EchoStream v0.1.1

Live AI captioning for vMix and OBS workflows. Pro adds NDI, translation, and text-to-speech.

1 Install

EchoStream is currently available to licensed customers. Purchase a Standard or Pro license below to receive your download link.

Already purchased? Check your email for the download link, or retrieve your license key.

2 Choose Your Audio Source

EchoStream captures audio from your production environment. On the Audio page, select your source:

Microphone — any local audio input device (WASAPI)
OMT (Open Media Transport) — network audio from vMix, OBS, or hardware encoders. Sources are auto-discovered on your LAN
NDI — industry-standard network audio from vMix, TriCaster, Wirecast, and other NDI-enabled tools. Auto-discovered via mDNS PRO

Tip: OMT and NDI sources are discovered automatically via DNS-SD. Make sure your production machines are on the same network subnet.

3 Select a Captioning Engine

Choose the AI speech recognition engine that best fits your hardware:

Engine	Best For	Languages
EchoCaption Ultra Parakeet TDT v3	Best accuracy, GPU recommended	25 (auto-detect)
EchoCaption Plus Faster Whisper	GPU or high-end CPU	99+
EchoCaption Lite Moonshine	CPU-friendly, lightweight	Major languages
EchoCaption Core WhisperCpp	CPU or GPU, versatile PRO	99+

4 Connect to vMix or OBS

On the Outputs page, enable your target application:

vMix — direct TCP API connection. Enter your vMix host IP and the title input name for caption overlays
OBS Studio — connects via WebSocket v5. Supports text source overlays and CEA-608 embedded captions
Browser Overlay — a built-in HTML overlay server you can add to any application as a browser source

vMix Caption Overlay Setup

EchoStream sends captions to vMix via the TCP API. You need a Title input in vMix with specific named text fields.

Option A: Use the included template (recommended)

Download the ready-made vMix XAML caption template. It's 1920×1080 with a semi-transparent bar at the bottom, pre-configured with the correct field names.

Download EchoStream_Captions.xaml

In vMix: Add Input → Title / XAML → browse to the downloaded file → Open.

Note: The template is also bundled with the installer at C:\Users\<you>\AppData\Local\EchoStream\templates\

Option B: Create your own title

If you prefer a custom design, create a Title input in vMix and add text fields with these exact names:

Field Name	Purpose
`CaptionLine1`	Previous completed caption line
`CaptionLine2`	Current line being spoken (rolling text)
`SpeakerName`	Speaker label (optional)

To name a field: open the Title Editor, select the text element, and set its Name property in the properties panel (not the visible text content).

Configure the connection

In EchoStream on the Outputs page:

Enable vMix
Set Host to your vMix machine IP (or localhost if same PC)
Set Title Input to the name of your Title input in vMix (e.g. EchoStream_Captions.xaml)
Set Overlay Channel (1–4) to assign the title to a vMix overlay channel
Enable Auto Show Overlay to automatically show/hide the overlay when captions start/stop

Tip: Use the Test Connection button on the Outputs page to verify EchoStream can reach vMix before starting captioning.

5 Start Captioning

Press Start on the Dashboard. Captions will appear in real time in your selected output. Typical end-to-end latency is under 400ms with a GPU.

6 Translation & Voice (Pro)

With a Pro license, you can also:

Translate captions in real time into 7+ target languages on the Translation page
Generate AI voice from captions or translated text using the Voice page, with multiple TTS providers

Standard vs Pro

EchoStream comes in two editions. Both include a 30-day free trial of all features.

Feature	Standard	Pro
Audio Sources
Microphone input	✓	✓
OMT network audio	✓	✓
NDI network audio	—	✓
Captioning
Live AI speech-to-text	✓	✓
3 ASR engines (4 with Pro)	✓	✓
99+ language recognition	✓	✓
SRT subtitle file export	✓	✓
Outputs
vMix caption overlay	✓	✓
OBS text source overlay	✓	✓
CEA-608 embedded captioning	✓	✓
Browser overlay server	✓	✓
Display customisation	✓	✓
AI Translation & Voice
Real-time caption translation	—	✓
AI text-to-speech voice generation	—	✓
Multiple TTS providers	—	✓
Integration
REST API for remote control	—	✓

Get EchoStream

Start with a 30-day free trial. Choose monthly or lifetime billing.

Standard

Captioning, vMix, OBS, overlay & browser source.

£60/mo

Cancel anytime

£1,050

One-time payment

Buy Lifetime

PRO

Pro

Everything in Standard plus NDI, translation, TTS & REST API.

£100/mo

Cancel anytime

£2,150

One-time payment

Buy Lifetime

Already purchased? Retrieve your license key.

System Requirements

Component	Minimum	Recommended
OS	Windows 10/11 (x64)	Windows 10/11 (x64)
CPU	4 cores	8+ cores (i7/Ryzen 7)
RAM	4 GB	16 GB+
GPU	None (CPU mode)	NVIDIA RTX 3070+
Storage	5 GB (with models)	10 GB+
Network	LAN (for OMT/NDI)	Gigabit Ethernet

GPU note: An NVIDIA GPU with CUDA support significantly improves captioning latency. Without a GPU, EchoStream runs in CPU mode using lighter models (EchoCaption Lite or Core).

Troubleshooting

No OMT/NDI sources found? Make sure your source devices are on the same network subnet. Check that UDP port 5353 (mDNS) is not blocked by your firewall.
High latency or slow captions? Switch to a lighter ASR engine (EchoCaption Lite or Core) or enable GPU acceleration in Settings if you have an NVIDIA GPU.
Captions not appearing in vMix? Verify the host IP and title input name on the Outputs page. Make sure vMix's TCP API is enabled (default port 8099).
Captions not appearing in OBS? Check that OBS WebSocket Server is enabled in OBS (Tools → WebSocket Server Settings). Default port is 4455.
NDI source greyed out? NDI input requires a Pro license. Go to Settings to activate or start a trial.
Translation or Voice pages locked? These are Pro features. Go to Settings to activate your Pro license.

Need Help?

Website: sebiulabs.co.uk