Logo Logo
directory schraeg
Camcorders· Cinema-Kamera· Computers· Drohnen· GPU· Kamera-Zubehör· Video-DSLR· accessories
Compositing· Color correction· DV Editing

Shooting· Sound
Forschung· KI· Reviews· Streaming
/// News


[10:16 Mon,6.March 2023   by ]    

According to Microsoft, one third of all videos played via the Edge browser only have a resolution of 480p or even less - either because the video was only filmed or available in this resolution or because the available network bandwidth is not sufficient for more data. Low-resolution videos are not nice to look at: they are blurry and texts in them are often difficult to read.

Microsoft and Nvidia now offer - independently of each other - a solution, namely the automatic upscaling of low-resolution videos in real time using AI. The magic word is "Video Super Resolution", a DeepLearning algorithm that runs on the graphics card and not only upscales such videos in real time, but also cleans them of annoying compression artifacts, such as block structures.


This is possible with any video that is played in the browser, for example YouTube, Netflix or another streaming platform. Nvidia&s RTX Video Super Resolution (VSR) had already been announced in January and is now freely available, Microsoft&s Video Super Resolution (VSR) is available via the Edge Beta program. Both require a powerful graphics card, but both have their specific advantages and disadvantages: Microsoft&s algorithm works with both Nvidia&s RTX 20/30/40 and AMD&s RX5700-RX7800 graphics cards, but only works in Microsoft&s Edge browser, Nvidia&s solution (of course) requires an Nvidia&s GPUs, namely one of the RTX 30 or 40 series, but in return not only videos are improved in Edge but also in Google Chrome, the most widely used browser.


Microsoft&s Video Super Resolution

The following additional conditions must also be met for Microsoft&s Video Super Resolution to be enabled:
- The video is playing at a resolution of less than 720p.
- The device is connected to the mains.
- Both the height and width of the video are greater than 192 pixels.
- The video is not protected with Digital Rights Management technologies such as PlayReady or Widevine.

Video Super Resolution is already active on a test basis for just under half of Edge users in the Canary channel (beta) (recognizable by an "HD" icon in the address bar), all others can activate it manually via "edge://flags/#edge-video-super-resolution".

Nvidia&s RTX Video Super Resolution

Nvidia&s RTX Video Super Resolution algorithm works with videos with a native resolution of 360p to 1440p and a frame rate of up to 144Hz in real time. It can not only upscale low-resolution videos but also works with higher-resolution videos. So it can upscale 1080p videos - as seen in the sample clip below - to a resolution of 4K, so they look sharp on a 4K monitor.

For Nvidia&s RTX Video Super Resolution to be active an update to the latest GeForce Game Ready driver, which is available now, or to the next NVIDIA Studio driver, which will be released in March, must be performed. Both Chrome (version 110.0.5481.105 or later) and Edge (version 110.0.1587.56) have recently been updated to support RTX VSR.

To enable the feature, the "Super Resolution" checkbox under "RTX Video Enhancement" must be clicked under the "Customize Video Image Settings" menu item in the NVIDIA Control Panel and a quality from one to four - from the lowest impact on GPU performance to the highest level of upscaling enhancement, must be selected.

What is Super Resolution?

While the rule used to be that when resolution was increased, no detail could be added that was not in the original image, a paradigm shift has been taking place for several years. Thanks to the new super-resolution technology, more detail can be added to an image (or now video) than was originally present. But how is this possible? Super Resolution methods do not use the rather primitive interpolation between known pixels for the "more image information", but invent details conclusively.

Super Resolution Algorithm compared to simple bicubic upscaling

This is made possible by Deep Learning technology, as these AI algorithms are very good at recognizing objects and making sense of them thanks to training with millions of images (or videos). When upscaling, neural networks make use of the typical patterns of natural objects - i.e. structures in the low-resolution images are reconstructed in a meaningful way when upscaling. So, for example, if a corresponding Super Resolution AI has seen millions of faces from different angles and in diverse lighting situations, it can subsequently add learned appropriate details to each coarse resolution-only face in an image. However, this is just as true for plants, cars or animals.

Link more infos at bei www.microsoftedgeinsider.com

deutsche Version dieser Seite: Nie wieder unscharfe Videos? Super Resolution im Browser per kostenloser KI


  Vorige News lesen Nächste News lesen 
bildDas könnte teuer werden: Achtung beim Filmdreh oder Fotoshooting vor Fototapeten bildWeiteres Update - Blackmagic DaVinci Resolve 18.1.4

related news:1E0Inspiring AI image variations at the click of a mouse with Stable Diffusion Reimagine / unCLIP 31.March 2023
Insta360 Flow - smartphone gimbal with AI tracking introduced 29.March 2023
Runway Gen-1 available- expensive AI video generator to get you started 28.March 2023
This is what the AI-generated winning films of the Runway AI Film Festival look like 27.March 2023
How to produce a SF clip in just 3h using AI tools 24.March 2023
Microsoft with generative AI models for everyone in the browser - Bing Image Creator 23.March 2023
Adobe announces Firefly: Generative AI models now directly from Adobe 21.March 2023
alle Newsmeldungen zum Thema KI

[nach oben]

Archiv Newsmeldungen


March - February - January

December - November - October - September - August - July - June - May - April - March - February - January























deutsche Version dieser Seite: Nie wieder unscharfe Videos? Super Resolution im Browser per kostenloser KI

last update : 31.März 2023 - 18:02 - slashCAM is a project by channelunit GmbH- mail : slashcam@--antispam:7465--slashcam.de - deutsche Version