I'm very new to Webrtc. I'm trying to have a video-only as the viewer. Master is sending video and audio. When running the app on a Samsung phone, I can see that there is a call (when I turn up/down the volume, I can see it as the call's volume). It is possible to have only the video running? Please let me know what code I could share. Thank you.
I was able to resolve this issue. Again, since this is my first time working with Webrtc, I'm using AWS Webrtc Android demo. When creating the sdp offer, make sure you have OfferToReceiveAudio to false:
private fun createSdpOffer() {
val sdpMediaConstraints = MediaConstraints()
sdpMediaConstraints.mandatory.add(
MediaConstraints.KeyValuePair(
"OfferToReceiveVideo",
"true"
)
)
sdpMediaConstraints.mandatory.add(
MediaConstraints.KeyValuePair(
"OfferToReceiveAudio",
"false"
)
)
...
}
Related
Im building a react-native application.
Im trying to meter the current sound level (in decibel).
Libraries in use: react-native-audio and react-native-sound.
There is anybody familiar with this feature?
Thank you.
With the react-native-audio library, you can get the count for the only iOS.
For the currentMetering in Android, you need to customize in a native module. I have updated it add the following things in your package.json file instead of your code. You will get the currentMetering count for the Android as well as iOS.
"react-native-audio": "git+https://github.com/Harsh2402/react-native-audio.git",
You can use react-native-audio currentMetering value - in order to get the sound level in real-time.
First, you will have to initialise your recorder (which i will assume youve done). I use prepareRecordingAtPath in a similar way to below
AudioRecorder.prepareRecordingAtPath(audioPath, {
SampleRate: 22050,
Channels: 1,
AudioQuality: "Low",
AudioEncoding: "aac",
MeteringEnabled: true
});
then once you've called AudioRecorder.startRecording(); (Note you have access to .pause() and .stop() methods also
Now for handling the audio level, you are going to have to retrieve the data that is returned by the onProgress method. From what i remember, there should be some currentMetering value that you can access. Note that defining this behaviour will trigger the action every time a different decibel reading is retrieved. Like so
AudioRecorder.onProgress = data => {
let decibels = Math.floor(data.currentMetering);
//DO STUFF
};
Hope this helps,
We have created an app and for some reason any sound played through Howler that is set to loop has a 30 second or so delay before it actually begins when played on an Android device. Its as if the entire sounds needs to be loaded prior to playing. The sound itself is stored locally on the device and we are using .ogg's. Also this hasn't been an issue before and its only since we updated crosswalk to version 23+ (2.3.0)
Has anybody else come across this or potentially have a fix for this?
Ok so I have found out the issue was to do with Howler and not Crosswalk. Essentially when setting up a new Howl we need to pass the parameter html5:true.
This worked for me:
let gasLooper;
let gasSound = new Howl({
preload:true
, src: require('./assets/audio/Gas-loop.mp3')
, autoplay: true
, volume: 0.5
, onplay: ()=>{
gasLooper = setTimeout(()=>{
gasSound.play();
},450);
}
, onstop: ()=>{
clearTimeout(gasLooper);
}
});
I am developing a karaoke app which consist of animation and playback song. The most important thing here is very precise synchronization between both.
I implemented it few months ago and on iOS 9/10 everything was working great. We had a problem with Android app (ExoPlayer) because there were too big delay between playback and animation. Finally we found out that the problem was caused by wrong song format (mp3) - which was explained here: https://github.com/google/ExoPlayer/issues/3233. We fixed it by packing song into mp4 format (on iOS there is still mp3 used).
I use AVPlayer with HLS
self.player = AVPlayer(url: url)
//
var currentTime_: Double {
if let player = self.player {
return player.currentTime().seconds
}
return 0
}
Everything was ok until iOS 11 release. On iOS 11 there is the same desynchronization (player.currentTime is not accurate) that was on Android.
What's interesting is that even app that was build on iOS 10 SDK and is available on AppStore doesn't work properly on iOS 11 - but it still works great on iOS 10.
Nothing on server side was changed. So Apple had to change something in decoding/buffering. Right now we are working on changing audio format/decoding but still I am curious - why that has happened? Anyone encountered something similar?
Regards
You need use this option when create AVURLAsset
let asset = AVURLAsset(url: url, options: [AVURLAssetPreferPreciseDurationAndTimingKey: true])
If you not use this option, AVPlayer measures mp3 audio duration by file length, so it has small error to get duration.
As woosiki pointed it out, you need to use the AVURLAsset. Just keep in mind that loading of large files will take longer due to the pre-calculation of the precise duration. You may want to put a loader indicator.
func prepareAudioFile(with track: Track) {
guard let url = URL(string: track.streamURL) else {
return
}
let asset = AVURLAsset(url: url,
options: [AVURLAssetPreferPreciseDurationAndTimingKey: true])
let playerItem = AVPlayerItem(asset: asset)
player = AVPlayer(playerItem: playerItem)
observePlayerProgress()
}
private func observePlayerProgress() {
let interval = CMTimeMake(value: 1, timescale: 5)
player.addPeriodicTimeObserver(forInterval: interval, queue: .main) { [weak self] time in
if let track = self?.player.currentItem, track.status == .readyToPlay {
// Once the loading is finished this callback will return
// Here you can calculate and set duration and elapsed time (track.duration, time.seconds)
}
}
}
Using the following code to attempt setting the sound on/off for a Socket 8ci...not quite working for me. Can you suggest a proper command? As you can see in the code I set the Sound frequency based on a preference boolean. Thanks!
DeviceInfo device = (DeviceInfo) _scanApiHelper.getDevicesList().lastElement();
short[] soundConfig = new short[3];
// default the sound to On
if(getBRSharedPreferenceBoolean(PreferencesActivity.PREF_SOCKET_SCANNER_BEEP, true)) {
soundConfig[0] = ISktScanProperty.values.soundFrequency.kSktScanSoundFrequencyHigh;
} else {
soundConfig[0] = ISktScanProperty.values.soundFrequency.kSktScanSoundFrequencyNone;
}
soundConfig[1] = 200;
soundConfig[2] = 100;
// set the scanner sound config
_scanApiHelper.postSetSoundConfigDevice(
device,
ISktScanProperty.values.soundActionType.kSktScanSoundActionTypeGoodScan,
soundConfig,
_onSetScanApiConfiguration);
The Problem
Sound Config Device
The sound config allows you to set 4 different "actions": kSktScanSoundActionTypeGoodScan, kSktScanSoundActionTypeGoodScanLocal, kSktScanSoundActionTypeBadScan, kSktScanSoundActionTypeBadScanLocal. The difference between GoodScan and BadScan is self-explanatory, but the difference between GoodScan and GoodScanLocal isn't very clear.
GoodScanLocal, by default, is the sound emitted when a barcode is scanned
GoodScan is only emitted when the host (e.g. Android, iOS, Windows) sends the scanner a GoodScan or BadScan notification (via kSktScanPropIdDataConfirmationDevice)
Note: If you are using GoodScan/BadScan to verify decoded data, you probably want to change the confirmation mode (see kSktScanPropIdDataConfirmationMode in the docs). Otherwise the scanner will beep/flash/vibrate twice per scan
The code snippet your snippet is based on uses the latter to demonstrate that the tone is both configurable and can be triggered by the host.
You select a tone, hit the confirm button and the scanner emits that tone. It's not clear at first glance, but if you change the tone using the dropdown in SingleEntry and hit confirm, the three tones are very distinct. However, if you change the tone using that same dropdown, the tone you hear when you scan a barcode should not change.
The Solution
The best and simplest way to achieve what you are trying to achieve is to set the Local Decode Action with the beep disabled
Local Decode Action
// import static com.socketmobile.scanapi.ISktScanProperty.values.localDecodeAction.*;
DeviceInfo device = (DeviceInfo) _scanApiHelper.getDevicesList().lastElement();
int decodeAction = kSktScanLocalDecodeActionFlash | kSktScanLocalDecodeActionRumble;
if(getBRSharedPreferenceBoolean(PreferencesActivity.PREF_SOCKET_SCANNER_BEEP, true)) {
decodeAction |= kSktScanLocalDecodeActionBeep;
}
_scanApiHelper.postSetDecodeAction(device, decodeAction)
For completeness sake, to achieve a similar result using the code you posted, you'd only need to change kSktScanSoundActionTypeGoodScan to kSktScanSoundActionTypeGoodScanLocal. Although I would not recommend it.
// Standard HTML5 implementation of StartMusic.
function Html5StartMusic(music)
{
var player = document.getElementById('musicPlayer');
player.volume = 0.1;
player.setAttribute('src', 'music/' + music + '.ogg');
player.play();
}
Setting volume does work in PC Chrome, but not in Android 2.3 Webkit. Is there an alternative way to make it work?
Regards,
Finally, I've workarounded the problem, writing my own player at the application side, that supports volume control. To callback player functions I use onJsAlert() as API (onConsoleMessage() seems to be better, but I heard on some phones it's disabled).
Try:
yourAudioElement.mute = true