Real-time multilingual live dubbing
DubStream
Carries a speaker's voice into 12 languages at 2.4-second median latency. Runs on our hardware at the venue, so a dropped uplink does not drop the translation.
Specification
DubStream, standard configuration
| Target languages | 12 concurrentBeyond 8, add a second GPU node |
|---|---|
| Median latency | 2.4 sSpeaker lips to delegate ear |
| Delivery | Delegate mobile, or venue receivers |
| Deployment | On-premise at the venueSurvives an uplink drop |
| Source audio | XLR from the desk, or SRT |
Questions we get
Procurement questions
01Where does the data sit?
In India by default, on infrastructure we control. Event-scoped data is purged on the retention schedule agreed in the contract, and we will put that schedule in writing before you sign.
02What does it cost?
It depends on scale and whether we supply hardware, so a figure quoted before we know those would be fiction. Tell us the event size and we return a costed range, usually inside three working days.
03What happens if it fails mid-event?
We build the failure path first. This runs on our hardware at the venue with a named engineer present on show days, so recovery does not depend on a support ticket or an internet connection.
See it on a real event
Demos on a laptop prove very little. We would rather show you this running at a live show.
Commercials
How it is bought
Priced per event on delegate count, show days and whether we supply hardware; or annually under a retainer for four or more events a year. We quote a range before the venue survey and a fixed figure after it.
Security & data residency
Data is held in India on infrastructure we control. Event-scoped personal data is purged on the retention schedule set in the contract. Sub-processors are listed on /security, and we will sign a DPDP-aligned data processing agreement.
Integrations
- ZoomConferencing
- Microsoft TeamsConferencing
- YouTube LiveStreaming
- SRT / RTMPTransport
At a glance
| Target languages | 12 concurrentBeyond 8, add a second GPU node |
|---|---|
| Median latency | 2.4 sSpeaker lips to delegate ear |
| Delivery | Delegate mobile, or venue receivers |
| Deployment | On-premise at the venueSurvives an uplink drop |
| Source audio | XLR from the desk, or SRT |