SlideShare a Scribd company logo
Scalable Media Processing
David Sayed
Principal Product Manager, Amazon Elastic Transcoder
Amazon Web Services

© 2013 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified, or distributed in whole or in part without the express consent of Amazon.com, Inc.
Agenda
•
•
•
•
•
•
•

Media workflows
Where AWS fits
Scalability
Cloud media processing approaches
Amazon Elastic Transcoder
Visual effects example
Broadcast playout to web publishing example
“I think cinema, movies and
magic have always been
closely associated. The very
earliest people who made
film were magicians.”
— Francis Ford
Coppola
Between This Glass…
… And This Glass
Media Workflows: The Real Magic
Media Workflows: The Real Magic

Insert your
favorite
workflows
here
Media Workflows: The Real Magic

Archive
Networks

Insert a few
more
workflows
here

Theatrical
DVD/BD
Online

MSOs
Mobile Apps
Media Workflows: The Real Magic

Archive
Networks

Insert a few
more
workflows
here

Theatrical
DVD/BD
Online

MSOs
Mobile Apps
Media Workflows: The Real Magic
Featurettes

Interviews
2D Movie
3D Movie
Archive
Materials
Stills

Archive
Networks

Insert even
more
workflows
here

Theatrical
DVD/BD
Online

MSOs
Mobile Apps
Where AWS Fits Into Media Processing

Compute &
Networking

Application
Services

Database
Storage & Content
Delivery

Deployment &
Management
Compute &
Networking

Playbac
k

Application
Services

Database
Storage & Content
Delivery

Track

Auth.

Protect

Package

QC

Process

Index

Ingest

Where AWS Fits Into Media Processing

Deployment &
Management
Where AWS Fits Into Media Processing
Analytics and Monetization

Amazon Web Services

Playbac
k

Track

Auth.

Protect

Package

QC

Process

Index

Ingest

Media Asset Management
Infrastructure Cost

Scalability

Unable to
serve
customers

Large
Capital
Expenditure

Predicted
Demand

Opportunity
Cost

Traditional
Hardware
Actual
Demand
Cloud
Computing

Time
Scalability, Is It Really That Important?

The business wants the new
library of content they just
signed on the VOD site
yesterday.
Scalability, Is It Really That Important?

A major event is taking place
and will be live streamed to a
worldwide audience.
Scalability, Is It Really That Important?

A second screen app ties-in
with a prime time event and
may have millions of
simultaneous users.
Infrastructure Cost

Scalability == Agility

Unable to
serve
customers

Large
Capital
Expenditure

Predicted
Demand

Opportunity
Cost

Traditional
Hardware
Actual
Demand
Cloud
Computing

Time
Media Processing Approaches

3 Phases
Cloud Media Processing Approaches: Phase 1

Phase 1
• Lift processing
from on-premise
and shift to the
cloud
Cloud Media Processing Approaches: Phase 2
Phase 2
Phase 1
• Lift processing
from on-prem and
shift to the cloud

• On-prem and cloud
hybrid model
• Refactor workflow
and cloud-optimize
Cloud Media Processing Approaches: Phase 3
Phase 3
Phase 2
Phase 1
• Lift processing from
on-prem and shift
to the cloud

• On-prem and cloud
hybrid model
• Refactor workflow
and cloud-optimize

• Cloud-native
architecture
A Transcoding Example

“Lift and Shift”
• Run existing
transcoding
software on
AWS
A Transcoding Example
Hybrid and
Refactor
“Lift and Shift”
• Run existing
transcoding software
on AWS

• Run cloud versions of
existing transcoding
software
• E.g. Digital Rapids,
Elemental, Harmonic,
Sorenson, Telestream
A Transcoding Example
Cloud Native
Hybrid and Refactor
“Lift and Shift”
• Run existing transcoding
software on AWS

• Run cloud versions of
existing transcoding
software
• E.g. Digital Rapids,
Elemental, Harmonic,
Sorenson, Telestream

• Managed cloud
transcoding service
• E.g. Amazon Elastic
Transcoder,
Encoding.com, Zencoder
Amazon Elastic Transcoder
• Built in response to customer demand
• Designed for file-based high volume transcoding
• Simple, low-cost pricing starting at 1.5 cents per
minute of output
• Focused on OTT delivery formats
Amazon Elastic Transcoder Features
•
•
•
•
•
•

Fully managed cloud transcoding service
Elastically scalable
Output to MP4, WebM and HLS
Integration across AWS services
Available in multiple AWS regions
API and SDK driven
How To Scale With Elastic Transcoder
• Architect around Elastic Transcoder
• Use transcoding pipelines
Amazon Elastic Transcoder Block Diagram
Transcoding Pipeline 1
Amazon S3
Output Bucket

Amazon S3
Input Bucket

Transcoding Pipeline 2
Amazon S3
Output Bucket

Amazon S3
Input Bucket

Transcoding Pipeline 4
Amazon S3
Output Bucket

720p

Transcoding Job n

Kindle Fire

Transcoding Job 2

Custom

Transcoding Job 1
How To Scale With Elastic Transcoder
• Architect around Elastic Transcoder
• Use transcoding pipelines
• Use notifications
Cloud Media Processing Summary
• Three different ways to transcode in the cloud
1. Lift and Shift
2. Hybrid and/or refactor
3. Cloud-native

• The right answer? It depends on your scenario
A Visual Effects Example
Visual Effects in Cloud
“Atomic Fiction has figured out how to use cloud
computing to its advantage, recently delivering 400
shots in only four months for Flight, Zemeckis’
return to live action. It’s a business model that’s
being viewed as the future of the industry.”
Bill Desowitz
Owner ImmersedInMovies.com
Contibuting Editor to Animation Scoop
<video>
Visual Effects Summary
•
•
•
•

High-resolution pre-release material
Only pay for what you use
Elasticity: scalability when you need it
Standard visual effects workflow using ZYNC
plug-ins with Maya, Mental Ray and Nuke
A Broadcaster Example:
Playout to Web Publishing
Project Objective
• Get broadcast content to VOD player ASAP
• Enable editorial clipping and review
• Create many different renditions for output
devices
• Elastically scale for high-demand situations (e.g.
breaking news and sporting events)
• Deliver agility for different types of processing
Playout To Web Workflow
SDI Feed

Amazon Elastic
Transcoder

Broadcast
Encoder
Business
Rules

Program
Data Feed

Live Ingest
Logic

Amazon S3
Re-assembled
Mezzanine

Elemental
Transcoder

DRM
QC

Editorial
Clipping
MAM

Amazon S3
Distribution
Renditions
Best Practices For This Project
•
•
•
•
•
•

Componentize, decouple and be stateless
Design for failure
If data is in motion, use queues
Run at least 3 of everything
Log everything
Use frameworks where possible
Project Results
• Quicker turnaround time
• Only pay for what they use
• Ability to support greater number of target
platforms
• Extremely resilient
• Better control over assets
Parting Thoughts
• Using AWS for scalable media processing is a
reality
• Different ways to architect:
– Lift and shift
– Hybrid and refactor
– Cloud-native

• Start now
More Information
aws.amazon.com/digital-media
aws.amazon.com/architecture
aws.amazon.com/training/architecture
aws.amazon.com/elastictranscoder
www.linkedin.com/in/davidsayed

More Related Content

What's hot (8)

PDF
8KMiles Cloud Solutions
8KMiles Software Services
 
PDF
Aws 101 A walk-through the aws cloud (2013)
Martin Yan
 
PDF
AWS reinvent 2019 - Transforming to feature-driven development
Bhuvaneswari Subramani
 
PDF
Preparing your IT infrastructure for thanksgiving
8KMiles Software Services
 
PDF
APAC Webinar Apr 10 - Architecting your Mobile App for the Cloud
8KMiles Software Services
 
PDF
Artem Zhurbila - 1 aws overview
Artem Zhurbila
 
PPTX
AWS 101 and the benefits of Migrating to the Cloud
CloudHesive
 
PDF
Fundamentals of Cloud Computing & AWS
Bhuvaneswari Subramani
 
8KMiles Cloud Solutions
8KMiles Software Services
 
Aws 101 A walk-through the aws cloud (2013)
Martin Yan
 
AWS reinvent 2019 - Transforming to feature-driven development
Bhuvaneswari Subramani
 
Preparing your IT infrastructure for thanksgiving
8KMiles Software Services
 
APAC Webinar Apr 10 - Architecting your Mobile App for the Cloud
8KMiles Software Services
 
Artem Zhurbila - 1 aws overview
Artem Zhurbila
 
AWS 101 and the benefits of Migrating to the Cloud
CloudHesive
 
Fundamentals of Cloud Computing & AWS
Bhuvaneswari Subramani
 

Viewers also liked (15)

PDF
AWS re:Invent - Med305 Achieving consistently high throughput for very large ...
asperasoft
 
PPTX
Aspera - Bridging On Premise and Cloud Deployments for Broadcast IT
François Quereuil
 
PPT
Aspera Solution Overview - IBM Software
Sebastian Osterc
 
PDF
The Pandora Security Model - Alessandro Confetti
Data Driven Innovation
 
PPTX
Diventare famosi con lo stack ELK - Alfonso Iannotta
Data Driven Innovation
 
PDF
OneLeap Solutions (1)
Amol Shenvi
 
PDF
Telestream Vidchecker
Benoît Godard
 
PDF
TELESTREAM Vantage - VIDELIO Cap'Ciné
Benoît Godard
 
PPTX
L'architettura di classe enterprise di nuova generazione - Massimo Brignoli
Data Driven Innovation
 
PPTX
Automation in Post-Production — Boris Polyak for NATEXPO 2016
Boris Polyak
 
PDF
Machine Learning Real Life Applications By Examples - Mario Cartia
Data Driven Innovation
 
PPTX
SoftLayer Storage Services Overview
Michael Fork
 
PPTX
Aspera on demand for AWS (S3 inc) overview
Bhavik Vyas
 
PPTX
Polyglot Persistence e Big Data: tra innovazione e difficoltà su casi reali -...
Data Driven Innovation
 
PDF
The Top Skills That Can Get You Hired in 2017
LinkedIn
 
AWS re:Invent - Med305 Achieving consistently high throughput for very large ...
asperasoft
 
Aspera - Bridging On Premise and Cloud Deployments for Broadcast IT
François Quereuil
 
Aspera Solution Overview - IBM Software
Sebastian Osterc
 
The Pandora Security Model - Alessandro Confetti
Data Driven Innovation
 
Diventare famosi con lo stack ELK - Alfonso Iannotta
Data Driven Innovation
 
OneLeap Solutions (1)
Amol Shenvi
 
Telestream Vidchecker
Benoît Godard
 
TELESTREAM Vantage - VIDELIO Cap'Ciné
Benoît Godard
 
L'architettura di classe enterprise di nuova generazione - Massimo Brignoli
Data Driven Innovation
 
Automation in Post-Production — Boris Polyak for NATEXPO 2016
Boris Polyak
 
Machine Learning Real Life Applications By Examples - Mario Cartia
Data Driven Innovation
 
SoftLayer Storage Services Overview
Michael Fork
 
Aspera on demand for AWS (S3 inc) overview
Bhavik Vyas
 
Polyglot Persistence e Big Data: tra innovazione e difficoltà su casi reali -...
Data Driven Innovation
 
The Top Skills That Can Get You Hired in 2017
LinkedIn
 
Ad

Similar to AWS 2013 LA Media Event: Scalable Media Processing (20)

PDF
Scale, baby, scale!
Julien SIMON
 
PDF
Netflix on Cloud - combined slides for Dev and Ops
Adrian Cockcroft
 
PPTX
Application Architecture Summit - Monitoring the Dynamic Cloud
New Relic
 
PPTX
re:Invent Recap-AWSMeetup
CloudHesive
 
PPTX
AWS re:Invent 2013 Scalable Media Processing in the Cloud
David Sayed
 
PPTX
AWS Meetup Fort Lauderdale Re:invent Recap
Anthony Palmer
 
PDF
[AWS에서의 미디어 및 엔터테인먼트] 클라우드에서의 브로드캐스팅 서비스
Amazon Web Services Korea
 
PPTX
Microservices? Dynamic Infrastructure? - Adventures in Keeping Your Applicati...
New Relic
 
PDF
Introducing to serverless computing and AWS lambda - Israel Clouds Meetup
Boaz Ziniman
 
PPTX
Build Web Applications using Microservices on Node.js and Serverless AWS
Mitoc Group
 
PPT
Continuous Delivery to the cloud - Innovate 2014
Sanjeev Sharma
 
PPTX
Microservices Architecture for Web Applications using AWS Lambda and more
Mitoc Group
 
PPTX
Microservices Architecture for Content Management Systems using AWS Lambda an...
Mitoc Group
 
PDF
Build an app on aws for your first 10 million users (2)
AWS Vietnam Community
 
PPTX
Microservices Architecture for Web Applications using Amazon AWS Cloud
Mitoc Group
 
PPTX
Serverless Microservices - Real life story of a Web App that uses AWS Lambda
Mitoc Group
 
PPTX
Docker Orchestration: Welcome to the Jungle! JavaOne 2015
Patrick Chanezon
 
PPTX
CMG2013 Workshop: Netflix Cloud Native, Capacity, Performance and Cost Optimi...
Adrian Cockcroft
 
PPTX
Microservices Architecture for Web Applications using Serverless Computing wi...
Mitoc Group
 
PDF
Azure Serverless Toolbox
Johan Eriksson
 
Scale, baby, scale!
Julien SIMON
 
Netflix on Cloud - combined slides for Dev and Ops
Adrian Cockcroft
 
Application Architecture Summit - Monitoring the Dynamic Cloud
New Relic
 
re:Invent Recap-AWSMeetup
CloudHesive
 
AWS re:Invent 2013 Scalable Media Processing in the Cloud
David Sayed
 
AWS Meetup Fort Lauderdale Re:invent Recap
Anthony Palmer
 
[AWS에서의 미디어 및 엔터테인먼트] 클라우드에서의 브로드캐스팅 서비스
Amazon Web Services Korea
 
Microservices? Dynamic Infrastructure? - Adventures in Keeping Your Applicati...
New Relic
 
Introducing to serverless computing and AWS lambda - Israel Clouds Meetup
Boaz Ziniman
 
Build Web Applications using Microservices on Node.js and Serverless AWS
Mitoc Group
 
Continuous Delivery to the cloud - Innovate 2014
Sanjeev Sharma
 
Microservices Architecture for Web Applications using AWS Lambda and more
Mitoc Group
 
Microservices Architecture for Content Management Systems using AWS Lambda an...
Mitoc Group
 
Build an app on aws for your first 10 million users (2)
AWS Vietnam Community
 
Microservices Architecture for Web Applications using Amazon AWS Cloud
Mitoc Group
 
Serverless Microservices - Real life story of a Web App that uses AWS Lambda
Mitoc Group
 
Docker Orchestration: Welcome to the Jungle! JavaOne 2015
Patrick Chanezon
 
CMG2013 Workshop: Netflix Cloud Native, Capacity, Performance and Cost Optimi...
Adrian Cockcroft
 
Microservices Architecture for Web Applications using Serverless Computing wi...
Mitoc Group
 
Azure Serverless Toolbox
Johan Eriksson
 
Ad

More from David Sayed (11)

PDF
Scaling Live OTT with DASH
David Sayed
 
PDF
HDR Formats and Trends
David Sayed
 
PDF
Seattle Video Tech Meetup August 2019: Optimal Multi-codec Streaming
David Sayed
 
PDF
Seattle Video Tech: The Future of SSAI on OTT Devices
David Sayed
 
PDF
On ABR Streaming and CDN Performance
David Sayed
 
PDF
NGBP ATSC 3.0 Overview
David Sayed
 
PDF
November 2018 Seattle Video Tech Meetup: SCTE-35 In-Band Event Signalling for...
David Sayed
 
PPTX
What's New with Amazon Elastic Transcoder November 2013
David Sayed
 
PDF
Internet Strategies Forum 2010 - Microsoft Media Platform
David Sayed
 
PPTX
IBC2010 Microsoft Media Platform Booth Demos
David Sayed
 
PDF
DSLR Shoot
David Sayed
 
Scaling Live OTT with DASH
David Sayed
 
HDR Formats and Trends
David Sayed
 
Seattle Video Tech Meetup August 2019: Optimal Multi-codec Streaming
David Sayed
 
Seattle Video Tech: The Future of SSAI on OTT Devices
David Sayed
 
On ABR Streaming and CDN Performance
David Sayed
 
NGBP ATSC 3.0 Overview
David Sayed
 
November 2018 Seattle Video Tech Meetup: SCTE-35 In-Band Event Signalling for...
David Sayed
 
What's New with Amazon Elastic Transcoder November 2013
David Sayed
 
Internet Strategies Forum 2010 - Microsoft Media Platform
David Sayed
 
IBC2010 Microsoft Media Platform Booth Demos
David Sayed
 
DSLR Shoot
David Sayed
 

Recently uploaded (20)

PPTX
CapCut Pro PC Crack Latest Version Free Free
josanj305
 
DOCX
Cryptography Quiz: test your knowledge of this important security concept.
Rajni Bhardwaj Grover
 
PPTX
Manual Testing for Accessibility Enhancement
Julia Undeutsch
 
PDF
Evolution: How True AI is Redefining Safety in Industry 4.0
vikaassingh4433
 
PDF
Peak of Data & AI Encore AI-Enhanced Workflows for the Real World
Safe Software
 
PDF
Transforming Utility Networks: Large-scale Data Migrations with FME
Safe Software
 
PDF
Linux schedulers for fun and profit with SchedKit
Alessio Biancalana
 
PDF
Survival Models: Proper Scoring Rule and Stochastic Optimization with Competi...
Paris Women in Machine Learning and Data Science
 
PDF
NASA A Researcher’s Guide to International Space Station : Fundamental Physics
Dr. PANKAJ DHUSSA
 
PDF
Next Generation AI: Anticipatory Intelligence, Forecasting Inflection Points ...
dleka294658677
 
PDF
“Computer Vision at Sea: Automated Fish Tracking for Sustainable Fishing,” a ...
Edge AI and Vision Alliance
 
PDF
NLJUG Speaker academy 2025 - first session
Bert Jan Schrijver
 
PDF
CIFDAQ Market Wrap for the week of 4th July 2025
CIFDAQ
 
PDF
“Squinting Vision Pipelines: Detecting and Correcting Errors in Vision Models...
Edge AI and Vision Alliance
 
PDF
Modern Decentralized Application Architectures.pdf
Kalema Edgar
 
PDF
Dev Dives: Accelerating agentic automation with Autopilot for Everyone
UiPathCommunity
 
PPTX
The Project Compass - GDG on Campus MSIT
dscmsitkol
 
PDF
Go Concurrency Real-World Patterns, Pitfalls, and Playground Battles.pdf
Emily Achieng
 
PDF
🚀 Let’s Build Our First Slack Workflow! 🔧.pdf
SanjeetMishra29
 
PDF
Software Development Company Keene Systems, Inc (1).pdf
Custom Software Development Company | Keene Systems, Inc.
 
CapCut Pro PC Crack Latest Version Free Free
josanj305
 
Cryptography Quiz: test your knowledge of this important security concept.
Rajni Bhardwaj Grover
 
Manual Testing for Accessibility Enhancement
Julia Undeutsch
 
Evolution: How True AI is Redefining Safety in Industry 4.0
vikaassingh4433
 
Peak of Data & AI Encore AI-Enhanced Workflows for the Real World
Safe Software
 
Transforming Utility Networks: Large-scale Data Migrations with FME
Safe Software
 
Linux schedulers for fun and profit with SchedKit
Alessio Biancalana
 
Survival Models: Proper Scoring Rule and Stochastic Optimization with Competi...
Paris Women in Machine Learning and Data Science
 
NASA A Researcher’s Guide to International Space Station : Fundamental Physics
Dr. PANKAJ DHUSSA
 
Next Generation AI: Anticipatory Intelligence, Forecasting Inflection Points ...
dleka294658677
 
“Computer Vision at Sea: Automated Fish Tracking for Sustainable Fishing,” a ...
Edge AI and Vision Alliance
 
NLJUG Speaker academy 2025 - first session
Bert Jan Schrijver
 
CIFDAQ Market Wrap for the week of 4th July 2025
CIFDAQ
 
“Squinting Vision Pipelines: Detecting and Correcting Errors in Vision Models...
Edge AI and Vision Alliance
 
Modern Decentralized Application Architectures.pdf
Kalema Edgar
 
Dev Dives: Accelerating agentic automation with Autopilot for Everyone
UiPathCommunity
 
The Project Compass - GDG on Campus MSIT
dscmsitkol
 
Go Concurrency Real-World Patterns, Pitfalls, and Playground Battles.pdf
Emily Achieng
 
🚀 Let’s Build Our First Slack Workflow! 🔧.pdf
SanjeetMishra29
 
Software Development Company Keene Systems, Inc (1).pdf
Custom Software Development Company | Keene Systems, Inc.
 

AWS 2013 LA Media Event: Scalable Media Processing

  • 1. Scalable Media Processing David Sayed Principal Product Manager, Amazon Elastic Transcoder Amazon Web Services © 2013 Amazon.com, Inc. and its affiliates. All rights reserved. May not be copied, modified, or distributed in whole or in part without the express consent of Amazon.com, Inc.
  • 2. Agenda • • • • • • • Media workflows Where AWS fits Scalability Cloud media processing approaches Amazon Elastic Transcoder Visual effects example Broadcast playout to web publishing example
  • 3. “I think cinema, movies and magic have always been closely associated. The very earliest people who made film were magicians.” — Francis Ford Coppola
  • 5. … And This Glass
  • 6. Media Workflows: The Real Magic
  • 7. Media Workflows: The Real Magic Insert your favorite workflows here
  • 8. Media Workflows: The Real Magic Archive Networks Insert a few more workflows here Theatrical DVD/BD Online MSOs Mobile Apps
  • 9. Media Workflows: The Real Magic Archive Networks Insert a few more workflows here Theatrical DVD/BD Online MSOs Mobile Apps
  • 10. Media Workflows: The Real Magic Featurettes Interviews 2D Movie 3D Movie Archive Materials Stills Archive Networks Insert even more workflows here Theatrical DVD/BD Online MSOs Mobile Apps
  • 11. Where AWS Fits Into Media Processing Compute & Networking Application Services Database Storage & Content Delivery Deployment & Management
  • 12. Compute & Networking Playbac k Application Services Database Storage & Content Delivery Track Auth. Protect Package QC Process Index Ingest Where AWS Fits Into Media Processing Deployment & Management
  • 13. Where AWS Fits Into Media Processing Analytics and Monetization Amazon Web Services Playbac k Track Auth. Protect Package QC Process Index Ingest Media Asset Management
  • 15. Scalability, Is It Really That Important? The business wants the new library of content they just signed on the VOD site yesterday.
  • 16. Scalability, Is It Really That Important? A major event is taking place and will be live streamed to a worldwide audience.
  • 17. Scalability, Is It Really That Important? A second screen app ties-in with a prime time event and may have millions of simultaneous users.
  • 18. Infrastructure Cost Scalability == Agility Unable to serve customers Large Capital Expenditure Predicted Demand Opportunity Cost Traditional Hardware Actual Demand Cloud Computing Time
  • 20. Cloud Media Processing Approaches: Phase 1 Phase 1 • Lift processing from on-premise and shift to the cloud
  • 21. Cloud Media Processing Approaches: Phase 2 Phase 2 Phase 1 • Lift processing from on-prem and shift to the cloud • On-prem and cloud hybrid model • Refactor workflow and cloud-optimize
  • 22. Cloud Media Processing Approaches: Phase 3 Phase 3 Phase 2 Phase 1 • Lift processing from on-prem and shift to the cloud • On-prem and cloud hybrid model • Refactor workflow and cloud-optimize • Cloud-native architecture
  • 23. A Transcoding Example “Lift and Shift” • Run existing transcoding software on AWS
  • 24. A Transcoding Example Hybrid and Refactor “Lift and Shift” • Run existing transcoding software on AWS • Run cloud versions of existing transcoding software • E.g. Digital Rapids, Elemental, Harmonic, Sorenson, Telestream
  • 25. A Transcoding Example Cloud Native Hybrid and Refactor “Lift and Shift” • Run existing transcoding software on AWS • Run cloud versions of existing transcoding software • E.g. Digital Rapids, Elemental, Harmonic, Sorenson, Telestream • Managed cloud transcoding service • E.g. Amazon Elastic Transcoder, Encoding.com, Zencoder
  • 26. Amazon Elastic Transcoder • Built in response to customer demand • Designed for file-based high volume transcoding • Simple, low-cost pricing starting at 1.5 cents per minute of output • Focused on OTT delivery formats
  • 27. Amazon Elastic Transcoder Features • • • • • • Fully managed cloud transcoding service Elastically scalable Output to MP4, WebM and HLS Integration across AWS services Available in multiple AWS regions API and SDK driven
  • 28. How To Scale With Elastic Transcoder • Architect around Elastic Transcoder • Use transcoding pipelines
  • 29. Amazon Elastic Transcoder Block Diagram Transcoding Pipeline 1 Amazon S3 Output Bucket Amazon S3 Input Bucket Transcoding Pipeline 2 Amazon S3 Output Bucket Amazon S3 Input Bucket Transcoding Pipeline 4 Amazon S3 Output Bucket 720p Transcoding Job n Kindle Fire Transcoding Job 2 Custom Transcoding Job 1
  • 30. How To Scale With Elastic Transcoder • Architect around Elastic Transcoder • Use transcoding pipelines • Use notifications
  • 31. Cloud Media Processing Summary • Three different ways to transcode in the cloud 1. Lift and Shift 2. Hybrid and/or refactor 3. Cloud-native • The right answer? It depends on your scenario
  • 32. A Visual Effects Example
  • 33. Visual Effects in Cloud “Atomic Fiction has figured out how to use cloud computing to its advantage, recently delivering 400 shots in only four months for Flight, Zemeckis’ return to live action. It’s a business model that’s being viewed as the future of the industry.” Bill Desowitz Owner ImmersedInMovies.com Contibuting Editor to Animation Scoop
  • 35. Visual Effects Summary • • • • High-resolution pre-release material Only pay for what you use Elasticity: scalability when you need it Standard visual effects workflow using ZYNC plug-ins with Maya, Mental Ray and Nuke
  • 36. A Broadcaster Example: Playout to Web Publishing
  • 37. Project Objective • Get broadcast content to VOD player ASAP • Enable editorial clipping and review • Create many different renditions for output devices • Elastically scale for high-demand situations (e.g. breaking news and sporting events) • Deliver agility for different types of processing
  • 38. Playout To Web Workflow SDI Feed Amazon Elastic Transcoder Broadcast Encoder Business Rules Program Data Feed Live Ingest Logic Amazon S3 Re-assembled Mezzanine Elemental Transcoder DRM QC Editorial Clipping MAM Amazon S3 Distribution Renditions
  • 39. Best Practices For This Project • • • • • • Componentize, decouple and be stateless Design for failure If data is in motion, use queues Run at least 3 of everything Log everything Use frameworks where possible
  • 40. Project Results • Quicker turnaround time • Only pay for what they use • Ability to support greater number of target platforms • Extremely resilient • Better control over assets
  • 41. Parting Thoughts • Using AWS for scalable media processing is a reality • Different ways to architect: – Lift and shift – Hybrid and refactor – Cloud-native • Start now

Editor's Notes

  • #2: Joined 2 years ago to focus on building a new media-related service.Couldn’t tell you what it was at last year’s eventNow it is Amazon Elastic TranscoderMedia credentialsWorked in event videoWorked on media technologies at MicrosoftTalk today about media processing in the cloud
  • #3: We’ll cover:Media workflowsWhere and how AWS fits into those workflowsWhy scalability is importantApproaches for cloud media processingNext I’ll talk about Amazon Elastic TranscoderWhat it doesWhy we built itHow it fits into a media workflowThen we’ll have a media processing example for video effectsFinally, I’ll talk about a broadcast example where the broadcaster wants to take content from their playout system and publish it to their web-based VOD service.But first I’m going to talk about magic.
  • #4: Movie aficionados often talk about the magic of film.After all movies are all about making the audience believe what’s taking place on the screen.And in fact as this well known director says, the history of film and magic is intertwined.Those of you who saw the film Hugo might recall, French magician Georges Méliès invented so many of the movie techniques that we take for granted today: stop trick (where an object disappears), multiple exposures, time-lapse photography, dissolves and hand coloring. What’s amazing is that he did this in the 1900s.
  • #5: Cameras capture the on-set magic.Beautifully lit scene from a lighting masterclass at Mole Richardson.Glass = ArriAlexa. Light gathering magical box.Super 35 sensor generates almost 20GB for every minute that’s shot.Many of you will be familiar with shooting ratio: the ratio of how much material that’s shot on camera to makes it into the final edit.Conservative number is 10:1. 10 minutes  1 minute.Today’s features are about 140 minutes in duration.With that shooting ratio that’s 26TB before you even think about multiple cameras.100TB for a movie is not unreasonable. We’ll come back to that a bit later.
  • #6: Now look at the glass at the other end: the screen.And it is a magical experience to watch a movie on a big screen.This one here is at the CGV Starium in Korea.Supposed to be largest in the world: 4000 sqft – almost a basketball court.They use a couple of Christie 2K projectors to power the screen so that’s about 3GB/minute.But I think the real magic is what needs to happen to get the beautiful big images to the beautiful big screen.
  • #7: The media workflow or the glass to glass workflow is the real magic.Behind every great magic trick is science and engineering. Sorry to disappoint you but David Copperfield doesn’t really levitate: he uses a harnessThe disappearing Statue of Liberty? Stage netting and lighting.All of which requires science and engineering.So what about our media workflows?Media workflowss are part of the engineering that make the magic of the movies possible.
  • #8: Well you probably all have at least dozen ways to get from camera to screen and then we have things like The Academy’s color workflow.So you’ve got your workflows and that gets you to one screen…
  • #9: But now you’ve got all these other outputs.
  • #10: But now you’ve got all these other outputs.And many of them have multiplesHow many online devices do you support?How many different edits, languages, and so on
  • #11: And an ever increasing number of content sources.In fact we hear that over the course of its life, a particular asset may have been converted 1,000 times.
  • #12: So where does AWS fit into media processing workflows?As you’ve heard today, we provide infrastructure services: compute, networking, database, storage and delivery and so on. We also provide application services and deployment and management services. Using these services, as your “software defined datacenter”, you can build media processing workflows.
  • #13: Typical operations in a media workflow would run on top of the AWS services. These operations could be provided by software that you’ve developed or they might be from another vendor like Aspera for ingest or Tektronix for video QC.
  • #14: On top of all that you’d have business level applications like what Ben’s team at Sony built.
  • #15: Now one thing you’ve heard us talk about is scalability. I said earlier that you can think of us as a datacenter in software. You need another compute instance, you can create one at the click of a mouse or ideally through code that creates and destroys instances as it needs to. And that’s elasticity – the ability to not only scale up but also to scale down when you need to.We say that you pay for what you use, so build your media processing in such a way to take advantage of that.
  • #16: So
  • #20: So now that you’re convinced that AWS is the way to go to build your media workflows on, how do you go about it?
  • #34: http://www.digitalartsonline.co.uk/features/motion-graphics/how-stand-out-in-vfx-animation/
  • #38: All broadcast content available through catch up servicesSome content is file-basedSome comes from broadcast playout systemContent usually needs some editorial clippingLots of output devices including set top boxes and streaming devices