AWS for M&E Blog

A new standard for remote QC and playback

Guest post by Jason Dvorkin (Sr. M&E Partner Solutions Architect, AWS Partner Program) and Scott Sharp (Head of Technology, GrayMeta).

 

Over the past year and during a global health crisis, media companies faced the challenge of connecting a remote workforce to content production and quality control (QC) workflows. Many studios increased the use of cloud-based remote production and post-production services to allow creative artists to access this content securely and from disparate locations. With more high-resolution assets moving to the cloud and the growth of cloud computing, performing quality control (QC) processes needed to change. And, with the move to the cloud, media companies can benefit from the use of artificial intelligence and machine learning (AI/ML) to flag critical moments within the content for review.

GrayMeta, an Amazon Web Services (AWS) Partner, has deep expertise using AI/ML metadata solutions to drive workflow efficiencies that translate into time and cost savings. GrayMeta’s Iris Anywhere product uses AWS compute, storage, and AI/ML services, allowing users to perform remote QC, quality assurance (QA), compliance, content review and approval from anywhere in the world.

QC on AWS

QC is often the start or end of media workflows. GrayMeta’s Iris is trusted by prominent service providers in the industry, including Deluxe Media Inc., Premiere Digital, and others. Iris features a robust set of capabilities, including video scopes, audio tools, and most importantly, the ability to play any type of video file or package in a single software solution, eliminating the need to purchase and maintain costly hardware.

As employees shifted to working from home, content production and post-production moved to the cloud to prevent significant disruption of work. QC needed to do the same. Iris Anywhere connects directly to a customer’s Amazon Simple Storage Solution (Amazon S3) bucket allowing a user to QC and frame-accurately play back high-resolution video assets through a web browser.

GrayMeta Iris Anywhere screenshot with QC tools and audio routing

GrayMeta Iris Anywhere screenshot with QC tools and audio routing

GrayMeta created its own Amazon Machine Image (AMI) that users can install on an Amazon Elastic Compute Cloud (Amazon EC2) instance. Based on the number of concurrent users and the type of play back content, GrayMeta recommends the appropriate type of Amazon EC2 instance to support the workload. As an example, Deluxe uses Iris Anywhere to QC ultra-high definition (UHD) Interoperable Master Format (IMF) packages with up to three concurrent users. For this use case, GrayMeta deployed Iris Anywhere on Amazon EC2 C5 instances. GrayMeta uses AWS Auto Scaling groups to provision workstations based on user demand.

GrayMeta Iris Anywhere workflow on AWS

GrayMeta Iris Anywhere workflow on AWS

Users interact with Iris Anywhere through a web browser, such as Google Chrome, Mozilla Firefox, Microsoft Edge, or Safari, without needing virtual desktop infrastructure (VDI). Iris Anywhere combines standalone QC scopes and tools such as a video waveform monitor, spectrascope, and vectorscope and audio tools like loudness monitor, phase meter, and level meters into a single piece of software. Iris Anywhere also supports the playback of High Dynamic Range (HDR) and UHD with the ability to simulate 4K playback. This eliminates the need for an operator to download the file from Amazon S3 to QC on a broadcast monitor. In addition, Iris Anywhere can display Dolby Vision Content Mapping Unit (CMU) metadata and visualize the composition playlist (CPL) of an IMF package. Dolby Vision CMU metadata and the IMF CPL can be displayed on a timeline, making it faster for an operator to identify the segments and validate the technical specifications of the asset.

Dolby Vision file playback options, including the ability to extract Dolby Vision CMU metadata

Dolby Vision file playback options, including the ability to extract Dolby Vision CMU metadata

Media intelligence meets QC

Today, most media companies have a lot of content. They could be preparing to launch a streaming or over-the-top (OTT) platform, or preparing for a theatrical release. Either way, these companies need a way to focus their operators on the right spots in an asset in order to QC more files in a day. Watching an asset end to end might be necessary for some assets, but by using AI/ML, the attention of an operator can be focused on problematic areas in the content. This can reduce an operator’s review time by 50%.  That’s why GrayMeta added the ability to view metadata generated by Amazon Rekognition and Amazon Transcribe in Iris Anywhere.

Amazon Rekognition is a machine learning based service that can analyze images and videos to detect objects, people, faces, text, scenes, activities, and inappropriate content. Having the metadata from Amazon Rekognition, an operator can quickly see the moment in time where inappropriate content appears on screen. They can quickly review these moments and flag points for an editor to fix. This combines Standards and Practices (S&P) workflows into the QC process.

Amazon Transcribe is an automatic speech recognition service that returns a text file of transcribed speech. Iris Anywhere can display closed captions embedded in a video file, or it can open a side car file. This allows operators to check the sync of the captions with the dialogue. If a video file does not have captions, the transcript created by Amazon Transcribe can be used to check audio synchronization, or to provide to a caption or localization vendor to prepare the video for distribution.

When a user opens an asset in Iris Anywhere, the metadata from Amazon Rekognition and Amazon Transcribe can be imported and displayed in a tool window. The operator has enriched metadata that can speed up their review, and reduce the number of people needed to review a file. Iris Anywhere’s annotation tool lets operators use the enriched metadata to create an edit decision list (EDL) that Compliance Editors can use to make edits to the content and prepare it for broadcast.

Collaboration

Iris Anywhere supports remote collaboration from anywhere in the world. An operator reviewing a file can bring in a second operator by launching an over-the-shoulder review session. One operator playing the asset will update in real time on the other user’s screen. Both users have the ability to control the playback and view the same information in the scopes. Iris Anywhere also allows time-based annotations to determine whether an asset passed or failed the QC process.

Operators can export a detailed QC report containing the technical metadata and annotations directly from Iris Anywhere and include it with the delivery of the media asset. This reduces the time it takes to create additional documentation or the need to pull information from multiple systems.

Customer Benefits

With the release of Iris Anywhere, QC no longer needs to happen within the four walls of a facility. Operators can be anywhere, and can playback media stored in Amazon S3 without time spent downloading an asset. With the ability to leverage metadata generated by Amazon Rekognition and Amazon Transcribe, operators can focus their attention on the key points of an asset, helping an organization save time and reduce operational costs.

GrayMeta is an Advanced AWS Technology Partner that allows companies to transform physical assets to digital, connect to and access content from anywhere, and create valuable metadata using machine learning & AI services. Learn how Iris Anywhere on AWS connects Media Intelligence to your business. Iris Anywhere is available from AWS Marketplace. Learn more about GrayMeta on AWS.

Jason Dvorkin

Jason Dvorkin

Jason is a Senior Industry Specialist BD supporting customers in the Media & Entertainment Vertical for AWS. Jason works with executives and technology owners to address their technical and non-technical transformation challenges, providing domain expertise for cloud adoption and achievement of key business objectives. Jason started his career producing live events and broadcasts for a professional sports team. He then went on to work for multiple tech start-ups in the areas of media asset management and media workflows orchestrion and machine learning and artificial intelligence. When Jason isn’t delivering results customers, he can be found spending time with his wife and two daughters, or on the ski slopes in Colorado.