Recherche avancée

Sur d’autres sites (5916)

• How to improve web camera streaming latency to v4l2loopback device with ffmpeg ?

  11 mars, par Made by Moses

  I'm trying to stream my iPhone camera to my PC on LAN.

  


  What I've done :

  


  


  1. HTTP server with html page and streaming script :

     


     I use WebSockets here and maybe WebRTC is better choice but it seems like network latency is good enough

     


  


  


  async function beginCameraStream() {
  const mediaStream = await navigator.mediaDevices.getUserMedia({
    video: { facingMode: "user" },
  });

  websocket = new WebSocket(SERVER_URL);

  websocket.onopen = () => {
    console.log("WS connected");

    const options = { mimeType: "video/mp4", videoBitsPerSecond: 1_000_000 };
    mediaRecorder = new MediaRecorder(mediaStream, options);

    mediaRecorder.ondataavailable = async (event) => {
      // to measure latency I prepend timestamp to the actual video bytes chunk
      const timestamp = Date.now();
      const timestampBuffer = new ArrayBuffer(8);
      const dataView = new DataView(timestampBuffer);
      dataView.setBigUint64(0, BigInt(timestamp), true);
      const data = await event.data.bytes();

      const result = new Uint8Array(data.byteLength + 8);
      result.set(new Uint8Array(timestampBuffer), 0);
      result.set(data, 8);

      websocket.send(result);
    };

    mediaRecorder.start(100); // Collect 100ms chunks
  };
}


  


  


  2. Server to process video chunks

     


  


  


  import { serve } from "bun";
import { Readable } from "stream";

const V4L2LOOPBACK_DEVICE = "/dev/video10";

export const setupFFmpeg = (v4l2device) => {
  // prettier-ignore
  return spawn("ffmpeg", [
    '-i', 'pipe:0',           // Read from stdin
    '-pix_fmt', 'yuv420p',    // Pixel format
    '-r', '30',               // Target 30 fps
    '-f', 'v4l2',             // Output format
    v4l2device, // Output to v4l2loopback device
  ]);
};

export class FfmpegStream extends Readable {
  _read() {
    // This is called when the stream wants more data
    // We push data when we get chunks
  }
}

function main() {
  const ffmpeg = setupFFmpeg(V4L2LOOPBACK_DEVICE);
  serve({
    port: 8000,
    fetch(req, server) {
      if (server.upgrade(req)) {
        return; // Upgraded to WebSocket
      }
    },
    websocket: {
      open(ws) {
        console.log("Client connected");
        const stream = new FfmpegStream();
        stream.pipe(ffmpeg?.stdin);

        ws.data = {
          stream,
          received: 0,
        };
      },
      async message(ws, message) {
        const view = new DataView(message.buffer, 0, 8);
        const ts = Number(view.getBigUint64(0, true));
        ws.data.received += message.byteLength;
        const chunk = new Uint8Array(message.buffer, 8, message.byteLength - 8);

        ws.data.stream.push(chunk);

        console.log(
          [
            `latency: ${Date.now() - ts} ms`,
            `chunk: ${message.byteLength}`,
            `total: ${ws.data.received}`,
          ].join(" | "),
        );
      },
    },
  });
}

main();


  


  After I try to open the v4l2loopback device

  


  cvlc v4l2:///dev/video10


  


  picture is delayed for at least 1.5 sec which is unacceptable for my project.

  


  Thoughts :

  


  

  • Problem doesn't seems to be with network latency

  


  


  latency: 140 ms | chunk: 661 Bytes | total: 661 Bytes
latency: 206 ms | chunk: 16.76 KB | total: 17.41 KB
latency: 141 ms | chunk: 11.28 KB | total: 28.68 KB
latency: 141 ms | chunk: 13.05 KB | total: 41.74 KB
latency: 199 ms | chunk: 11.39 KB | total: 53.13 KB
latency: 141 ms | chunk: 16.94 KB | total: 70.07 KB
latency: 139 ms | chunk: 12.67 KB | total: 82.74 KB
latency: 142 ms | chunk: 13.14 KB | total: 95.88 KB


  


   150ms is actually too much for 15KB on LAN but there can some issue with my router

  


  

  • As far as I can tell it neither ties to ffmpeg throughput :

  


  


  Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'pipe:0':
  Metadata:
    major_brand     : iso5
    minor_version   : 1
    compatible_brands: isomiso5hlsf
    creation_time   : 2025-03-09T17:16:49.000000Z
  Duration: 00:00:01.38, start:
0.000000, bitrate: N/A
    Stream #0:0(und): Video: h264 (Baseline) (avc1 / 0x31637661), yuvj420p(pc), 1280x720, 4012 kb/s, 57.14 fps, 29.83 tbr, 600 tbn, 1200 tbc (default)
    Metadata:
      rotate          : 90
      creation_time   : 2025-03-09T17:16:49.000000Z
      handler_name    : Core Media Video
    Side data:
      displaymatrix: rotation of -90.00 degrees

Stream mapping:
  Stream #0:0 -> #0:0 (h264 (native) -> rawvideo (native))

[swscaler @ 0x55d8d0b83100] deprecated pixel format used, make sure you did set range correctly

Output #0, video4linux2,v4l2, to '/dev/video10':
  Metadata:
    major_brand     : iso5
    minor_version   : 1
    compatible_brands: isomiso5hlsf
    encoder         : Lavf58.45.100

Stream #0:0(und): Video: rawvideo (I420 / 0x30323449), yuv420p, 720x1280, q=2-31, 663552 kb/s, 60 fps, 60 tbn, 60 tbc (default)
    Metadata:
      encoder         : Lavc58.91.100 rawvideo
      creation_time   : 2025-03-09T17:16:49.000000Z
      handler_name    : Core Media Video
    Side data:
      displaymatrix: rotation of -0.00 degrees

frame=   99 fps=0.0 q=-0.0 size=N/A time=00:00:01.65 bitrate=N/A dup=50 drop=0 speed=2.77x
frame=  137 fps=114 q=-0.0 size=N/A time=00:00:02.28 bitrate=N/A dup=69 drop=0 speed=1.89x
frame=  173 fps= 98 q=-0.0 size=N/A time=00:00:02.88 bitrate=N/A dup=87 drop=0 speed=1.63x
frame=  210 fps= 86 q=-0.0 size=N/A time=00:00:03.50 bitrate=N/A dup=105 drop=0 speed=1.44x
frame=  249 fps= 81 q=-0.0 size=N/A time=00:00:04.15 bitrate=N/A dup=125 drop=0 speed=1.36
frame=  279 fps= 78 q=-0.0 size=N/A time=00:00:04.65 bitrate=N/A dup=139 drop=0 speed=1.31x


  


  


  • I also tried to write the video stream directly to video.mp4 file and immediately open it with vlc but it only can be successfully opened after 1.5 sec.

    


  


  • I've tried to use OBS v4l2 input source instead of vlc but the latency is the same

    


  


  


  Update №1

  


  When i try to stream actual .mp4 file to ffmpeg it works almost immediately with 0.2sec delay to spin up the ffmpeg itself :

  


  cat video.mp4 | ffmpeg -re -i pipe:0 -pix_fmt yuv420p -f v4l2 /dev/video10 & ; sleep 0.2 && cvlc v4l2:///dev/video10


  


  So the problem is apparently with streaming process

  


• Enterprise web analytics : Quick start guide (and top tools)

  10 juillet, par Joe — Analytics Tips

  Without data, you’ll get lost in the sea of competition.

  This is even more important for large organisations.

  Data helps you :

  • Optimise customer experiences
  • Navigate complex business decisions
  • Create a roadmap to sustainable brand growth
  • Data can power differentiation, especially within fiercely competitive sectors.

  How do you get the benefits of data in a large organisation ?

  Enterprise web analytics.

  In this guide, we’ll cover everything you need to know about enterprise web analytics to enhance website performance, improve customer experiences and increase conversions.

  What is enterprise web analytics ?

  Enterprise web analytics help large organisations capture, analyse, and act on website data to optimise customer experiences and make informed decisions. By providing insight into customer interactions, user behaviour and preferences, they’re vital in helping big businesses improve their websites.

  

  Enterprise web analytics can extract data from web pages and reveal a range of performance metrics, including :

  • Pageviews
  • Average time on page
  • Actions per visit
  • Bounce rate
  • Conversions
  • Traffic sources
  • Device type
  • Event tracking
  • And more

  You can track this data daily or access monthly reports, which will give you valuable insights into optimising user engagement, improving your website’s search engine traffic, and meeting business goals like increased conversion rates.

  For large organisations, web analytics isn’t just about measuring traffic. Instead, it’s an asset you can use to identify issues in your web strategy so you can gain insights that will fuel sustainable business growth.

  An advanced analytics strategy goes beyond the digital channels, page views and bounce rates of traditional analytics.

  Instead, modern web analytics incorporates behavioural analytics for deeper analysis and insight into user experiences. These advanced features include :

  • Heatmaps (or scroll maps) to track scroll behaviour on each page
  • User flow reports to see the pages your users visit in the customer journey
  • Session recordings to analyse user interactions (step-by-step)

  Taking a two-pronged approach to web analytics that includes both traditional and behavioural metrics, organisations get a clearer picture of users and their brand interactions.

  Different needs of enterprise companies

  Let’s dive deeper into the different needs of enterprise companies and how enterprise web analytics can help solve them :

  Access more storage

  Let’s face it. Large organisations have complex IT infrastructures and vast amounts of data.

  The amount of data to capture, analyse and store isn’t slowing down anytime soon.

  Enterprise web analytics can help handle and store large amounts of data in ways that serve the entire organisation.

  Enable cross-organisational data consumption

  It’s one thing to access data in a small company. You’ve got yourself and a few employees. That’s easy.

  But, it’s another thing to enable an organisation with thousands of employees with different roles to access complex data structures and large amounts of data.

  Enterprise web analytics allows big companies to enable their entire workforce to gain access to the data they need when they need it.

  Increase security

  As mentioned above, large organisations can use enterprise web analytics to help hundreds or even thousands of employees access their web data.

  However, some data shouldn’t be accessed by every type of employee. For example, some organisations may only want certain data accessed by executives, and some employees may not need to access certain types of data that may confuse or overwhelm them.

  Enterprise web analytics can help you grant access to certain types of data based on your role in the company, ensuring the security of sensitive data in your organisation.

  Improve privacy

  You can keep your data secure from internal breaches with enterprise web analytics. But, how do you protect customer data ?

  With all-inclusive privacy measures.

  To ensure that your customers’ privacy and data are protected, choose a web analytics solution that’s compliant with the latest and most important privacy measures, such as GDPR, LGPD and CCPA.

  Taking a privacy-first approach to data helps ensure your protection from potential legal action or fines.

  Enterprise web analytics best practices

  Want to make sure you get the most out of your web analytics strategy ?

  

  Be clear on what metrics you want to track

  You can track a ton of data in your organisation, but you may not need to. To ensure you’re not wasting time and resources tracking irrelevant numbers, you should make sure you’re clear from day one on the metrics you want to track.

  Start by making a list of key data points relevant to your business.

  For example, if you have an online marketplace, you’ll want to track specific ecommerce metrics like conversion rate, total visits, bounce rates, traffic source, etc.

  Don’t take data at face value

  Numbers alone can’t tell you the whole story of what’s happening in your organisation. It’s crucial you add context to your data, no matter what.

  Dozens of factors could impact your data and visitors’ interactions with your site, so you should always try to look beyond the numbers to see if there are other factors at play.

  For example, you might see that your site traffic is down and think your search engine optimisation (SEO) efforts aren’t working. Meanwhile, there could have been a major Google algorithm update or some sort of seasonality in a key market.

  On the other hand, you might see some positive signals that things are going well with your organic social media strategy because you saw a large influx of traffic from Instagram. But, there could be more to the story.

  For example, an Instagram influencer with five million followers may have just posted a reel reviewing your product or service without you knowing it, leading to a major traffic spike for your website.

  Remember to add notes to your web analytics data if necessary to ensure you can reference any insights from your data to maintain that point of context.

  Ensure your data is accurate

  With web analytics, data is everything. It will help you see where your traffic is coming from, how your users are behaving, and gain actionable insights into how you can improve your website and user experience.

  But if your data isn’t accurate, your efforts will be futile.

  Accurate data is crucial for launching an effective web analytics strategy. Data sampling and simple tracking errors can lead to inaccurate numbers and misleading conclusions. 

  If a tool relies on cookies to collect data, then it’s relying on a faulty data collection system. Cookies give users the option to opt out of tracking, making it challenging to get a clear picture of every user interaction.

  For example, some platforms like Google Analytics use data sampling to make predictions about traffic rather than relying on accurate data collection, leading to inaccurate numbers and conclusions.

  To ensure you’re making decisions based on accurate data, find a solution that doesn’t rely on inaccurate data collection methods like data sampling or cookies.

  Lean on visual data tools to improve analysis

  Enterprise organisations deal with a ton of data. There are endless data points to track, and it can be easy to lose track of what’s going on with the bigger picture.

  One of the best ways to interpret your data is to use a data visualisation tool to integrate with your web analytics solution, like Looker or PowerBI.

  Make sure your chosen platform lets you export your data easily so you can link it with a visual support tool.

  With Matomo, you can easily export your data into Google BigQuery to warehouse your customer data and visualise it through other tools (without the need for APIs, scripts or additional tools).

  Use advanced web analytics

  Web analytics is quite broad, and different tools will offer various features you can access in your analytics dashboard.

  Take advantage of advanced features that utilise both traditional and behavioural data for deeper insights.

  • Use heatmaps to better understand what parts of your web pages your visitors are focusing on to improve conversion rates.
  • Review session recordings to see the exact steps your customers take as they interact with your website.
  • Conduct A/B tests to see which call to action, headline, or image provides the optimal user experience.

  There are dozens of advanced features available, so take the time to make sure your chosen tool has everything you need.

  Choose a privacy-focused tool

  Obviously, not every tool is created equal, and most of the software on the market isn’t suitable for enterprise businesses.

  As a large organisation, the most important step is to choose a trusted enterprise web analytics tool to ensure it’s capable of fitting within a company of your size.

  It needs to have great infrastructure and be able to handle large amounts of data.

  Another crucial factor is to check that the tool is compatible with your website or app. Does it integrate easily with it ? What about your other software ? Will it integrate with those as well and fit into your current tech stack ?

  Most importantly, you need a platform that can provide the data and insights your organisation needs.

  Make sure the tool you choose is GDPR-compliant and privacy-friendly. The last thing you want is to be sued or fined because you chose the wrong software. 

  Consumers are growing more cautious about privacy and data risks, so picking a privacy-focused tool will help build trust with customers.

  Top 5 enterprise web analytics tools

  Now that you understand enterprise web analytics and how to get the most out of it, it’s time to talk about tools.

  You need to make sure you’re using the right web analytics software to improve productivity, optimise website performance and grow your brand without compromising on the infrastructure required for large organisations to thrive.

  Here are five of the best enterprise solutions available :

  Features and pricing comparison

  	GDPR compliant	On-premise option	100% data ownership	Traditional analytics	Behavioural analytics	Awarded best enterprise software
  Matomo	✔️	✔️	✔️	✔️	✔️	✔️
  Amplitude	✔️	✔️			✔️
  Adobe	✔️			✔️	✔️
  GA360				✔️
  Contentsquare	✔️	✔️			✔️	✔️

  Use Matomo to power your website analytics

  Web analytics help enterprise organisations reach new users, improve engagement with current users or grow their web presence.

  These advanced solutions support cross-organisational data consumption, enhance data privacy and security and allow brands to create the web experiences they know customers will love.

  

  Matomo can help you unlock the potential of your website strategy with traditional and behavioural analytics and accurate data. Trusted by over 1 million websites, Matomo’s open-source software is an ethical web solution that helps organisations of all sizes improve decision-making and customer experiences without compromising on privacy or security.

  Start your free 21-day trial now. No credit card required.

• flutter_ffmpeg is discontinued with no alternatives ?

  13 mars, par Rageh Azzazy

  Based on this article by Taner Sener
https://tanersener.medium.com/saying-goodbye-to-ffmpegkit-33ae939767e1

  


  and after the discontinuation of flutter_ffmpeg and ffmpeg_kit_flutter packages.

  


  most packages for executing video editing commands were built depending on them and so won't work After April 1, 2025 as mentioned in the package documentation and Taner's article.

  


  example of packages depending on flutter_ffmpeg or ffmpeg_kit_flutter like

  


  

  • video_trimmer

  


  • zero_video_trimmer

  


  • flutter_video_trimmer

  


  • video_trim

  


  • bemeli_compress

  


  • video_trimmer_pro

  


  • ... others

  


  


  and editing video using video_editor or video_editor_2 or video_editor_pits has become a problem

  


  and I believe downloading the binaries and doing the whole thing locally is not just tedious but illegal as well.

  


  so I broke down exactly what I need to edit videos in my flutter app

  


  and I found those package but not yet tested them

  


  Trimming : flutter_native_video_trimmer

  


  Compression : video_compress or video_compress_plus

  


  Muting : video_compress handles this

  


  Encoding : Not sure, I just need a simple MP4 video files

  


  Cropping : I can't find a solution for video cropping

  


  I don't need to rotate, reverse or do any fancy stuff to the videos, just those five functions.

  


  so now only remaining solution for this approach is to find a simple video cropping function and an encoder that work on flutter IOS & Android

  


  so my question is not looking for external library recommendation but
do you have any ideas how to crop a video in flutter natively ?