Glossary
VoIP (Voice over IP)
Phone calls as data. VoIP digitises voice and sends it as packets over the internet or a private IP network, replacing the dedicated circuit that the traditional phone network reserves for every call.
General definition
VoIP (Voice over IP) converts speech into digital audio, compresses it with a codec, and sends it across an IP network in small packets that are reassembled and played back at the other end. The public switched telephone network (PSTN) reserves a fixed circuit for the length of every call; VoIP shares ordinary network capacity with everything else, which is why it is cheaper and why it can run inside any app that has an internet connection.
A VoIP call has two halves. Signaling sets the call up and tears it down: who is calling whom, which codecs both sides support, and where to send the audio. Media is the audio itself. In the telephony world signaling is usually SIP (RFC 3261) and media travels over RTP (RFC 3550). In the browser and mobile world, WebRTC supplies the media engine, encryption and NAT traversal, and the app chooses its own signaling channel.
- Codecs: G.711 is the classic PSTN codec; Opus (RFC 6716) is the modern default for internet calls and is mandatory in WebRTC
- Quality depends on the network: latency, jitter and packet loss matter more than raw bandwidth, since a voice stream needs only tens of kilobits per second
- Reaching a phone number requires a gateway that bridges IP calls to the PSTN; pure in-app calling between signed-in users does not
- Emergency calling and caller ID rules apply to VoIP services that connect to the telephone network
VoIP today covers everything from hosted business phone systems and SIP trunks to consumer apps and the voice and video calling built into chat products. The common thread is that voice is just another data stream, so it can share a connection, an identity system and a server with messaging, and it can be secured, logged and deployed the same way.
In the Ethora ecosystem
Ethora’s in-app calling is VoIP in the WebRTC sense: voice and video calls run inside your app, between users who are already signed in to your chat, with media routed through an SFU media server so that group calls scale. There is no phone number and no PSTN gateway involved. A call works across web, React Native, iOS and Android through the same SDK as chat, and audio-only calls are available as an opt-in so a product can offer voice without video. See the voice and video call API page for the calling features.
Because calling and chat share one backend, the same rules apply to both: users are provisioned by your system, access follows your app’s own controls, and the deployment can be managed cloud, a dedicated server in your own cloud account, or self-hosted. For regulated teams that is usually the deciding factor: calls stay on infrastructure you control rather than passing through a third-party telephony network.
