Cinegy Daniel2 GPU Codec Enters Public Beta
Cinegy has been teasing audiences at tradeshows with 8K recording and playback demos, highlighting the performance advantages of its new NVIDIA GPU accelerated video codec, Daniel2.
The last public Daniel2 teaser occurred at NAB 2017 in April, where Cinegy showed a dual-core 13-inch MacBookPro connected via Thunderbolt3 to an external Akitio Node “eGPU“ box with an NVIDIA GTX1080 inside, which was attached to a brand new Dell Ultrasharp UP3218K 8K monitor. This setup played 8K @ 60fps - or when vsync was turned off, at up to 180fps. Another new feature that Cinegy has added to its bag of tricks is that during the realtime playback, 3D LUT-based color profiles can be applied. This provides initial professional color grading on the fly, which is not meant to replace packages such as BMD Resolve or Adobe Speedgrade but is a very useful feature to get an immediate impression of how a piece of video could look after post production.
This works on any Windows PC or notebook with a high-end NVIDIA graphics card. An 8K monitor is not required. ”Cinegy Player 3” can scale in and out as well as pan and zoom during playback, so HD and UHD monitors can be used for 8K playback. If attached screens support 10-bit color fidelity, this can also be enabled using NVIDIA consumer cards.
The teasing is finally over because Cinegy today announced that it has entered public beta with Daniel2 GPU codec-based applications. Two applications are now available for free testing from Cinegy’s Daniel2 website at www.Daniel2.com – Cinegy Player 3 and an Adobe Premier export plugin, so users can create their own HD, UHD/4K or 8K output in the Daniel2 format. For those who want to try 8K playback immediately there is a small selection of 4K and 8K files on the Daniel2 website for immediate download.
Cinegy co-founder and CTO Jan Weigner said, “Cinegy’s Daniel2 GPU codec is the world fastest professional video codec, leaving any other codec light years behind. Daniel2 can decode up to 1100 frames per second at 8K, or 7680x4320 pixels to be precise, which is 16x the resolution of full HD - translating into more than 17000 frames of full HD decoded per second.”
Cinegy also has an ingest tool for live 8K recording, which was first presented at IBC 2016. It enables 60fps 8K recording and simultaneous playback of one, or multiple 8K streams while performing realtime effects, color correction, scaling and titling. The Daniel2-powered Cinegy Player 3 allows jog and shuttle, scrubbing, zoom and pan while playing at 60fps 8K or faster, again while still recording. This makes it perfect for a number of applications, the foremost being instant replay for sports. Daniel2’s bitrates and PSNR are comparable to those of AVID’s DNxHR or Apple’s ProRes.
Weigner added, “This performance secret of the Daniel2 codec is that it was architected and developed from the ground up to be GPU based. All legacy dependencies and old codec architectures were discarded. This means that Daniel2 is only compatible to itself, but the benefits of this approach are enormous and outweigh the drawbacks by far.”
The Daniel2 codec is an acquisition and production codec designed for recording from camera sources, editing and post-production as well as playout. Daniel2 is targeted at the same space in the production workflow as AVID’s DNxHR, Apple’s ProRes or Sony’s XAVC.
Weigner concluded, “A problem one faces when designing 4K, 8K or, soon, 16K systems that need to handle and manipulate multiple streams in realtime, is that even if you could decode the streams using the CPU, you probably still want to use the power of the GPU for effects and filters, which creates a system bus bottleneck to transfer the decoded streams into the GPU’s memory. This is where Daniel2 shines because it streams a fraction of the size of uncompressed counterparts are read from disk or via the network and passed to the GPU to be decompressed faster than the uncompressed frames can be copied.
“So the Daniel2-enabled Cinegy Player3 provides three wins at once: the system bus uses far less bandwidth; less space or bandwidth is consumed on disk or network; and the CPU is free to do other tasks as it no longer needs to decode multiple streams.”