DVB subtitles for the chains that run on transport streams
Teletext, DVB-SUB and DVB-TTML, generated live and muxed into your SRT stream by the Kobe Muxer, SyncWords' own broadcast-grade MPEG-TS stack. One family, three formats, chosen by what your decoders support.
- Teletext (DVB-TXT)
- DVB-SUB, any script
- DVB-TTML
- Kobe Muxer
- Per-language PID mapping
The family
Three formats, one decision: what does your chain decode?
The interesting choice in DVB is between the three members of the family. All three ride the SRT / MPEG-TS path, each language on its own PID, labeled in the service descriptor so set-top boxes show the right selector.
Teletext (DVB-TXT)
The workhorse. Decoded by practically everything in a European broadcast chain. Per-language page mapping auto-assigned from page 888, level-1 colours, double height, row and alignment control. Its character sets cover part of the language set, which is its one real limit.
DVB-SUB
Bitmap subtitles, rasterised before they leave us. The decoder never needs the font, so any script at all renders correctly: Arabic, CJK, Cyrillic, Hebrew. The dependable answer when the language set outruns Teletext.
DVB-TTML
The W3C-based format the standard is heading toward, with modern styling, precise positioning and accurate timing. Decoder support is still limited, so confirm your chain decodes it before building on it.
Kobe Muxer
A transport-stream stack built in-house
Kobe Muxer is SyncWords' proprietary MPEG transport stream multiplexer. It carries the whole broadcast pipeline: embedded 608, Teletext, DVB-SUB, DVB-TTML and dubbed audio tracks, inserted without re-encoding the video.
Any framerate, UHD-ready
Built for high-bitrate streams, including live 4K and UHD channels, with the timing accuracy linear TV requires.
Per-language PID mapping
Each subtitle language travels on its own PID in the same stream and is labeled in the service descriptor, so downstream systems find it without manual mapping.
Fast insertion
Sub-2s caption delay in SRT, where typical captioning solutions display live captions within 4 to 8 seconds. Add a few seconds of buffer and captions land with no perceptible delay.
Presentation control
Format-by-format control
Placement and styling are configurable per service and stay attached to it between runs.
- Teletext
- Row 1-23, one to four lines, start or center alignment, double height, the seven Teletext level-1 colours, and per-language page mapping.
- DVB-SUB
- Start or center alignment, font size as a percentage of video height, bottom margin, and full RGBA control of foreground and background.
- DVB-TTML
- Maximum line length from 12 to 120 characters and a bottom margin from 0 to 90% of the viewport.
- Dubbed audio
- Dub tracks join the same stream: codec per track, explicit or auto-assigned PIDs, and controllable position in the PMT relative to the original audio.
Frequently asked questions
Which DVB format should we choose?
Start from your decoders. Teletext is decoded almost everywhere in European chains and is the default. If your language set includes scripts Teletext cannot carry, DVB-SUB renders any script because it ships as a bitmap. DVB-TTML is the most capable on paper and where the standard is heading, but decoder support is still limited, so treat it as an option to verify rather than assume.
What is the Kobe Muxer?
SyncWords' proprietary broadcast-grade MPEG transport stream multiplexer. It writes embedded EIA-608, Teletext, DVB-SUB, DVB-TTML and dubbed audio tracks into SRT streams at almost any framerate, keeping subtitles synchronized with the video, all in the cloud with no hardware and no transcoding.
Can several subtitle languages ride in one SRT stream?
Yes. Each language is carried on its own PID inside the same stream and labeled in the service description table, so you do not open a new output per market. Live AI dubs can travel on the same stream as additional audio PIDs.
What delay can we expect on captions?
Sub-2s caption delay in SRT. If you prefer captions exactly in sync with the audio, a few seconds of configurable buffer absorbs the processing time; the video path itself adds a half-second latency in SRT workflows. See latency, processing time and delay.
How does SyncWords handle redundancy and scale?
The platform runs on Amazon EKS with primary and backup resources in any AWS region, dual-pipeline A/B failover with SCTE-35 pass-through, and no per-channel ceiling or licence. Create a service per stream, then run it event-based or 24/7.
Does this help with our accessibility obligations?
It gives you the output side of the job: DVB subtitles and embedded captions in the formats and languages your regulators and distributors ask for, on live programming. Compliance itself rests with the broadcaster; our output is built to support it.
Check it against your decoders
The only test that matters is your own chain. Our engineers will set up a proof of concept on your SRT feed so you can verify every format downstream.