Recherche avancée

Sur d’autres sites (4632)

• aarch64 : h264pred : Optimize the inner loop of existing 8 bit functions

  12 avril 2021, par Martin Storsjö

  aarch64 : h264pred : Optimize the inner loop of existing 8 bit functions
  Move the loop counter decrement further from the branch instruction,
  
  this hides the latency of the decrement.
  In loops that first load, then store (the horizontal prediction cases),
  
  do the decrement after the load (where the next instruction would
  
  stall a bit anyway, waiting for the result of the load).
  In loops that store twice using the same destination register,
  
  also do the decrement between the two stores (as the second store
  
  would need to wait for the updated destination register from the
  
  first instruction).
  In loops that store twice to two different destination registers,
  
  do the decrement before both stores, to do it as soon before the
  
  branch as possible.
  This gives minor (1-2 cycle) speedups in most cases (modulo measurement
  
  noise), but the horizontal prediction functions get a rather notable
  
  speedup on the Cortex A53.
  Before :                     Cortex A53     A72     A73
  
  pred8x8_dc_8_neon :                60.7    46.2    39.2
  
  pred8x8_dc_128_8_neon :            30.7    18.0    14.0
  
  pred8x8_horizontal_8_neon :        42.2    29.2    18.5
  
  pred8x8_left_dc_8_neon :           52.7    36.2    32.2
  
  pred8x8_mad_cow_dc_0l0_8_neon :    48.2    27.7    25.7
  
  pred8x8_mad_cow_dc_0lt_8_neon :    52.5    33.2    34.7
  
  pred8x8_mad_cow_dc_l0t_8_neon :    52.5    31.7    33.2
  
  pred8x8_mad_cow_dc_l00_8_neon :    43.2    27.0    25.5
  
  pred8x8_plane_8_neon :            112.2    86.2    88.2
  
  pred8x8_top_dc_8_neon :            40.7    23.0    21.2
  
  pred8x8_vertical_8_neon :          27.2    15.5    14.0
  
  pred16x16_dc_8_neon :              91.0    73.2    70.5
  
  pred16x16_dc_128_8_neon :          43.0    34.7    30.7
  
  pred16x16_horizontal_8_neon :      86.0    49.7    44.7
  
  pred16x16_left_dc_8_neon :         87.0    67.2    67.5
  
  pred16x16_plane_8_neon :          236.0   175.7   173.0
  
  pred16x16_top_dc_8_neon :          53.2    39.0    41.7
  
  pred16x16_vertical_8_neon :        41.7    29.7    31.0
  After :
  
  pred8x8_dc_8_neon :                59.0    46.7    42.5
  
  pred8x8_dc_128_8_neon :            28.2    18.0    14.0
  
  pred8x8_horizontal_8_neon :        34.2    29.2    18.5
  
  pred8x8_left_dc_8_neon :           51.0    38.2    32.7
  
  pred8x8_mad_cow_dc_0l0_8_neon :    46.7    28.2    26.2
  
  pred8x8_mad_cow_dc_0lt_8_neon :    55.2    33.7    37.5
  
  pred8x8_mad_cow_dc_l0t_8_neon :    51.2    31.7    37.2
  
  pred8x8_mad_cow_dc_l00_8_neon :    41.7    27.5    26.0
  
  pred8x8_plane_8_neon :            111.5    86.5    89.5
  
  pred8x8_top_dc_8_neon :            39.0    23.2    21.0
  
  pred8x8_vertical_8_neon :          27.2    16.0    14.0
  
  pred16x16_dc_8_neon :              85.0    70.2    70.5
  
  pred16x16_dc_128_8_neon :          42.0    30.0    30.7
  
  pred16x16_horizontal_8_neon :      66.5    49.5    42.5
  
  pred16x16_left_dc_8_neon :         81.0    66.5    67.5
  
  pred16x16_plane_8_neon :          235.0   175.7   173.0
  
  pred16x16_top_dc_8_neon :          52.0    39.0    41.7
  
  pred16x16_vertical_8_neon :        40.2    33.2    31.0
  Despite this, a number of these functions still are slower than
  
  what e.g. GCC 7 generates - this shows the relative speedup of the
  
  neon codepaths over the compiler generated ones :
                             Cortex A53    A72    A73
  
  pred8x8_dc_8_neon :               0.86   0.65   1.04
  
  pred8x8_dc_128_8_neon :           0.59   0.44   0.62
  
  pred8x8_horizontal_8_neon :       1.51   0.58   1.30
  
  pred8x8_left_dc_8_neon :          0.72   0.56   0.89
  
  pred8x8_mad_cow_dc_0l0_8_neon :   0.93   0.93   1.37
  
  pred8x8_mad_cow_dc_0lt_8_neon :   1.37   1.41   1.68
  
  pred8x8_mad_cow_dc_l0t_8_neon :   1.21   1.17   1.32
  
  pred8x8_mad_cow_dc_l00_8_neon :   1.24   1.19   1.60
  
  pred8x8_plane_8_neon :            3.36   3.58   3.76
  
  pred8x8_top_dc_8_neon :           0.97   0.99   1.43
  
  pred8x8_vertical_8_neon :         0.86   0.78   1.18
  
  pred16x16_dc_8_neon :             1.20   1.06   1.49
  
  pred16x16_dc_128_8_neon :         0.83   0.95   0.99
  
  pred16x16_horizontal_8_neon :     1.78   0.96   1.59
  
  pred16x16_left_dc_8_neon :        1.06   0.96   1.32
  
  pred16x16_plane_8_neon :          5.78   6.49   7.19
  
  pred16x16_top_dc_8_neon :         1.48   1.53   1.94
  
  pred16x16_vertical_8_neon :       1.39   1.34   1.98
  In particular, on Cortex A72, many of these functions are slower
  
  than the compiler generated code, while they're more beneficial on
  
  e.g. the Cortex A73.
  Signed-off-by : Martin Storsjö <martin@martin.st>
  

  • [DH] libavcodec/aarch64/h264pred_neon.S

• FFMPEG conversion (h.264) taking long time for short videos

  15 février 2023, par Sara

  I am trying to record the video and upload into the aws s3 server. Vuejs as front end and php Laravel as backend, I was not using any conversion before saving it to s3. Due to this if any recording recorded from android cannot be played in apple device due to some codecs..
To over come this, I am using ffmpeg to encode in X264() format to make it play in apple and android device regardless on which device the recording is done.

  


  1 min video taking 6-7 minutes using ffmpeg. I thought may be aws s3 taking time to save, i commented "saving to s3 bucket code" still very slow to save temp public folder in php.

  


  please check the code if i am missing anything to make conversion quick. if any solution update answer with reference link or code snippet with reference to my code below.

  


  public function video_upload(Request $request)
    {
                // Response Declaration   
                $response=array();
                $response_code = 200;
                $response['status'] = false;
                $response['data'] = [];
                // Validation
                // TODO: Specify mimes:mp4,webm,ogg etc 
                $validator = Validator::make(
                $request->all(), [
                'file' => 'required'
                ]
                );
                if ($validator->fails()) {
                $response['data']['validator'] = $validator->errors();
                return response()->json($response);
                }
                try{
                    $file = $request->file('file');
                    //convert
                    $ffmpeg = FFMpeg\FFMpeg::create();
                    $video = $ffmpeg->open($file);
                    $format = new X264(); 
                    //end convert
                    $file_name =  str_replace (' ', '-', Hash::make(time()));
                    $file_name = preg_replace('/[^A-Za-z0-9\-]/', '',$file_name).'.mp4';
                    
                    $video->save($format, $file_name);
                    $file_folder = 'uploads/video/';
                    // Store the file to S3
                    
                    // $store = Storage::disk('s3')->put($file_folder.$file_name, file_get_contents($file));
                    $store = Storage::disk('s3')->put($file_folder.$file_name, file_get_contents($file_name));
                    if($store){
                        // Replace old file if exist
                        //delete the file from public folder
                        $file = public_path($file_name);
                        if (file_exists($file)) {
                            unlink($file);
                        }

                        if(isset($request->old_file)){
 
                            if(Storage::disk('s3')->exists($file_folder.basename($request->old_file))) {
                                 Storage::disk('s3')->delete($file_folder.basename($request->old_file));
                            }
                        }
                    }
                    $response['status'] = true;
                    $response['data']= '/s3/'.$file_folder. $file_name;

                }catch (\Exception $e) {
                    $response['data']['message']=$e->getMessage()."line".$e->getLine();
                    $response_code = 400;
                }
                return response()->json($response, $response_code);
    }


  


  Its blocking point for me. I cannot let user to wait 5-6 mins to upload 1 min video.

  


• Encoding video for real time streaming

  15 septembre 2019, par lilroo

  Miguel Grinberg’s article video streaming with flask, covers sending multiple JPG formatted images to a client through a multipart response to stream video to the browser.

  After doing some reading, I quickly realized that MJPEG is not a real video format, JPEGs are not optimized for video streaming and MJPEG cannot be streamed to a HTML5 video element.

  for my application (a screen sharing application written in python), I need to be able to encode and stream relatively smooth video (smoother than MJPEG anyway) in semi-realtime to the browser.

  I think what I need to do is encode a video stream using a tool like FFmpeg and then send chunks of that output to a browser using a multipart response - however it doesn’t look like there is a way to get FFmpeg (using python bindings) to write to an in-memory buffer.

  Is ffmpeg up to this task ?

  I appreciate that this question may sound nieve, but I would be thankful for being pointed in the right direction.