The Sterowanie24 team has created and published a web-based remote lighting control project and exhibition, and they are using OvenMediaEngine to stream it with low latency.
Their wonderful project has been going on for ten years and is still developing. This is full of fun interactive elements with IP cameras. Turn on/off the lights, change the text on the signboard, make a sound, and more. You can control these elements in Poland with low latency, wherever you are. If you're interested, click HERE to experience low latency interactivity.
Here we go!
Q. What kind of project did you do?
We created a non-commercial project in which we needed an independent web-camera streaming system. This project first started in 2012 and is being improved every year.
Our project allows you to control Christmas lighting via a website and observe effects on a preview of a video broadcast. We operate as Sterowanie24.pl in Poland. This year, we also translated the site for foreign users as Smartpixels.app. And you can be admired the whole with the camera transmitting live images.
Also, our project enjoys increasing popularity year by year, translating into the number of visits to the website. We have entries from virtually all over the world.
Q. How is your project structured?
Our project can be divided into the controller, web application, and live streaming parts.
The web application runs with Node.js – it generates a website with a control panel. And the user page is delivered from the nearest server through Cloudflare. Infrastructure is designed for high availability and the ability to scale quickly.
We have our own video streaming solutions. We use WebRTC and HTTP live Streaming technologies to deliver camera footage to users. As cameras, we usually use consumer surveillance IP cameras that are calibrated to ensure the highest possible image quality.
Q. Why did you choose/use OvenMediaEngine?
We needed a free system with the lowest possible latency. Furthermore, it had to work on both mobile devices (Android and iPhone) and desktops (Windows, Linux, OS X). OvenMediaEngine enabled us to implement such a system. Currently, video delay is around two seconds due to latency generated by IP cameras and image processing. For now, receiving transmissions is done by Nginx-RTMP, transcoding via FFmpeg, but this will change in the future.
Q. What have you achieved with OvenMediaEngine?
Since the beginning of December, we have been delivering a stable web cameras view to our users. In addition, OvenMediaEngine allowed us to deliver 0.5Gbps (peak) in WebRTC for 100 simultaneous users without any slowdowns. Please see the network usage chart we shared below.
Q. What is the good/bad point of OvenMediaEngine that you think?
Our problem is a performance with a large amount of users. However, we created a streaming cluster based on several servers. In addition, we have prepared a proprietary load balancer that distributes users so that the user always goes to the least loaded server.
Q. What improvements are needed for OvenMediaEngine?
We want OvenMediaEngine to provide features like Autoscaling, WebRTC and HLS with ABR, JPG thumbnails generator with some cache, Thumbnails using MP4, RTMP Pull to save bandwidth on edge WebRTC servers, and Some load balancer on the OvenMediaEngine side. Of course, we know that some of the above lists are already included in the roadmap through discussion.
If you are more interested in this project and need more details, please click HERE and HERE to read articles as well. Sadly, their exhibition is only held in winter. So we plan to talk about their event on AirenBlog again in December 2022. So we look forward to seeing how you come back.
Once again, special thanks to fcqpl, and thanks to you for reading our fourth use case.