How Ryff is Inventing Intelligent Images
Our data centric world has changed the way we think of and use information. We do not give a thought to searching for data on absolutely any topic, at any time, anywhere. Once we had to visit libraries, consult experts and read extensively to find information, now it is freely and easily accessible at the press of a few buttons.
However, this is only true of information - so far, it has eluded us for images. It is true that powerful visual computing solutions exist to search through video libraries to search for scenes with cats, certain actors or other criteria. Think of Amazon X-ray or Google Photos for example. While we can search and sometimes find images, we cannot use them the way we use data.
Ryff’s mission is nothing less than to re imagine and reinvent the moving picture. We asked the question what if we could make every image intelligent enough to be dynamic and programmatic? We believe we can, and we have demonstrated it in action.
Together the team at Ryff represents leadership from each of the disciplines needed to develop intelligent images further. Spencer Stephens is our Chief Science Adviser – previously he was CTO at Sony Pictures. Mark Turner is COO and previously worked at Technicolor, Dolby and Microsoft. Evan Smyth, our CTO, worked for Jeffrey Katzenberg at Dreamworks and Nova.
At our recent launch event we announced our partnerships with the worlds largest independent TV production company Endemol Shine and Hollywood's leading marketing agency LA Associates. Separately Indian IT firm Tech Mahindra will partner with us to develop and deliver the Ryff Placer™ platform.
Tomorrows intelligent images
Today, each frame of a film or TV program is constructed using dozens, hundreds or even thousands of different assets and elements. Real-life object information is kept in a manifest file and digital assets have metadata which tells us what they are and where they are stored. Each frame (during construction) is therefore full of immensely valuable image information. For example, which actors are in a scene, where the scene takes place, genre, objects, etc., etc. During the video construction process, each frame can be edited, re-shot, re-composited, etc. – until it is final. Then it is flattened, and all that data is lost forever – this is almost identical to what happens to an image made in Photoshop.
What if the valuable data wasn’t lost when it was flattened?
Each individual asset in a frame is what we call a media element, or a MEL. In this great short video by Roy Peker, he brilliantly shows how scenes can be composited or constructed using different techniques and elements. Ryff has invented a way for those elements to be still be editable all the way to the final viewer. For millions of viewers. Individual to each. Unique to each screen.
Video games ability to deliver unique viewing
In a modern video game this happens automatically. Millions of players enjoy being on the same map at the same time, playing the same game. What each player sees and experiences is a totally unique experience. Of course, what they see is not photorealistic, although Battlefield V is pretty amazing. If this can be done for video games we thought then surely we can deliver individual experiences to TV viewers too.
Ryff has exclusively licensed Cyclops technology from The Mill, part of Technicolor, which enables us to dynamically change assets in a frame to the final viewer . We also work with the folks at Epic Games to use the Unreal Engine to deliver absolutely photo realistic CGI. Film is delivered to us 30 times a second with each frame fixed. Unreal engine delivers us each frame at 90 times a second, or more, at resolutions of 4k above and they are interactive.
Our team is unique in combining expertise from the film, television, video games, animation and rendering in both real time and offline. By combing the best and latest techniques from all of these worlds we are able to deliver images which may be edited, be connected to a URL, updated, removed or inserted into the viewers frame up to and including in real time.
Massive opportunity to redefine product placement
Our first market for intelligent images is product placement. As viewers increasingly move away from traditional TV and interstitial advertising to OTT and VOD, we are seeing a dramatic rise in product placement. In 2017 it grew 14% and this year in North America it will exceed $10 Billion. To quote Loussiné Sarkissian, “Targeting Millennials is no easy task. Their media consumption habits differ from other generations and advertisers have to adapt to reach them. Using product placement seems to satisfy everyone. Brands advertise their products, movie producers and show creators add money to their budget and authenticity to their productions.”
The Ryff Placer™ platform allows for a viewer to see products placed into a scene that suits them, that are of targeted interest to them and which can be clicked on in an OTT service and even given an option to immediately purchase said item. Traditional broadcast content can be customized to insert brands and products in local languages suitable for the target demographic.
More content than ever is reaching record audiences, but yield per asset is dropping. This is because with 487 new TV shows launching in 2018, it is impossible to ensure that even the best traditional advert or product placement reaches your target audience. How can you cover every possible show on every channel? The Ryff Placer™ platform fixes this by allowing you to insert desired products into every show dynamically.
The Ryff Placer™ platform can be demonstrated now.
At our September 12th announcement event in Beverly Hills, we demonstrated the Ryff Placer™ platform in action. Choosing one of 10 available personas we were able to show a scene from our partner Endemol Shines’ ‘Big Brother’ that had 15 different product placements depending on the persona’s profile. A child was shown soda, an adult alcoholic beverages and a Hispanic female a Spanish language product. This was a OTT stream that attendees were able to access via their smartphones to see the Ryff Placer™ platform working live at the event.