Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Pushing WebRTC Up That Broadcast Hill
Search
danjenkins
June 01, 2023
Technology
0
62
Pushing WebRTC Up That Broadcast Hill
danjenkins
June 01, 2023
Tweet
Share
More Decks by danjenkins
See All by danjenkins
Building ICEPerf.com
danjenkins
0
74
Using GStreamer to build real-time applications with Golang
danjenkins
0
1.1k
VON: Builder - Controlling Physical Devices with WebRTC
danjenkins
0
78
When Worlds Collide: The Blurring Lines of Real-Time and Traditional Media
danjenkins
0
130
Passkeys, FIDO2, WebAuthn… What does it all mean?
danjenkins
0
150
Spiderman's Uncle Ben and his advice for casting your Web(RTC)
danjenkins
0
110
Don't talk SRT about my WHIP
danjenkins
0
780
WebRTC 101 - How to get started building your first WebRTC application
danjenkins
1
200
WebRTC Reborn London Node User Group
danjenkins
0
130
Other Decks in Technology
See All in Technology
PO初心者が考えた ”POらしさ”
nb_rady
0
210
What’s new in Android development tools
yanzm
0
310
スタートアップに選択肢を 〜生成AIを活用したセカンダリー事業への挑戦〜
nstock
0
170
How Do I Contact HP Printer Support? [Full 2025 Guide for U.S. Businesses]
harrry1211
0
120
整頓のジレンマとの戦い〜Tidy First?で振り返る事業とキャリアの歩み〜/Fighting the tidiness dilemma〜Business and Career Milestones Reflected on in Tidy First?〜
bitkey
2
16k
Lazy application authentication with Tailscale
bluehatbrit
0
210
Should Our Project Join the CNCF? (Japanese Recap)
whywaita
PRO
0
340
Zephyr RTOSを使った開発コンペに参加した件
iotengineer22
1
220
20250707-AI活用の個人差を埋めるチームづくり
shnjtk
4
3.8k
ビギナーであり続ける/beginning
ikuodanaka
3
750
Sansanのデータプロダクトマネジメントのアプローチ
sansantech
PRO
0
150
MobileActOsaka_250704.pdf
akaitadaaki
0
120
Featured
See All Featured
The Web Performance Landscape in 2024 [PerfNow 2024]
tammyeverts
8
690
Become a Pro
speakerdeck
PRO
29
5.4k
GraphQLの誤解/rethinking-graphql
sonatard
71
11k
The MySQL Ecosystem @ GitHub 2015
samlambert
251
13k
Code Reviewing Like a Champion
maltzj
524
40k
Faster Mobile Websites
deanohume
307
31k
The Language of Interfaces
destraynor
158
25k
Building a Scalable Design System with Sketch
lauravandoore
462
33k
How To Stay Up To Date on Web Technology
chriscoyier
790
250k
Embracing the Ebb and Flow
colly
86
4.7k
Code Review Best Practice
trishagee
69
18k
Chrome DevTools: State of the Union 2024 - Debugging React & Beyond
addyosmani
7
740
Transcript
Streaming Tech Sweden 2023 Pushing WebRTC Up That Broadcast Hill
Dan Jenkins
Dan Jenkins WebRTC & Real Time Media Geek @dan_jenkins @
[email protected]
Founder - Everycast Labs / Nimble Ape Organiser CommCon (19-23 June - UK)
Broadcast Bridge - bring remote collaborators into your professional work
fl ow @everycastlabs broadcastbridge.app everycastlabs.uk
Pushing WebRTC Up That Broadcast Hill
WebRTC Is Now 12 Years Old.
And It Has Barely Been A "Thing" Inside Broadcast Media.
So First... Let's Define WebRTC
WebRTC Allows Two Peers behind NATs To Connect to One
Another
It Uses an ICE Process To Make Those Connections.
WebRTC Allows for Media and Data To Be Sent Over
a Peer Connection
WebRTC Is Encrypted by Default.
A Peer Doesn't Need To Be a Client
An SFU or MCU Is Also a Peer.
WebRTC Media Is Sent as RTP
Which Allows for RTP Extensions To Be Used.
RTX
Absolute Capture Time
Audio Levels
Absolute Send Time
Video Orientation
Playout Delay
Video Content Type
That's Just a Few.
It's a Fantastic Toolset To Have at Your Fingertips.
Some May Say It's Rather Prosumer https://twitter.com/OpenBroadcastSy/status/1662075098172358657
But In Order To Understand Some Of The Whys...
We Need To Look at WebRTCs Past
Google Acquired Global IP Solutions...
And Then Open Sourced WebRTC (YAY!)
The IETF And The W3C Were Then Involved With Standardisation
But at its Core...
WebRTC Was Designed for Conference Calls
"Real Time Media on the Web"
And the IETF Process Was Led by VoIP People
Real Time Media = (Web) Real Time Media
And This Led to Decisions.
Like Using SDP To Handle Negotiation of a Session.
But The Biggest* Decision Was Having No Standard Signalling Method.
(*At Least in my mind)
People Could Use What They Wanted.
JSON Over Websocket
SIP Over Websocket
XMPP
GraphQL
Some Proprietary Binary Format Over Websocket
*N
You Could Do It Your Way
Whatever Made Sense for Your Application
Include SDP? Sure!
Get Rid Of SDP And Fill It In Later?
That Works Too!
But No Standard Signalling Isn't Ecosystem Friendly
SRT, RTMP, NDI, HLS & DASH (And many others)
They Just Work (Mostly)
The Lack of a Common Signalling Layer Was Both Great
and Terrible
Now...
The Lack Of Common Signalling Wasn't The Only Issue.
Lack Of Controllable Latency
Lack Of Encode/ Decode Control In A Browser
Lack Of Control Generally.
Pray Be To The Browser Gods.
But.
WebRTC Has Its Own Superpowers
None
WebRTCs Superpower Is That It Works on Billions of Devices
Without Installing A Thing
Use The Right Tool For The Right Job.
WebRTC In A Browser Is The Right Tool.
For The Right Job.
It Might Just Be A Different Job Than You Expected.
Now...
Let's Take A Look at WebRTCs Present
Works In All Browsers Yes, even on Mobile (Safari, Chrome,
Firefox - Edge is a clone)
Supports New Codecs (AV1, H265)
OPUS RED.
More Than Just libWebRTC (Pion, WebRTC.rs, aiortc, |pipe| to name
a few)
SVC (One Video encoded once resulting in one stream with
multiple qualities)
...But Different Support in Every Browser
Simulcast (Multiple encoded videos sent independently as multiple streams) (Now
includes VP9 and AV1 in Chrome)
...But Different Support in Every Browser
X & Y Features
...But Different Support in Every Browser
Scaling Isn't Simple
Scaling To 100,000s Of Connections
It's Tricky
And Expensive
So WebRTCs Present Is Just Tricky.
However.
Something Is Becoming Simpler...
Standardised & Simple Signalling
WHIP (WebRTC HTTP Ingest Protocol)
WHEP (WebRTC HTTP Egress Protocol) https://eyevinntechnology.medium.com/whip-whpp-for-webrtc-based-broadcast-streaming-2cf469e95299
They Go Hand In Hand.
Let's Take A Look At WHIP
At The Heart Of Both Is A HTTP POST Request/
Response
POST /whip/endpoint/test HTTP/1.1 Host: localhost:7080 Content-Type: application/sdp Authorization: Bearer verysecret
Accept-Encoding: gzip, de fl ate Connection: Keep-Alive Content-Length: 1162 v=0 [..sdp] (WHIP) POST with an OFFER
HTTP/1.1 201 Created Location: /whip/resource/test Content-Type: application/sdp Connection: keep-alive Transfer-Encoding:
chunked Content-Length: 569 v=0 [..sdp] (WHIP) POST RESPONSE WITH ANSWER
WHEP Is the SAME
Super Simple.
And Finally Means Ecosystem Friendliness
Both WHIP and WHEP Are Getting Industry Support
Dolby.io
Cloudflare
Osprey Talon Encoders
OBS
Flowcaster
LiveU
And Lastly...
Broadcast Bridge
What's Coming in WebRTCs Future?
Or...
What Do I Hope We Will See Soon?
WebCodec Support
DRED (Deep Audio Redundancy)
Bringing All Browsers Up To The Same Spec
LOL.
This Is a Web Problem Generally.
So What's In It For You?
Lack of Control?
Issues Scaling?
You're Not Selling Me on This Dan...
But Seriously.
Use WebRTC Right...
And It Will Work Well for You.
WebRTC Isn't Right for Broadcasting to 100,000 People
WebRTC Isn't Right for Broadcasting to 100,000 People (Usually)
WebRTC Is Not Right for Transporting Media Inside a Studio.
WebRTC Is Right for Real Time Collaboration
Where You Need the Lowest Friction
WebRTC Is Right for Weird Network Scenarios
It Is a Tool in the Toolbelt.
And Even Though Web Browsers Frustrate Me
A LOT.
It Is Only Because of the Web Browser That WebRTC
Has Its Superpower
It Works Everywhere
Not Everything Is a Nail Needing a Hammer.
I've Seen Stranger Things Than WebRTC Being Used in Broadcast
That's Me!
Cheers!
Thanks! We're hiring in the UK! https://jobs.everycastlabs.uk @dan_jenkins @
[email protected]
@everycastlabs