Recherche avancée

Sur d’autres sites (6475)

• MediaCodec - save timing info for ffmpeg ?

  18 novembre 2014, par Mark

  I have a requirement to encrypt video before it hits the disk. It seems on Android the only way to do this is to use MediaCodec, and encrypt and save the raw h264 elementary streams. (The MediaRecorder and Muxer classes operate on FileDescriptors, not an OutputStream, so I can’t wrap it with a CipherOutputStream).

  Using the grafika code as a base, I’m able to save a raw h264 elementary stream by replacing the Muxer in the VideoEncoderCore class with a WriteableByteChannel, backed by a CipherOutputStream (code below, minus the CipherOutputStream).

  If I take the resulting output file over to the desktop I’m able to use ffmpeg to mux the h264 stream to a playable mp4 file. What’s missing however is timing information. ffmpeg always assumes 25fps. What I’m looking for is a way to save the timing info, perhaps to a separate file, that I can use to give ffmpeg the right information on the desktop.

  I’m not doing audio yet, but I can imagine I’ll need to do the same thing there, if I’m to have any hope of remotely accurate syncing.

  FWIW, I’m a total newbie here, and I really don’t know much of anything about SPS, NAL, Atoms, etc.

  /*
  
   * Copyright 2014 Google Inc. All rights reserved.
  
   *
  
   * Licensed under the Apache License, Version 2.0 (the "License");
  
   * you may not use this file except in compliance with the License.
  
   * You may obtain a copy of the License at
  
   *
  
   *      http://www.apache.org/licenses/LICENSE-2.0
  
   *
  
   * Unless required by applicable law or agreed to in writing, software
  
   * distributed under the License is distributed on an "AS IS" BASIS,
  
   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  
   * See the License for the specific language governing permissions and
  
   * limitations under the License.
  
   */
  
  
  
  
  
  import android.media.MediaCodec;
  
  import android.media.MediaCodecInfo;
  
  import android.media.MediaFormat;
  
  import android.util.Log;
  
  import android.view.Surface;
  
  
  
  import java.io.BufferedOutputStream;
  
  import java.io.File;
  
  import java.io.FileOutputStream;
  
  import java.io.IOException;
  
  import java.nio.ByteBuffer;
  
  import java.nio.channels.Channels;
  
  import java.nio.channels.WritableByteChannel;
  
  
  
  /**
  
   * This class wraps up the core components used for surface-input video encoding.
  
   * <p>
  
   * Once created, frames are fed to the input surface.  Remember to provide the presentation
  
   * time stamp, and always call drainEncoder() before swapBuffers() to ensure that the
  
   * producer side doesn't get backed up.
  
   * </p><p>
  
   * This class is not thread-safe, with one exception: it is valid to use the input surface
  
   * on one thread, and drain the output on a different thread.
  
   */
  
  public class VideoEncoderCore {
  
      private static final String TAG = MainActivity.TAG;
  
      private static final boolean VERBOSE = false;
  
  
  
      // TODO: these ought to be configurable as well
  
      private static final String MIME_TYPE = "video/avc";    // H.264 Advanced Video Coding
  
      private static final int FRAME_RATE = 30;               // 30fps
  
      private static final int IFRAME_INTERVAL = 5;           // 5 seconds between I-frames
  
  
  
      private Surface mInputSurface;
  
      private MediaCodec mEncoder;
  
      private MediaCodec.BufferInfo mBufferInfo;
  
      private int mTrackIndex;
  
      //private MediaMuxer mMuxer;
  
      //private boolean mMuxerStarted;
  
      private WritableByteChannel outChannel;
  
  
  
      /**
  
       * Configures encoder and muxer state, and prepares the input Surface.
  
       */
  
      public VideoEncoderCore(int width, int height, int bitRate, File outputFile)
  
              throws IOException {
  
          mBufferInfo = new MediaCodec.BufferInfo();
  
  
  
          MediaFormat format = MediaFormat.createVideoFormat(MIME_TYPE, width, height);
  
  
  
          // Set some properties.  Failing to specify some of these can cause the MediaCodec
  
          // configure() call to throw an unhelpful exception.
  
          format.setInteger(MediaFormat.KEY_COLOR_FORMAT,
  
                  MediaCodecInfo.CodecCapabilities.COLOR_FormatSurface);
  
          format.setInteger(MediaFormat.KEY_BIT_RATE, bitRate);
  
          format.setInteger(MediaFormat.KEY_FRAME_RATE, FRAME_RATE);
  
          format.setInteger(MediaFormat.KEY_I_FRAME_INTERVAL, IFRAME_INTERVAL);
  
          if (VERBOSE) Log.d(TAG, "format: " + format);
  
  
  
          // Create a MediaCodec encoder, and configure it with our format.  Get a Surface
  
          // we can use for input and wrap it with a class that handles the EGL work.
  
          mEncoder = MediaCodec.createEncoderByType(MIME_TYPE);
  
          mEncoder.configure(format, null, null, MediaCodec.CONFIGURE_FLAG_ENCODE);
  
          mInputSurface = mEncoder.createInputSurface();
  
          mEncoder.start();
  
  
  
          // Create a MediaMuxer.  We can't add the video track and start() the muxer here,
  
          // because our MediaFormat doesn't have the Magic Goodies.  These can only be
  
          // obtained from the encoder after it has started processing data.
  
          //
  
          // We're not actually interested in multiplexing audio.  We just want to convert
  
          // the raw H.264 elementary stream we get from MediaCodec into a .mp4 file.
  
          //mMuxer = new MediaMuxer(outputFile.toString(),
  
          //        MediaMuxer.OutputFormat.MUXER_OUTPUT_MPEG_4);
  
  
  
          mTrackIndex = -1;
  
          //mMuxerStarted = false;
  
          outChannel = Channels.newChannel(new BufferedOutputStream(new FileOutputStream(outputFile)));
  
      }
  
  
  
      /**
  
       * Returns the encoder's input surface.
  
       */
  
      public Surface getInputSurface() {
  
          return mInputSurface;
  
      }
  
  
  
      /**
  
       * Releases encoder resources.
  
       */
  
      public void release() {
  
          if (VERBOSE) Log.d(TAG, "releasing encoder objects");
  
          if (mEncoder != null) {
  
              mEncoder.stop();
  
              mEncoder.release();
  
              mEncoder = null;
  
          }
  
          try {
  
              outChannel.close();
  
          }
  
          catch (Exception e) {
  
              Log.e(TAG,"Couldn't close output stream.");
  
          }
  
      }
  
  
  
      /**
  
       * Extracts all pending data from the encoder and forwards it to the muxer.
  
       * </p><p>
  
       * If endOfStream is not set, this returns when there is no more data to drain.  If it
  
       * is set, we send EOS to the encoder, and then iterate until we see EOS on the output.
  
       * Calling this with endOfStream set should be done once, right before stopping the muxer.
  
       * </p><p>
  
       * We're just using the muxer to get a .mp4 file (instead of a raw H.264 stream).  We're
  
       * not recording audio.
  
       */
  
      public void drainEncoder(boolean endOfStream) {
  
          final int TIMEOUT_USEC = 10000;
  
          if (VERBOSE) Log.d(TAG, "drainEncoder(" + endOfStream + ")");
  
  
  
          if (endOfStream) {
  
              if (VERBOSE) Log.d(TAG, "sending EOS to encoder");
  
              mEncoder.signalEndOfInputStream();
  
          }
  
  
  
          ByteBuffer[] encoderOutputBuffers = mEncoder.getOutputBuffers();
  
          while (true) {
  
              int encoderStatus = mEncoder.dequeueOutputBuffer(mBufferInfo, TIMEOUT_USEC);
  
              if (encoderStatus == MediaCodec.INFO_TRY_AGAIN_LATER) {
  
                  // no output available yet
  
                  if (!endOfStream) {
  
                      break;      // out of while
  
                  } else {
  
                      if (VERBOSE) Log.d(TAG, "no output available, spinning to await EOS");
  
                  }
  
              } else if (encoderStatus == MediaCodec.INFO_OUTPUT_BUFFERS_CHANGED) {
  
                  // not expected for an encoder
  
                  encoderOutputBuffers = mEncoder.getOutputBuffers();
  
              } else if (encoderStatus == MediaCodec.INFO_OUTPUT_FORMAT_CHANGED) {
  
                  // should happen before receiving buffers, and should only happen once
  
                  //if (mMuxerStarted) {
  
                  //    throw new RuntimeException("format changed twice");
  
                  //}
  
                  MediaFormat newFormat = mEncoder.getOutputFormat();
  
                  Log.d(TAG, "encoder output format changed: " + newFormat);
  
  
  
                  // now that we have the Magic Goodies, start the muxer
  
                  //mTrackIndex = mMuxer.addTrack(newFormat);
  
                  //mMuxer.start();
  
                  //mMuxerStarted = true;
  
              } else if (encoderStatus &lt; 0) {
  
                  Log.w(TAG, "unexpected result from encoder.dequeueOutputBuffer: " +
  
                          encoderStatus);
  
                  // let's ignore it
  
              } else {
  
                  ByteBuffer encodedData = encoderOutputBuffers[encoderStatus];
  
                  if (encodedData == null) {
  
                      throw new RuntimeException("encoderOutputBuffer " + encoderStatus +
  
                              " was null");
  
                  }
  
  
  
                  /*
  
                     FFMPEG needs this info.
  
                  if ((mBufferInfo.flags &amp; MediaCodec.BUFFER_FLAG_CODEC_CONFIG) != 0) {
  
                      // The codec config data was pulled out and fed to the muxer when we got
  
                      // the INFO_OUTPUT_FORMAT_CHANGED status.  Ignore it.
  
                      if (VERBOSE) Log.d(TAG, "ignoring BUFFER_FLAG_CODEC_CONFIG");
  
                      mBufferInfo.size = 0;
  
                  }
  
                  */
  
  
  
                  if (mBufferInfo.size != 0) {
  
                      /*
  
                      if (!mMuxerStarted) {
  
                          throw new RuntimeException("muxer hasn't started");
  
                      }
  
                      */
  
  
  
                      // adjust the ByteBuffer values to match BufferInfo (not needed?)
  
                      encodedData.position(mBufferInfo.offset);
  
                      encodedData.limit(mBufferInfo.offset + mBufferInfo.size);
  
  
  
                      try {
  
                          outChannel.write(encodedData);
  
                      }
  
                      catch (Exception e) {
  
                          Log.e(TAG,"Error writing output.",e);
  
                      }
  
                      if (VERBOSE) {
  
                          Log.d(TAG, "sent " + mBufferInfo.size + " bytes to muxer, ts=" +
  
                                  mBufferInfo.presentationTimeUs);
  
                      }
  
                  }
  
  
  
                  mEncoder.releaseOutputBuffer(encoderStatus, false);
  
  
  
                  if ((mBufferInfo.flags &amp; MediaCodec.BUFFER_FLAG_END_OF_STREAM) != 0) {
  
                      if (!endOfStream) {
  
                          Log.w(TAG, "reached end of stream unexpectedly");
  
                      } else {
  
                          if (VERBOSE) Log.d(TAG, "end of stream reached");
  
                      }
  
                      break;      // out of while
  
                  }
  
              }
  
          }
  
      }
  
  }
  
  </p>

• Inside WebM Technology : VP8 Intra and Inter Prediction

  20 juillet 2010, par noreply@blogger.com (Lou Quillio)
  Continuing our series on WebM technology, I will discuss the use of prediction methods in the VP8 video codec, with special attention to the TM_PRED and SPLITMV modes, which are unique to VP8.

  
  

  First, some background. To encode a video frame, block-based codecs such as VP8 first divide the frame into smaller segments called macroblocks. Within each macroblock, the encoder can predict redundant motion and color information based on previously processed blocks. The redundant data can be subtracted from the block, resulting in more efficient compression.

  
  Image by Fido Factor, licensed under Creative Commons Attribution License.
  Based on a work at www.flickr.com

  
  

  A VP8 encoder uses two classes of prediction :

  • Intra prediction uses data within a single video frame
  • Inter prediction uses data from previously encoded frames

  The residual signal data is then encoded using other techniques, such as transform coding.

  
  

  VP8 Intra Prediction Modes

  VP8 intra prediction modes are used with three types of macroblocks :

  • 4x4 luma
  • 16x16 luma
  • 8x8 chroma

  Four common intra prediction modes are shared by these macroblocks :
  

  • H_PRED (horizontal prediction). Fills each column of the block with a copy of the left column, L.
  • V_PRED (vertical prediction). Fills each row of the block with a copy of the above row, A.
  • DC_PRED (DC prediction). Fills the block with a single value using the average of the pixels in the row above A and the column to the left of L.
  • TM_PRED (TrueMotion prediction). A mode that gets its name from a compression technique developed by On2 Technologies. In addition to the row A and column L, TM_PRED uses the pixel P above and to the left of the block. Horizontal differences between pixels in A (starting from P) are propagated using the pixels from L to start each row.

  For 4x4 luma blocks, there are six additional intra modes similar to V_PRED and H_PRED, but correspond to predicting pixels in different directions. These modes are outside the scope of this post, but if you want to learn more see the VP8 Bitstream Guide.

  
  

  As mentioned above, the TM_PRED mode is unique to VP8. The following figure uses an example 4x4 block of pixels to illustrate how the TM_PRED mode works :

  

  Where C, As and Ls represent reconstructed pixel values from previously coded blocks, and X₀₀ through X₃₃ represent predicted values for the current block. TM_PRED uses the following equation to calculate X_ij :

  
  

  X_ij = L_i + A_j - C (i, j=0, 1, 2, 3)

  
  

  Although the above example uses a 4x4 block, the TM_PRED mode for 8x8 and 16x16 blocks works in the same fashion.

  TM_PRED is one of the more frequently used intra prediction modes in VP8, and for common video sequences it is typically used by 20% to 45% of all blocks that are intra coded. Overall, together with other intra prediction modes, TM_PRED helps VP8 to achieve very good compression efficiency, especially for key frames, which can only use intra modes (key frames by their very nature cannot refer to previously encoded frames).

  
  

  VP8 Inter Prediction Modes

  
  

  In VP8, inter prediction modes are used only on inter frames (non-key frames). For any VP8 inter frame, there are typically three previously coded reference frames that can be used for prediction. A typical inter prediction block is constructed using a motion vector to copy a block from one of the three frames. The motion vector points to the location of a pixel block to be copied. In most video compression schemes, a good portion of the bits are spent on encoding motion vectors ; the portion can be especially large for video encoded at lower datarates.

  
  

  Like previous VPx codecs, VP8 encodes motion vectors very efficiently by reusing vectors from neighboring macroblocks (a macroblock includes one 16x16 luma block and two 8x8 chroma blocks). VP8 uses a similar strategy in the overall design of inter prediction modes. For example, the prediction modes "NEAREST" and "NEAR" make use of last and second-to-last, non-zero motion vectors from neighboring macroblocks. These inter prediction modes can be used in combination with any of the three different reference frames.

  
  

  In addition, VP8 has a very sophisticated, flexible inter prediction mode called SPLITMV. This mode was designed to enable flexible partitioning of a macroblock into sub-blocks to achieve better inter prediction. SPLITMV is very useful when objects within a macroblock have different motion characteristics. Within a macroblock coded using SPLITMV mode, each sub-block can have its own motion vector. Similar to the strategy of reusing motion vectors at the macroblock level, a sub-block can also use motion vectors from neighboring sub-blocks above or left to the current block. This strategy is very flexible and can effectively encode any shape of sub-macroblock partitioning, and does so efficiently. Here is an example of a macroblock with 16x16 luma pixels that is partitioned to 16 4x4 blocks :

  
  

  

  
  

  where New represents a 4x4 bock coded with a new motion vector, and Left and Above represent a 4x4 block coded using the motion vector from the left and above, respectively. This example effectively partitions the 16x16 macroblock into 3 different segments with 3 different motion vectors (represented below by 1, 2 and 3) :

  
  

  

  
  

  Through effective use of intra and inter prediction modes, WebM encoder implementations can achieve great compression quality on a wide range of source material. If you want to delve further into VP8 prediction modes, read the VP8 Bitstream Guide or examine the reconintra.c and rdopt.c files in the VP8 source tree.

  
  

  Yaowu Xu, Ph.D. is a codec engineer at Google.

  
  

  

• Inside WebM Technology : VP8 Intra and Inter Prediction

  20 juillet 2010, par noreply@blogger.com (Lou Quillio)
  Continuing our series on WebM technology, I will discuss the use of prediction methods in the VP8 video codec, with special attention to the TM_PRED and SPLITMV modes, which are unique to VP8.

  
  

  First, some background. To encode a video frame, block-based codecs such as VP8 first divide the frame into smaller segments called macroblocks. Within each macroblock, the encoder can predict redundant motion and color information based on previously processed blocks. The redundant data can be subtracted from the block, resulting in more efficient compression.

  
  Image by Fido Factor, licensed under Creative Commons Attribution License.
  Based on a work at www.flickr.com

  
  

  A VP8 encoder uses two classes of prediction :

  • Intra prediction uses data within a single video frame
  • Inter prediction uses data from previously encoded frames

  The residual signal data is then encoded using other techniques, such as transform coding.

  
  

  VP8 Intra Prediction Modes

  VP8 intra prediction modes are used with three types of macroblocks :

  • 4x4 luma
  • 16x16 luma
  • 8x8 chroma

  Four common intra prediction modes are shared by these macroblocks :
  

  • H_PRED (horizontal prediction). Fills each column of the block with a copy of the left column, L.
  • V_PRED (vertical prediction). Fills each row of the block with a copy of the above row, A.
  • DC_PRED (DC prediction). Fills the block with a single value using the average of the pixels in the row above A and the column to the left of L.
  • TM_PRED (TrueMotion prediction). A mode that gets its name from a compression technique developed by On2 Technologies. In addition to the row A and column L, TM_PRED uses the pixel P above and to the left of the block. Horizontal differences between pixels in A (starting from P) are propagated using the pixels from L to start each row.

  For 4x4 luma blocks, there are six additional intra modes similar to V_PRED and H_PRED, but correspond to predicting pixels in different directions. These modes are outside the scope of this post, but if you want to learn more see the VP8 Bitstream Guide.

  
  

  As mentioned above, the TM_PRED mode is unique to VP8. The following figure uses an example 4x4 block of pixels to illustrate how the TM_PRED mode works :

  

  Where C, As and Ls represent reconstructed pixel values from previously coded blocks, and X₀₀ through X₃₃ represent predicted values for the current block. TM_PRED uses the following equation to calculate X_ij :

  
  

  X_ij = L_i + A_j - C (i, j=0, 1, 2, 3)

  
  

  Although the above example uses a 4x4 block, the TM_PRED mode for 8x8 and 16x16 blocks works in the same fashion.

  TM_PRED is one of the more frequently used intra prediction modes in VP8, and for common video sequences it is typically used by 20% to 45% of all blocks that are intra coded. Overall, together with other intra prediction modes, TM_PRED helps VP8 to achieve very good compression efficiency, especially for key frames, which can only use intra modes (key frames by their very nature cannot refer to previously encoded frames).

  
  

  VP8 Inter Prediction Modes

  
  

  In VP8, inter prediction modes are used only on inter frames (non-key frames). For any VP8 inter frame, there are typically three previously coded reference frames that can be used for prediction. A typical inter prediction block is constructed using a motion vector to copy a block from one of the three frames. The motion vector points to the location of a pixel block to be copied. In most video compression schemes, a good portion of the bits are spent on encoding motion vectors ; the portion can be especially large for video encoded at lower datarates.

  
  

  Like previous VPx codecs, VP8 encodes motion vectors very efficiently by reusing vectors from neighboring macroblocks (a macroblock includes one 16x16 luma block and two 8x8 chroma blocks). VP8 uses a similar strategy in the overall design of inter prediction modes. For example, the prediction modes "NEAREST" and "NEAR" make use of last and second-to-last, non-zero motion vectors from neighboring macroblocks. These inter prediction modes can be used in combination with any of the three different reference frames.

  
  

  In addition, VP8 has a very sophisticated, flexible inter prediction mode called SPLITMV. This mode was designed to enable flexible partitioning of a macroblock into sub-blocks to achieve better inter prediction. SPLITMV is very useful when objects within a macroblock have different motion characteristics. Within a macroblock coded using SPLITMV mode, each sub-block can have its own motion vector. Similar to the strategy of reusing motion vectors at the macroblock level, a sub-block can also use motion vectors from neighboring sub-blocks above or left to the current block. This strategy is very flexible and can effectively encode any shape of sub-macroblock partitioning, and does so efficiently. Here is an example of a macroblock with 16x16 luma pixels that is partitioned to 16 4x4 blocks :

  
  

  

  
  

  where New represents a 4x4 bock coded with a new motion vector, and Left and Above represent a 4x4 block coded using the motion vector from the left and above, respectively. This example effectively partitions the 16x16 macroblock into 3 different segments with 3 different motion vectors (represented below by 1, 2 and 3) :

  
  

  

  
  

  Through effective use of intra and inter prediction modes, WebM encoder implementations can achieve great compression quality on a wide range of source material. If you want to delve further into VP8 prediction modes, read the VP8 Bitstream Guide or examine the reconintra.c and rdopt.c files in the VP8 source tree.

  
  

  Yaowu Xu, Ph.D. is a codec engineer at Google.