Overview
Videolinq is a cloud-based platform that adds real-time captions and subtitles to live video streams. It supports AI-generated and human-edited captions and injects them into live broadcasts distributed to destinations such as YouTube, Facebook, LinkedIn, and HTML5 and proprietary compatible video players. Caption insertion is powered by proprietary technology that requires compatibility with specific video stream structures.
How it Works
Live Video Source: SDI, HDMI, or virtual camera source.
Encoder: Hardware or software encoder sends a live RTMP or SRT stream to Videolinq.
Caption Source:
AI-generated captions (real-time speech recognition)
Human editor (to correct or enhance captions in real-time)
External caption feeds (e.g., StreamText, 1CapApp)
Videolinq Processing:
Injects CEA-608/708 captions into RTMP or HLS output via SEI user data.
Optionally adds open (burned-in) captions.
Data feed output displayed on HTML pages or as a sidecar format (WebVTT manifest).
Distribution:
Output stream sent to video platforms or RTMP/SRT endpoints with captions included.
Why Encoders Do Not Work
Not all RTMP or SRT encoders are compatible with Videolinq's caption insertion process. Successful caption injection depends on the encoder's ability to:
Support a standard container structure (RTMP/SRT) with full metadata.
Preserve or expose SEI (Supplemental Enhancement Information) layers for CEA-608/708 insertion.
Maintain proper keyframe alignment and access points needed for post-ingest modification.
Most web-based meeting and streaming platforms, such as StreamYard and Restream, use simplified RTMP workflows that typically do not expose SEI metadata. However, in some tests, Restream has shown variable behavior depending on the source encoder, and StreamYard may allow SEI-based injection if the upstream feed supports it. These platforms are considered conditionally supported but are not recommended for critical captioning workflows. Zoom RTMP output has been verified to support SEI-based caption insertion by Videolinq, assuming proper configuration.
Supported Encoders
Enterprise-Grade Encoders (Recommended for Broadcast Workflows)
Brand | Model(s) | Caption Support | Notes |
Haivision | Makito X4, Makito X, MB6, Rack Series | SEI + VANC | Broadcast-grade, multi-profile support, RTMP/SRT |
MediaKind (Ericsson) | CE1, RX8200, AVP4000 | SEI | DVB/SCTE/CEA compliant, ideal for contribution workflows |
Evertz | ev670-XE, Mediator X, 5701MSC | SEI | Modular platform with SDI and IP workflows |
Imagine Communications | SelenioFlex Live, Selenio Network Processor | SEI | SEI metadata insertion and multi-platform delivery |
Harmonic | Electra X, ProStream 9100, ViBE EM4000 | SEI + DVB | Supports all broadcast caption standards |
ATEME | TITAN Live, KYRION CM5000 | SEI | Advanced OTT/broadcast encoder |
Cisco | D9036 Encoder, DCM Series | SEI | Legacy encoder with SEI support in RTMP |
Elemental (AWS) | Live Encoder, Statmux Encoder | SEI | Cloud/hardware hybrid, AWS-integrated |
Sencore | OmniHub X, MRD 7000, DMG 7000 | SEI + VANC | SDI ingest with compliant SEI metadata injection |
Telestream | Lightspeed Live, Vantage with Live Encode Wirecast | SEI | End-to-end support for live and file-based captions |
Nevion | Virtuoso MI | SEI | High-end IP contribution encoder |
Professional / Semi-Pro Encoders (SEI-Compatible for Videolinq)
Brand | Model(s) | Caption Support | Notes |
Matrox | Monarch HDX, Monarch EDGE | SEI + VANC | Dual input, RTMP/SRT outputs with proper metadata |
Teradek | Cube 655/755, Prism Flex, Prism Mobile | SEI + VANC | Encoder with SEI metadata for RTMP/SRT |
AJA | HELO, HELO Plus | SEI + VANC passthrough | Compatible via SDI; requires external caption source |
Magewell | Ultra Encode SDI, Ultra Encode AIO | SEI (firmware dependent) | Good for compact IP production workflows |
Osprey | Talon G1, Talon UHD, Talon 4K-SC | SEI via config | Supports RTMP/SRT, caption metadata if enabled |
NewTek | TriCaster TC1, TriCaster 2 Elite | SEI injection supported | Verified RTMP output allows caption insertion |
OBS | OBS Studio | SEI injection supported | Verified RTMP output allows caption insertion |
vMix | All vMix products | SEI injection supported | Verified RTMP output allows caption insertion |
Zoom | Zoom paid plans with RTMP output | SEI injection supported | Verified RTMP output allows caption insertion |
Not Supported Encoders
Brand | Model(s) | Limitation | Notes |
Blackmagic | Web Presenter HD, Web Presenter 4K | No SEI/VANC support | Encoded RTMP is non-modifiable |
YoloLiv | YoloBox, YoloBox Pro, Instream | SEI insertion | Designed for ease-of-use, lacks metadata structure |
Epiphan | Pearl Mini, Pearl-2 | Partial (internal only) | No SEI in RTMP output; VANC only to internal record |
LiveU | Solo, LU300, LU800 | No SEI injection | Focused on transmission, not caption workflows |
Mevo | Mevo Start, Mevo Plus | No metadata support | Simplified streaming, not broadcast-grade |
OBSBOT | Tiny 4K, Tail Air | No captioning capabilities | No SEI, RTMP-only, AI camera streamers |
StreamYard | Web-based encoder | Conditional | SEI not guaranteed; not recommended for critical caption workflows |
Restream | Web-based encoder | Conditional | May work depending on upstream encoder; verify on case-by-case basis |
Conclusion
To enable real-time caption injection via Videolinq, your encoder must support RTMP or SRT containers that preserve SEI metadata and allow CEA-608/708 injection. Use only verified encoders that meet these standards for compliant, editable live streams. Avoid encoders that seal or strip metadata layers, as they block downstream caption insertion.
We recommend using OBS, vMix, Zoom RTMP output, or any encoder listed in the supported sections above for full compatibility.