-
Origin.
One stream in.
-
Global and regional masters.
The one copy is pushed into the network over UDP with error correction and carried on to the endnodes. No edge has to pull each segment again.
-
Endnodes.
The local point near the audience; each endnode serves its viewers locally.
-
Viewers.
The last mile is unchanged; every viewer still receives their own copy of the picture.
One copy on the backbone. Whatever the audience.
SmartCast is Viewcast’s delivery layer built on Smart PTMP: origin to global masters to regional masters to endnodes to viewers. One stream crosses the long-distance network however large the audience gets, and fans out only where the audience is.
The path, in order
SmartCast delivers a live stream in four steps: the origin sends one copy to global masters, global masters feed regional masters, regional masters feed endnodes, and endnodes fan out to viewers.
Schematic. Node positions illustrate the hierarchy; they are not an inventory.
What happens when everyone tunes in at once?
One stream crosses the long-distance network however large the audience gets.
Live monitoring
The 24/7 network operations centre has a live dashboard for every stream, every endnode and every viewer.
Origin, distant edge and endnode capacity
- What happens to the origin and delivery network when viewers join at once?
- The origin sends one copy to global masters, global masters feed regional masters, regional masters feed endnodes, and endnodes fan out to viewers. One stream crosses the long-distance network however large the audience gets. Add another ten thousand viewers and the long-distance part carries no extra traffic at all.
- Is a viewer joining a cold fetch from a distant edge?
- Because the endnode already carries the stream, tuning in is not a cold fetch from a distant edge: channel switching is immediate, with no buffering on start.
- How does an endnode serve its viewers?
- Each endnode serves its viewers locally.
Why does it take seconds to switch channels on a streaming TV app?
Because the endnode already carries the stream, tuning in is not a cold fetch from a distant edge: channel switching is immediate, with no buffering on start.
Notes
- The feel of linear television, which is difficult to reproduce on a network that opens a new session for every tune-in.
- Standards-based delivery to ordinary players: HLS, SRT and MPEG-TS output; no special client software is required.
Questions about channel switching
- Why is the feel of linear television difficult to reproduce?
- The feel of linear television is difficult to reproduce on a network that opens a new session for every tune-in.
- Why is channel switching immediate on SmartCast?
- Because the endnode already carries the stream, tuning in is not a cold fetch from a distant edge: channel switching is immediate, with no buffering on start.
- Do viewers need special client software?
- Standards-based delivery to ordinary players: HLS, SRT and MPEG-TS output; no special client software is required.
What it costs you, and what it does not
Customers typically save 30% or more on what they currently pay their delivery network, and the saving grows as the audience grows.
Notes
- We price against what the equivalent unicast delivery would cost you, so the bigger your peaks, the wider the gap.
- It works over the ordinary internet, with no cooperation needed from telecoms companies.
Schematic. Not a quotation; your saving depends on your peaks.
Take just the delivery layer: what changes, and what stays
You can take SmartCast on its own: it plugs in as a transparent carve-out behind your existing playout, typically live in four to six weeks.
It is the delivery leg that changes, not your stack. Your playout stays where it is; SmartCast plugs in behind it as a transparent carve-out.
What reaches the viewer is standards-based delivery to ordinary players: HLS, SRT and MPEG-TS output, and no special client software is required.
It works over the ordinary internet, with no cooperation needed from telecoms companies, and it is cloud-agnostic: AWS, Azure, GCP or bare metal.
How the switch runs on your own chain is easiest to go through against the live system. The fastest useful conversation is a walkthrough of the live production environment, with the monitoring open and your own numbers held against it.
Notes
- If it does not perform, you flip the traffic back. That is the entire risk of trying it.
- It is the delivery leg that changes, not your stack. Cloud-agnostic: the fabric runs on AWS, Azure, GCP or bare metal.
- Or run the whole chain in one portal with SmartPlay: playout from your schedule, delivery, ads stitched into each viewer’s stream and measurement.
Questions about switching the delivery layer
- What changes in our stack when we switch the delivery layer?
- It is the delivery leg that changes, not your stack. SmartCast plugs in as a transparent carve-out behind your existing playout, and it is cloud-agnostic: AWS, Azure, GCP or bare metal.
- Do viewers need special software?
- No. Delivery is standards-based, to ordinary players: HLS, SRT and MPEG-TS output; no special client software is required.
- Do telecoms companies need to cooperate?
- No. It works over the ordinary internet, with no cooperation needed from telecoms companies.
- How long does it take, and what if it does not perform?
- We plug in as a transparent carve-out behind your existing playout, typically live in four to six weeks. If it does not perform, you flip the traffic back. That is the entire risk of trying it.
Should I use multiple CDNs for live streaming, or is one integrated delivery provider enough?
One platform, not five vendors. Playout, server-side ad insertion, DRM, analytics and the delivery layer itself: one integrated stack, one contract, one console.
Compare scope first: the integrated stack, or just the delivery layer.
Cloud-agnostic: AWS, Azure, GCP or bare metal. It is the delivery leg that changes, not your stack.
Questions about choosing a delivery setup
- What is included in one integrated stack?
- Playout, server-side ad insertion, DRM, analytics and the delivery layer itself: one integrated stack, one contract, one console.
- Can we take only the delivery layer?
- You can also take just the delivery layer: we plug in as a transparent carve-out behind your existing playout.
- Does a cloud-agnostic delivery layer change our stack?
- Cloud-agnostic: AWS, Azure, GCP or bare metal. It is the delivery leg that changes, not your stack.
Counted when viewers watch in our player
When viewers watch in our own player, viewing sessions are counted per channel and per country, without tracking pixels or third-party scripts.
Notes
- We describe it as a measurement architecture and nothing more: IP addresses are personal data, and what that means for your consent obligations is a question for your counsel, not a marketing line.
- A 24/7 network operations centre watches every stream, every endnode and every viewer, with per-channel viewer series and a geographic heatmap.
Where it is proven
SmartCast has carried live production traffic since 2018: DanceTV, an independent company and a customer of eight years, runs nine channels around the clock to more than 140 countries on it, per internal analytics.
FlixSnip, a European FAST operator, runs on SmartPlay under a signed contract.
Chart notes
2024 – present · Comcast / Xfinity. Comcast / Xfinity: signed vendor contract; active integration engagement, 2024 to present.
- Years in productionsince 2018
- 8
- Channels, 24/7
- 9
- Countriesper internal analytics
- 140+
- Consecutive production days
- 2,900+
- Uptimeper internal analytics
- 99.97%