVideoCodecKey,
[NSNumber numberWithInt:frameSize.width], AVVideoWidthKey,
[NSNumber numberWithInt:frameSize.height], AVVideoHeightKey,
nil];
AVAssetWriterInput* writerInput = [AVAssetWriterInput
assetWriterInputWithMediaType:AVMediaTypeVideo
outputSettings:videoSettings];
NSMutableDictionary *attributes = [[NSMutableDictionary alloc] init];
[attributes setObject:[NSNumber numberWithUnsignedInt:kCVPixelFormatType_32ARGB] forKey:(NSString*)kCVPixelBufferPixelFormatTypeKey];
[attributes setObject:[NSNumber numberWithUnsignedInt:frameSize.width] forKey:(NSString*)kCVPixelBufferWidthKey];
[attributes setObject:[NSNumber numberWithUnsignedInt:frameSize.height] forKey:(NSString*)kCVPixelBufferHeightKey];
AVAssetWriterInputPixelBufferAdaptor *adaptor = [AVAssetWriterInputPixelBufferAdaptor
assetWriterInputPixelBufferAdaptorWithAssetWriterInput:writerInput
sourcePixelBufferAttributes:attributes];
[videoWriter addInput:writerInput];
// fixes all errors
writerInput.expectsMediaDataInRealTime = YES;
//Start a session:
BOOL start = [videoWriter startWriting];
NSLog(#"Session started? %d", start);
[videoWriter startSessionAtSourceTime:kCMTimeZero];
CVPixelBufferRef buffer = NULL;
buffer = [self pixelBufferFromCGImage:[first CGImage]];
BOOL result = [adaptor appendPixelBuffer:buffer withPresentationTime:kCMTimeZero];
if (result == NO) //failes on 3GS, but works on iphone 4
NSLog(#"failed to append buffer");
if(buffer)
CVBufferRelease(buffer);
[NSThread sleepForTimeInterval:0.05];
int fps = 1;
int i = 0;
for (UIImage *imgFrame in array)
{
if (adaptor.assetWriterInput.readyForMoreMediaData)
{
i++;
NSLog(#"inside for loop %d ",i);
CMTime frameTime = CMTimeMake(5, fps);
CMTime lastTime=CMTimeMake(i, fps);
CMTime presentTime=CMTimeAdd(lastTime, frameTime);
buffer = [self pixelBufferFromCGImage:[imgFrame CGImage]];
BOOL result = [adaptor appendPixelBuffer:buffer withPresentationTime:presentTime];
if (result == NO)
{
NSLog(#"failed to append buffer");
NSLog(#"The error is %#", [videoWriter error]);
}
if(buffer)
CVBufferRelease(buffer);
[NSThread sleepForTimeInterval:0.05];
}
else
{
NSLog(#"error");
i--;
}
[NSThread sleepForTimeInterval:0.02];
}
//Finish the session:
[writerInput markAsFinished];
[videoWriter finishWritingWithCompletionHandler:^{
}];
CVPixelBufferPoolRelease(adaptor.pixelBufferPool);
LRProjectListViewController * proj = (LRProjectListViewController*)[[(UINavigationController*)DELEGATE.window.rootViewController viewControllers] firstObject];
[proj playVideoWithUrl:path];
}

Answer 1

It depends on the frame rate you are writing out in your composition. Assuming that it's 30FPS, then in your for loop, you'll need 30 frames in the array for each second of video.
for (UIImage *imgFrame in array)
You can modify your code to write out to appendPixelBuffer multiple times for each frame as an alternative. If you want it to play for 5 seconds, then you'd need to loop (FPS * 5) times. BUT without manipulating the frame, that's just going to give you a static picture for 5 seconds.
Without knowing the details of what you want to achieve, I think that this is not the best way to handle this problem.
Instead, look at using CAAnimations with AVVideoCompositionCoreAnimationTool. This will work directly with AVAssetExportSession which is MUCH easier than playing with AVAssetWriter unless you have a very specific requirement (typically you need to manipulate the actual video frames).