The Vu VOD Upscaler is a breakthrough technology to specifically enrich the video and audio playback for VOD (Video on Demand) content like Netflix, YouTube, Amazon Prime, and others. High-production VOD content is made with the latest in camera and VFX technology, but in most cases, viewers do not get optimum experience because of internet speed and quality constraints. Vu has developed proprietary upscaling technology that combines multiple advanced machine learning algorithms working in parallel with a state-of-the-art DSP (digital signal processor) to provide the best video/audio playback for VOD content. This gives high-end, cinema-quality playback for VOD content.
Vu stealth launched the VOD Upscaler technology with its Premium Android range of TVs. In addition, more than 3,000 units have been installed in the month of March. Vu has received raving reviews for its picture and audio quality – some even rating it better than Samsung and LG. Vu has the legacy of information technology which gives it the unique capability to create a technology like VOD Upscaler, which converges information technology into televisions. The VOD Upscaler technology has been in development for more than 3 years, with 107 man-years of engineering effort to create it.
What is the Vu VOD Upscaler?
The Vu VOD Upscaler is a combination of software algorithms cutting edge digital signal processing technology that gives cinema-quality video and audio playback for VOD content which is highly compressed in order to transmit it over the Internet. What is compression and why does it lead to a drop in picture quality? One second of VOD content shot using the latest in camera and post-production technology generates data that’s 1,900 megabytes in size (as compared normal cable/dish video which is 80 megabytes for 1 second of video). In order to stream it over the Internet, it's compressed to a size of 3 megabytes. The way it does it is by reducing detail – for example, there are millions of shades of blue within a 4 billion color palette. Compression reduces it to a few thousand. Standard televisions just reproduce these colors without attempting to recreate the original detail. The same goes for factors like brightness, saturation, and hue. This also creates artifacts, blocks, and jitters in the image.
What does the Vu VOD Upscaler do?
The VOD Upscaler uses a number of different image processing algorithms combined with some machine learning techniques to make the video appear with the same quality as it was before compression. For doing this its using a number of different algorithms like motion-compensated interpolation, intelligent color, and brightness codebook lookups – all working in parallel. For example, when reproducing colour, it takes a compressed color pixel, analyses the changes to that pixel over the last 60 frames and then refers to an internal database of 4 billion colours and guesses the best possible color pixel to use. The same goes for brightness – standard TVs allow the user to increase the brightness by increasing white balance which makes the brightness the same unnatural. The Vu VOD upscaler again is analyzing data over a large number of previous frames, looking at the color being reproduced and then setting the brightness at a pixel level rather than just increasing white balance. Lastly, the VOD upscaler uses a number of denoising filters to remove blocks and artifacts that form. In short the VOD upscaler is doing 200 times more work on every video frame as compared to a standard TV.
What does the Vu VOD upscaler do for audio?
The VOD upscaler helps produce better quality audio in 2 ways. Using advanced algorithms it can differentiate between vocal sounds and music/bass effects and then give the user the control to choose whether they would prefer vocal emphasis or music/bass emphasis for that particular video. Secondly, it implements a number of ambient noise cancellation techniques that you see in high-end sound systems that cost a lot of money that gives a crystal clear sound at high volume levels without crackling or interference.