Recherche avancée

Sur d’autres sites (11358)

• Download m3u8 video from web page with authorization

  31 août 2024, par maya_soloveva

  I try download m3u8 video from webpage with access by authorization.

  


  I use ffmpeg command.
I get headers from network tab.

  


  ffmpeg -y -loglevel verbose -headers "$(cat headers.txt)" -i "https://apps.foxford.ru/webinar-foxford/storage/api/v2/backends/yandex/sets/hls.webinar.foxford.ru::420e27e1-1c7d-43e9-9195-fc8b99e6f50a/objects/long.v2.yandex.video.360p.master.m3u8" output2.mp4 -c copy -f mp4 file:foxford_01.mp4.part


  


  headers.txt content :

  


  Host: apps.foxford.ru
User-Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:129.0) Gecko/20100101 Firefox/129.0
Accept: */*
Accept-Language: en-US,en;q=0.5
Accept-Encoding: gzip, deflate, br, zstd
Referer: https://apps.foxford.ru/webinar/?audience=foxford.ru&brand=foxford&embedded_origin=https%3A%2F%2Ffoxford.ru&external_id=5aefae2d-e810-49ef-8344-d3807535fd9e&lesson_id=250799&scope=Z2lkOi8vc3RvZWdlL1VsbXM6OlJvb21zOjpXZWJpbmFyLzUwNDcwMA&utm_referrer=https%3A%2F%2Ffoxford.ru%2Fcourses%2F9193%2Flessons%2F250799&backurl=https%253A%252F%252Fapps%252Efoxford%252Eru%252Fapi%252Fv1%252Fredirs%252Faudiences%252Ffoxford%252Eru%252Fapps%252Fwebinar
DNT: 1
Sec-Fetch-Dest: empty
Sec-Fetch-Mode: cors
Sec-Fetch-Site: same-origin
authorization: Bearer eyJhbGciOiJFUzI1NiJ9.eyJleHAiOjE3MjUwODU1MzUsImlzcyI6InN2Yy5mb3hmb3JkLnJ1IiwiYXVkIjoidXNyLmZveGZvcmQucnUiLCJzdWIiOiJaMmxrT2k4dmMzUnZaV2RsTDFWelpYSTZPbEIxY0dsc0x6RXdPRFU0TXpJeSJ9.NE9IsraTHFKicr3lkjzTTEkGLoGP-bL74nlX1umcPjC62h9MGXjPUFXL53jsnixd7b-WPtIc8N2WhsC6Ipt8IQ
Connection: keep-alive
Cookie: _csrf_token=%2Ba0goi1UnpRzFxpPOIbnZBaxtY0YUlZyBnyGohC5NbfBaVzLvwpEtM0x%2BjEPvJqJrcvsy2qMZHTStvusdxeUmQ%3D%3D; advcake_trackid=4b724357-eb38-b2f0-6919-4a4363aef58a; _tm_lt_sid=1720871531732.305536; client_timezone=Europe/Moscow; _ga=GA1.2.506102438.1720871532; tmr_lvid=7b674da8235ea642c3bfacf244eeae53; tmr_lvidTS=1720871533984; _ym_uid=1720871534834198811; _ym_d=1720871534; _tt_enable_cookie=1; _ttp=El4ZowXxQWdsnA31_DDrCt_AT_4; popmechanic_sbjs_migrations=popmechanic_1418474375998%3D1%7C%7C%7C1471519752600%3D1%7C%7C%7C1471519752605%3D1; _foxford_cookie_consent=yes; advcake_session_id=2fb2830e-9b68-2529-f9e4-2dec39973ef9; adtech_uid=ba6ca4c8-055a-4f46-b06c-6df09c20e8a4%3Afoxford.ru; top100_id=t1.7729862.2001929594.1723667001899; t3_sid_7729862=s1.1788907302.1725084603508.1725085234295.9.11; _ym_isad=2; mindboxDeviceUUID=9d785862-f8db-4ba8-b7bc-ccd9dbba98e1; directCrm-session=%7B%22deviceGuid%22%3A%229d785862-f8db-4ba8-b7bc-ccd9dbba98e1%22%7D; qrator_jsid=1725084559.622.G14rJEU37yUAYfzF-0l4eiu40m3uudgv95tu3faluhe7up6nv; mindboxSessionIsOpen=true
Priority: u=4
Pragma: no-cache
Cache-Control: no-cache
TE: trailers


  


  ffmpeg returns HTTP error Server returned 401 Unauthorized (authorization failed) :

  


  

  


  Headers from network tab :


  


• Firebase and FFmpeg : Download URL error for videos longer than one hour

  11 août 2024, par braunlich

  Summary of the problem : I'm developing a video processing feature in a React application that uses Firebase as the backend. The video file uploads to Firebase Storage without issues. However, when attempting to use FFmpeg to extract the audio as a FLAC file, the process works for videos under an hour but fails for longer videos, producing the error : "Error : File could not be read ! Code=-1".

  


  Code context : The issue occurs during the process of grabbing the video URL from Firebase Storage and extracting audio using FFmpeg. Here is the code snippet :

  


  try {
const videoDownloadURL = await getDownloadURL(uploadTask.snapshot.ref);
console.log('Video file available at', videoDownloadURL);

await ffmpeg.load();
await ffmpeg.writeFile(file.name, await fetchFile(file));
await ffmpeg.exec(['-i', file.name, '-q:a', '0', '-map', 'a', '-ac', '1', 'output.flac']); // Ensuring mono channel
const audioData = await ffmpeg.readFile('output.flac');
const audioBlob = new Blob([audioData.buffer], { type: 'audio/flac' });
const audioStorageRef = ref(storage, `${user.uid}/${videoId}.flac`);
const audioUploadTask = uploadBytesResumable(audioStorageRef, audioBlob, {
    contentType: 'audio/flac',
    customMetadata: {
        uploadedBy: user.uid,
        videoId: videoId
    }
});

audioUploadTask.on('state_changed',
    (snapshot) => {
        const audioProgress = Math.round((snapshot.bytesTransferred / snapshot.totalBytes) * 100 * 0.1);
        setUploadProgress(Math.round(videoProgress + audioProgress));
        console.log(`Audio upload progress: ${Math.round(audioProgress)}%`);
    },
    (error) => {
        console.error("Audio upload error:", error);
    },
    async () => {
        try {
            const audioDownloadURL = await getDownloadURL(audioUploadTask.snapshot.ref);
            console.log('Audio file available at', audioDownloadURL);
            audioUploadComplete = true;
        } catch (error) {
            console.error("Audio download URL error:", error);
        }
    }
);

} catch (error) { console.error("Download URL error:", error); }


  


  Issue description : The above code successfully extracts audio from videos shorter than one hour. However, for videos around two to three hours long, the code breaks with the following error message :
error code

  


  I suspect this might be related to FFmpeg or Firebase's handling of larger files, but I'm not sure how to debug this further or what adjustments to make.

  


  Research and attempts :

  


  I've confirmed that the video files are successfully uploaded to Firebase Storage.

  


  The issue appears to be related to either the FFmpeg processing or the retrieval of the download URL for larger files.

  


  I've searched for similar issues but haven't found a solution specific to handling larger video files with FFmpeg and Firebase.

  


  Request for help :

  


  Any insights on why this error might be occurring for larger files ? Suggestions for debugging or modifying the code to handle larger videos efficiently.

  


• hwcontext_vulkan : rewrite upload/download

  18 juillet 2024, par Lynne

  hwcontext_vulkan : rewrite upload/download
  This commit was long overdue. The old transfer dubiously tried to
  
  merge as much code as possible, and had very little in the way
  
  of optimizations, apart from basic host-mapping.
  The new code uses buffer pools for any temporary bufflers, and
  
  handles falling back to buffer-based uploads if host-mapping fails.
  Roundtrip performance difference :
  
  ffmpeg -init_hw_device "vulkan=vk:0,debug=0,disable_multiplane=1" -f lavfi \
  
   i color=red:s=3840x2160 -vf hwupload,hwdownload,format=yuv420p -f null -
  7900XTX :
  
  Before : 224fps
  
  After : 502fps
  Ada, with proprietary drivers :
  
  Before : 29fps
  
  After : 54fps
  Alder Lake :
  
  Before : 85fps
  
  After : 108fps
  With the host-mapping codepath disabled :
  
  Before : 32fps
  
  After : 51fps
  

  • [DH] libavutil/hwcontext_vulkan.c
  • [DH] libavutil/vulkan.c