Rack
Mcaster
Channels leave the studio ready for every distributor
Every distributor takes the signal its own way, so a broadcaster ends up building an in-house centre of technical expertise instead of producing content. Mcaster takes that job over: ingest from the studio over SDI, ASI, NDI and ST2110, instrument-grade quality control, transcoding, multiplexing, and MPTS out to every distribution point. It ships as a project — survey, design and estimate, commissioning, and support.
A content aggregator retransmitting about 800 channels retired fifty hardware encoders for one platform.
Figures
What changed where it has already been done
- −60%complaints about channel delivery after a fleet of hardware encoders was retired
- 50 → 0hardware encoders retired: about 800 channels now run on one platform
- 3 monthsto migrate about 800 channels, with retransmission never stopping
- −40%operating costs at a nation-wide DVB-T2 broadcaster running 22 HD channels
The first three figures come from one installation — a content aggregator that replaced 50 hardware encoders with Mcaster. The −40% is a different project: a nation-wide DVB-T2 broadcast system on 50+ servers, holding 99.99% uptime.
Case
800 channels off 50 hardware encoders
A content aggregator retransmits about 800 channels. It wanted to stop depending on hardware that was supposed to never break, and to find out what its customers were complaining about before the complaints arrived.
Before
Fifty legacy hardware encoders, each of which had to be nursed back by hand when it failed. Reliability rested on the hardware not breaking, and the operator saw the input but not what actually reached the customer — so a delivery problem surfaced as a phone call, not as a graph.
After
- 50 hardware encoders replaced by one platform
- 60% fewer customer complaints
- Automatic failover around failing equipment, with no manual recovery
- Full visibility of input and output, pushes and sessions included
The same head-end at another scale: a nation-wide DVB-T2 system on 50+ servers, 22 HD channels, 99.99% uptime and 40% lower operating costs. Read the DVB-T2 case
Who asks
Two different conversations
A broadcaster decides whether preparing the signal is its own business at all. An integrator decides what to put in the specification. The answers are different, so they are written down separately.
You are the broadcaster
The real choice is not between vendors. It is whether you keep funding an in-house centre of expertise for every distributor's way of taking a signal, or buy that competence as a delivered system.
- What does it cost to keep people who know every distributor's format?
- What happens to the air when a piece of hardware fails at 3am?
- Who carries responsibility for the result of the rollout?
- How much time goes into establishing whose side broke?
You are the integrator
You are assembling the object and answering to the broadcaster for it. What matters is a specification that survives the tender, and a clear line between what we commission and what you do.
- What goes into the specification: hardware, capture cards, modules
- Where the commissioning line runs between Flussonic and the integrator
- Which parts of the maintenance contract you can carry yourself
Not this
If your job is a different one
Said up front, so a call does not end with «this is not for us». In these cases you need another product of ours, not Mcaster.
Your channels go to subscribers over IP. If there are viewers at the end of the line and you need packages, EPG and billing rather than an MPTS for a distributor, take
CatenaYou broadcast inside your own company. Live streams to every employee, training and emergency notification in a closed perimeter — take
AgoraYou want an engine, not a delivered system. Mcaster is a project with commissioning and acceptance criteria. If you are building your own pipeline on a streaming core, take
Sapsan
Signal
From the studio to every distribution point
Four stages in one rack, instead of a chain of separate boxes each with its own vendor, its own spare part and its own way of failing.
Ingest
SDI, ASI, NDI and ST2110 come in through dedicated capture cards; IP sources arrive over SRT, RIST and MPEG-TS.
One rack takes both the broadcast contour and the IP one, so you stop running two parallel systems with two sets of people.
Capture cards for SDI and ASI go into the project, and which ones is decided during the survey.
Instrument-grade control
An MPEG-TS analyzer watches the stream continuously: PSI/SI tables, bitrate, continuity counters, losses.
A problem shows up on the input rather than in a distributor's phone call, and because the history is kept, the argument about whose side broke ends with a graph.
The history goes to the Prometheus that ships with the system; how far back it reaches is set during commissioning.
Preparation
Constant-bitrate transcoding, recording to disk, and packaging into HLS and DASH.
The same channel goes out as MPTS to a cable operator and as HLS to an OTT platform, so a new distributor does not mean new hardware.
Constant bitrate keeps the bandwidth predictable, which is what lets channel density per server be calculated in advance rather than discovered.
Delivery
The MPTS multiplexer assembles services into a transport stream and generates PSI/SI for DVB, ATSC and ISDB.
Connecting a new distributor is a configuration change, not the purchase of another multiplexer.
The set of output standards is fixed in the project, and the module licensing follows from it.
Failure
Reliability comes from redundancy, not from heroic hardware.
Something fails
A capture card, a server, or an incoming source goes down. On a fleet of hardware encoders this is the moment an engineer is woken up.
Mcaster routes around it
Switching to the reserve happens automatically, with no manual recovery. The aggregator that moved off 50 encoders stopped restoring equipment by hand altogether.
The channel stays up
In the nation-wide DVB-T2 system this holds 99.99% availability across 50+ servers, with terrestrial signal redundancy behind it.
Both sides
Delivery both sides can see
The monitoring runs to the end of the chain — not only the input, but the pushes and the play sessions. The same window is open to your customer: the operator who takes your channels and delivers them to households watches how delivery to them is doing, without opening a ticket.
Before
The customer calls. You look at the input and it is fine. Then come the days of correspondence in which you establish whose side broke. Usually nobody is found at fault, and the bad taste stays.
After
The customer opens the page and sees their own delivery. Part of the enquiries never start at all. The ones that do start from a fact rather than from a version of events.
Support stops growing with the number of operators you are connected to, and the SLA becomes something to observe rather than something to promise: you arrive with a window onto the delivery, not with a number in a contract.
What is visible is exactly what goes through Mcaster: if part of the delivery runs around it, the picture will be incomplete.
At the aggregator running about 800 channels this is what the 60% drop in complaints was made of. Read the case study
Day 2
What it is like to live with
A head-end is bought once and operated for years. These are the properties that show up in the second year, not in the demo.
An immutable image
The system ships as one ready image together with Flussonic OS. Configuration and statistics live on a separate partition — back it up on its own, or wipe it and return to factory settings.
It runs in a closed perimeter
No internet access is required for operation, which is what a broadcast contour usually has to be.
Upgrading stays your decision
Builds are dated — 25.03.1 is the first build of March 2025 — and nothing updates itself. If behaviour has to change, we announce it at least six months ahead.
Support from the vendor
A request from your account gets a number; a confirmed defect becomes a development task with a regression test, and comes back to you in a release, with notice.
Growth is more channels, not another system
New channels, new distributors and a move to 4K are handled by extending the same head-end — the nation-wide system reached 50+ servers this way.
Stages
How a head-end actually gets built
Mcaster is not downloaded and tried out. It is designed for your channel list and your distributors, then commissioned against criteria agreed in advance.
Survey
We collect the business requirements: the channel list, which distributors take what and in which format, the sites, and the redundancy you need.
Design and estimate
Equipment, the set of modules out of the twenty-plus available, and the licensing programme. The output is a specification you can take into procurement.
Commissioning
Installation and configuration up to production operation, accepted against the criteria written down before the work started.
Operation
Scheduled maintenance, software updates and configuration changes, plus support — with a named technical manager where that is needed.
There is no trial, on purpose
Design, installation and configuration are done by engineers trained for it, and commissioning runs against prepared acceptance criteria. That is how the result is guaranteed rather than hoped for.
An integrator can carry part of it
Commissioning, scheduled maintenance and first-line support can be run by a trained integrator. Where the line falls is fixed in the project, not argued about later.
Supply
What you actually buy
A delivered head-end, in one of three forms — plus the work around it, which is part of the contract rather than an extra.
Mcaster Appliance
A ready hardware-and-software system with proven characteristics: Flussonic OS, capture cards, and the module set already assembled and verified.
Software
Server firmware to install on hardware you buy yourself, when the rack is already decided.
Project solution
The system designed and adapted to your task, including commissioning and ongoing support.
Module licensing
Over twenty modules — ingest, analyzer, transcoder, DVR, multiplexer, packager, monitoring. Which ones you need is decided by the project, so you pay for the contour you actually run.
Maintenance and SLA
Scheduled maintenance and technical support, with a named technical manager available where the operation calls for one.
Project
Get to a project
Tell us your channel list, your distributors and how each of them takes the signal. Our engineers will lay out the head-end architecture, the equipment specification and the estimate — together with the acceptance criteria that commissioning is signed off against.
Not sure whether the job is a head-end or subscriber IPTV? Compare with Catena