Recherche avancée

Sur d’autres sites (3125)

• +"archived version" download link

  6 janvier 2013, par scottschiller

  m doc/download/index.html +"archived version" download link

• how to download portion of video which was uploaded into AWS s3 bucket, though Nodejs SDKs

  22 février 2024, par rama rangeswara reddy

  I have uploaded a 1GB .mp4 file to an AWS S3 bucket. Using the AWS-SDK provided by the npm package, I am able to download the entire video. However, I have a specific requirement to generate a thumbnail at the 6-second mark of the video. Currently, I download the entire 1GB video to my local machine and then generate the thumbnail at the desired duration.

  


  To optimize server resources and reduce disk load, I plan to download only the first 10 seconds of the video, which should be approximately 10MB or less in size. By doing so, I can significantly reduce download time and server load while still fulfilling my requirement of generating the thumbnail at the 6-second mark. Therefore, instead of downloading the entire 1GB video, I aim to download only the 10MB segment corresponding to the first 10 seconds of the video.

  


  I am using nodejs, expressJS, as backed Technologies.

  


  `

  


  `async function downloadS3FileToLocalDirAndReturnPath(videoKey) {
    return new Promise(async (resolve, reject) => {
        try {
            AWS.config.update({
                accessKeyId: config.AWS.KEYS.accessKeyId,
                secretAccessKey: config.AWS.KEYS.secretAccessKey,
                region: config.AWS.KEYS.region,
                httpOptions: config.AWS.KEYS.httpOptions
            });
            const s3 = new AWS.S3();

            // Specify the local file path where you want to save the downloaded video
            const localFilePath = `${os.tmpdir()}/${Date.now()}_sre.mp4`;

            // Configure the parameters for the S3 getObject operation
            const params = {
                Bucket: config.AWS.S3_BUCKET,
                Key: videoKey
            };

            const result = await s3.getObject(params).promise();
            const fileContent = result.Body;
            fs.writeFileSync(localFilePath, fileContent);
            resolve(localFilePath);
        } catch (error) {
            reject(error);
        }
    });
}`


  


  this code was working fine to download the whole video , but i need to download only first 10 seconds duration

  


  S3 : How to do a partial read / seek without downloading the complete file ?

  


  I tried this ,before posting this question with above post, video was downloading , it was not playing , by throwing this error , the file contains no playable streams

  


  async function generateThumbnails(videoKey) {

const s3 = new AWS.S3();

const params = {
    Bucket: KEYS.bucket,
    Key: videoKey, // Specify the key of the video file in S3
    Range: `bytes=0-${1024 * 800}`, // Specify the range of bytes you want to retrieve
};

const file = fs.createWriteStream(`/tmp/${Date.now()}_rama.mp4`);

const s3Stream = s3.getObject(params).createReadStream();

s3Stream.pipe(file);

s3Stream.on("error", (error) => {
    console.log("Error Occured while File downloading!! ");
});

s3Stream.on("finish", () => {
    console.log("File downloaded Successfully ");
});


  


  }

  


• How to download .m3u8 in once time

  11 octobre 2024, par Nabi K.A.Z.

  I have a .m3u8 file on remote host, with contain fixed numbers of chunk .ts file name, and not stream :

  



  #EXTM3U
#EXT-X-VERSION:3
#EXT-X-TARGETDURATION:11
#EXT-X-MEDIA-SEQUENCE:0
#EXTINF:9.736,
media_0.ts
#EXTINF:9.96,
media_1.ts
#EXTINF:10.0,
media_2.ts
#EXTINF:10.0,
media_3.ts
#EXTINF:10.0,
media_4.ts
#EXTINF:10.2,
media_5.ts
#EXTINF:10.0,


  



  When I use this command :

  



  # ffmpeg -i "http://example.com/chunklist.m3u8" file.mp4

frame=  582 fps=9.4 q=28.0 size=    1536kB time=00:00:23.21 bitrate= 542.1kbits/s dup=2 drop=4 speed=0.375x


  



  It works. But It get frame by frame video and very long time needed. (It takes time almost to playback the video.)

  



  But since the path of all the .ts files are known. (http://example.com/media_0.ts, http://example.com/media_1.ts, ...) There must be a way to get and merge them all at the same time.

  



  But How in ffmpeg directly ?!

  



  EDIT (try a solution) :

  



  For one solution, I know how can concatenation files with ffmpeg.

  



  ffmpeg -i "concat:0.ts|1.ts|2.ts|3.ts|4.ts|5.ts" -c copy output.mp4


  



  This ffmpeg command was great, and works in less 1 sec time !

  



  So try to download all .ts files with CURL with this command :

  



  curl \
http://example.com/media_0.ts -o 0.ts \
http://example.com/media_1.ts -o 1.ts \
http://example.com/media_2.ts -o 2.ts \
http://example.com/media_3.ts -o 3.ts \
http://example.com/media_4.ts -o 4.ts \
http://example.com/media_5.ts -o 5.ts


  



  But you can see result :

  



    % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  687k  100  687k    0     0  75108      0  0:00:09  0:00:09 --:--:-- 74111
100  652k  100  652k    0     0  59404      0  0:00:11  0:00:11 --:--:-- 53400
100  673k  100  673k    0     0  48675      0  0:00:14  0:00:14 --:--:-- 55781
100  657k  100  657k    0     0  63573      0  0:00:10  0:00:10 --:--:-- 62494
100  671k  100  671k    0     0  39019      0  0:00:17  0:00:17 --:--:-- 40863
100  692k  100  692k    0     0  63480      0  0:00:11  0:00:11 --:--:-- 80049


  



  See, total download time was 72 sec, while the total duration of all parts is 59 sec ! that this time is very long !

  



  So sorry, download all parts and then concat that, was not good solution.

  



  EDIT 2

  



  I try for another .m3u8 file on the another server with difference URL :

  



  Download and concat together :

  



  ffmpeg -i "concat:\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_0.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_1.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_2.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_3.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_4.ts|\
http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_5.ts\
" -c copy -y output.ts


  



  Another command with input.txt URLs file.

  



  ffmpeg -f "concat" -i "input.txt" -c copy -y output.ts


  



  input.txt file :

  



  file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_0.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_1.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_2.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_3.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_4.ts'
file 'http://184.72.239.149/vod/smil:BigBuckBunny.smil/media_w442897525_b560000_5.ts'


  



  Or this command some time if needed :

  



  ffmpeg -f "concat" -safe "0" -protocol_whitelist "file,http,https,tcp,tls" -i "input.txt" -c copy -y output.ts


  



  Finally, for that download speed was good, MAYBE my server target has limited bandwidth. :-(