Fluendo awarded SPIRIT Open Call 2 grant: Introducing AIVATAR
daas-vdi, gstreamer, codecs, fluendo-codec-pack, fluendo-ai-plugins, remote-desktop-coding-tools, raven, lynxFluendo selected for SPIRIT grant to develop AI-powered avatar telepresence.
Lynx
Dive into the future of DaaS & VDI possibilities with cutting-edge codecs
Discover how we’re pioneering the next generation of Desktop-as-a-Service by integrating LCEVC codecs. Experience real-time interactivity and crystal-clear text quality designed for the most demanding professional environments.
Zero lag. Absolute clarity.
Exploring next-generation codecs, our first steps are towards testing LCEVC. The emphasis is on identifying where this standard can significantly impact, considering the unique video requirements for DaaS. Low latency ensures real-time interactions, which are crucial to a seamless DaaS experience. The extraordinary text quality is vital for readability and effective communication. By aligning these features with specific DaaS use cases, the project aims to unlock the potential of the LCEVC standard to enhance the DaaS user experience and expand its usability.
Zero lag. Absolute clarity.
Exploring next-generation codecs, our first steps are towards testing LCEVC. The emphasis is on identifying where this standard can significantly impact, considering the unique video requirements for DaaS. Low latency ensures real-time interactions, which are crucial to a seamless DaaS experience. The extraordinary text quality is vital for readability and effective communication. By aligning these features with specific DaaS use cases, the project aims to unlock the potential of the LCEVC standard to enhance the DaaS user experience and expand its usability.
What has been done
Evaluating the impact of screen content codecs in the DaaS & VDI Industry
This project compared and fine-tuned different codecs optimized for screen content streaming in the cloud. Using various metrics and tools, we measured the codecs’ bandwidth, quality, CPU usage, and CO2 emissions. Our results show that screen content codecs can reduce the bandwidth and improve the content’s quality while lowering the cloud servers’ CPU usage and CO2 emissions. Our project demonstrates the potential of these technologies in modernizing thin client solutions and contributing to a greener world.
What has been done
Evaluating the impact of screen content codecs in the DaaS & VDI Industry
This project compared and fine-tuned different codecs optimized for screen content streaming in the cloud. Using various metrics and tools, we measured the codecs’ bandwidth, quality, CPU usage, and CO2 emissions. Our results show that screen content codecs can reduce the bandwidth and improve the content’s quality while lowering the cloud servers’ CPU usage and CO2 emissions. Our project demonstrates the potential of these technologies in modernizing thin client solutions and contributing to a greener world.
How we do it
Methodology
- 01
Use Case
Servers with limited cloud CPU can handle 1 or 2 clients each. Limited frame rate (20 fps) and high bitrates (16/32 Mbps) are available.
- 02
Test clips
A broad spectrum of temporal and spatial complex screen content library, as defined in JCTVC-T1015-r1, has been assembled for our tests.
- 03
Metrics
A broad spectrum of temporal and spatial complex screen content library, as defined in JCTVC-T1015-r1, has been assembled for our tests.
- 04
Parameters
We fix quality, latency, and video resolution to maximize our metrics and let bitrate roam free.
What we did
Our achievements
High-Fidelity Text Streaming
Video quality needed to remain constant or even increase during tests to match our customers’ needs. All our current results follow that premise. With the improved bitrate reduction of our codecs, we can also propose quality improvements on text content, a traditional pain point on virtual desktops UX, using a higher chroma subsampling without impacting bitrate or latency. Different testing scenarios and content types have achieved an average BD-Rate reduction of -30.93%.
METRICS
- VMAF
RESULTS
- Reduced bitrate at same quality
- SCC at 1080p
LCEVC Compute Efficiency
The complexity of the LCEVC codec is intrinsically tied to the complexity of the encoder used for its base layer.
For conventional video codecs such as AVC, HEVC, or AV1, coding complexity decreases as the resolution is reduced, assuming other parameters remain constant. We have demonstrated a 20% reduction in compute resource usage over the x264 base codec.
30% Faster Encoding
Our findings suggest that fine-tuning the codec can lead to approximately a 30% reduction in encoding time without compromising bitrate or potential savings. However, the utilization of LCEVC does not automatically result in improvements. Careful consideration of the target content and use case requirements is essential for configuring the codec to achieve maximum quality and efficiency.
LCEVC Compute Efficiency
Our working prototype is a remote desktop application developed using GStreamer for multimedia processing and WebRTC for real-time communication. It has the following parts:
- GStreamer pipeline multimedia management (broadcast a desktop live and receive navigation events from the client)
- WebRTC communication (send/receive video and data channel with navigation events)
- GStreamer elements to evaluate LCEVC: lynxenc & lynxdec.
When testing our application on ultra-low latency settings, we achieved a glass-2-glass latency of ~40ms.
Bits & Bytes
Explore our blog, one byte at a time. Our team unpack our latest news, industry insights and in-depth articles to connect you with the multimedia world.Blog
Read more about our work
MPEG-5 LCEVC in remote desktop streaming
daas-vdi, codecs, remote-desktop-coding-tools, lynxEnhancing FreeRDP performance with MPEG-5 LCEVC for DaaS and VDI.
Video codec optimized for desktop virtualization with LCEVC
daas-vdi, codecs, remote-desktop-coding-tools, lynxOptimizing DaaS/VDI performance with LCEVC and x264.