Voximplant. Blog

Category Archives: News

Hot Summer Speech-to-Text Updates


Following Google’s release of new Speech API, we are happy to announce improved quality of call records transcription. No changes are required, your old code like will be producing more accurate results. In addition to that, you can now specify output format to get transcription in form of JSON file with timestamps for further processing. We are working on some examples that will be available in HowTo’s in the nearest future.

H.264 video call recording is now available (in mp4 format)

h264 avc mp4We are happy to announce that video calls that use H.264 video codec can now be recorded. Recorded video calls that use H.264 will be stored as mp4 files (calls with video in VP8 format are stored as webm files). There are few ways to enable H.264 in your video calls:
– Use H264first Config param (Web SDK)
– Use H264first flag in CallSettings (Web SDK)
– Mobile SDKs are being updated to let developers enable H264 in video calls

According to the following research from Chris Koehncke using H.264 instead of VP8 can save up to 50% in terms of power consumption https://www.chriskranky.com/webrtc-impacts-on-battery-life/ , in addition to that mp4/avc is widely used and supported by most players in both online and offline worlds.

See the example of video recording below (Firefox->Voximplant, no post-processing)

P.S. Feel free to ping us if you have any problems with video call recording

Voximplant mobile SDK for Unity in now available

Voximplant Unity SDK
We are happy to announce our new mobile SDK for Unity platform. Now Unity developers can use the SDK to embed real-time voice and video communication into VR/AR apps and games in minutes, we will take care of complexity and infrastructure. The SDK is compatible with Google Daydream. Please visit https://unity.voximplant.com/ to learn more.

Tagged in , , , , ,

VoIP Push Notifications support for iOS SDK has arrived

Dear Voximplant developers, we are happy to announce that Voximplant now supports VoIP Push Notifications for calls to iOS SDK. We tried to make everything as simple for developer as possible; the following steps are required to enable VoIP Push Notifications for your app:

1. Upload your Apple VoIP Push certificate at https://manage.voximplant.com/#certificates and specify (bind) Voximplant application(s) where you will be using VoIP Push functionality

2. Add

require(Modules.PushService);

in the beginning of your VoxEngine scenario

3. Now VoxEngine’s callUser and callUserDirect functions will start using VoIP Pushes to notify the SDK about incoming calls. (same applies to helper functions: forwardCallToUser and forwardCallToUserDirect)

4. Get new version of iOS SDK at http://voximplant.com/docs/references/mobilesdk/ios/

That’s it!

P.S. Android Push support is on its way to production, expect it right after the New Year in January 2017!

New billing for video calls running via servers

We are happy to announce that currently video calls that go via Voximplant servers (not P2P ones) are billed per megabyte of video traffic. Since the resolution of a video stream can be different (so as fps, video codec and other params) it means that per minute billing was a bad idea for this case – 1 minute of video call generates different load on our infrastructure depending on various parameters. So now the formula for video call going via Voximplant servers looks as follows:

SDK-platform / platform-SDK call – $0.002 per minute per call leg
+ $0.001 per megabyte of video traffic sent to the platform

See http://voximplant.com/pricing/ for the details

Tagged in , ,

High quality Speech Recognition is now available

RecognitionHi everyone! We’ve been busy working on new cool features for you, many developers asked us about high quality speech recognition for both audio call records transcription and real-time recognition scenarios. We are happy to announce that this functionality is now available for all our developers.

Voximplant Web SDK v4 beta is now available with Edge/ORTC support

WebSDKv4We have released new version of Web SDK, it has been fully rewritten. New version will help us to accelerate the development process and includes a lot of new features and improvements. Key features of Voximplant’s Web SDK 4.0 include:

  • Support for audio calls in Microsoft Edge. With Microsoft recently updating the browser’s web and audio support, Voximplant is one of the first companies to make its platform compatible with Edge. With Web SDK 4.0, web apps are empowered to handle incoming and outgoing voice calls from within the Edge web browser. Video calls in Edge will be functional as soon as VP9 or h.264 codec support is added to the browser.
  • Video can be enabled during an existing audio call. Now developers can create web apps that offer a Skype-like “enable video” option during active audio calls. Video can be enabled or disabled at any time during the call.
  • Both local and remote audio and video streams can be modified by applying filters. Developers can now modify video and audio streams in realtime. Possible applications of this feature include using video and audio filters to add watermarks, hide or alter faces, mask voices, etc.
  • h.264 video codec can be set as “high priority”. The “h.264” video codec is known for its valuable speed in modern hardware, especially in mobile devices. Developers can now force this codec to provide better video quality with less CPU usage.

Additional enhancements include:

  • Reduced latency of audio and video calls.
  • Improved syncing between audio and video tracks.
  • Calls within the Chrome browser now maintain better performance – even with poor network conditions.
  • Peer-to-peer connections can now be enabled in scenarios using JavaScript in the cloud.
  • Audio and video codec priorities can be manually configured.
  • Full support for WebRTC to offer users the best in voice and video quality.

New SDK also supports asynchronous initialization and drops support for Flash, which can still be used with legacy systems via 3.x versions of the SDK. Complete documentation for Voximplant’s Web SDK 4.0 beta can be found at https://voximplant.com/docs/references/websdk4/

HD Audio Playback / Streaming

This post is short and simple: your mp3 or ogg files played on VoxEngine scenario level with call.startPlayback or using Player will be played on the Web or Mobile SDK side in HD quality (48KHz), or on SIP side if it does support wideband audio codecs (Speex or Opus). It also appeared that Opus has 3 encoding presets – auto / speech / music, currently we use auto, but maybe we will let developers decide which preset can be used on VoxEngine scenario level later.

HD Audio Conferencing

HD Audio Conference

We were quite busy working on HD audio conferencing support and now it’s available for all VoxImplant developers. In HD mode audio is being mixed at 48KHz, all audio sources with lower sample rate will be resampled to 48KHz. Web and mobile SDK uses Opus audio codec and offers the best possible audio quality. Up to 100 participants can be connected to a conference.

It’s simple to enable HD mode for a conference in VoxEngine scenario:

// HD mode can be enabled for the conference by passing hd_audio parameter set to true
var conf = VoxEngine.createConference({ hd_audio: true });

You can learn more about building audio conferencing services using VoxImplant in our tutorial. HD audio conference demo available at https://demos.voximplant.com/hdconf/ (optimized for WebRTC browsers) , the demo app source code is available at https://github.com/voximplant/voxhdconf

HD Audio Recording

HD_recording

We’ve been offering audio recording for a while, but our audio recorder was saving audio in 8KHz, since it’s the most used format in good old telephony and when we have calls to/from PSTN. When we work with VoIP endpoints that do support wideband audio (WebRTC, SIP, mobile SDKs) we can save audio with much better quality without downsampling it to 8KHz. We chose 48 KHz as the base sample rate for HD audio recorder, since WebRTC/Opus can offer this quality, audio from endpoints with lower sample rate will be re-sampled.

It’s easy to use HD audio recording, just add hd_audio parameter in the recorder params:

  1. // Example for the record method of the Call class function
  2. call.record({hd_audio: true});
  3. // Example for the Recorder class
  4. require(Modules.Recorder);
  5. // ...
  6. var recorder = VoxEngine.createRecorder({hd_audio: true});
  7. call.sendMediaTo(recorder);

Feel free to start recording your calls in HD today, wideband audio conferencing is coming soon, stay tuned!

Older posts

Categories

Recent Posts

Aug 17, 2017 11:59
Messaging update
Aug 14, 2017 03:22
Hot Summer Speech-to-Text Updates
Aug 02, 2017 12:47
Screen Sharing in VoxImplant
May 02, 2017 12:08
H.264 video call recording is now available (in mp4 format)
Apr 17, 2017 10:58
Voximplant mobile SDK for Unity in now available
Feb 13, 2017 10:22
Authorization using tokens instead of passwords
Feb 03, 2017 01:45
Push Notifications for Android
Feb 03, 2017 01:45
Push Notifications for iOS
Jan 19, 2017 08:38
Speech-to-text: ASR
Jan 19, 2017 08:38
Speech-to-text: transcription

Archive

Tags