diff --git a/src/main/java/eu/siacs/conversations/ui/RtpSessionActivity.java b/src/main/java/eu/siacs/conversations/ui/RtpSessionActivity.java
index 33d17315d..ba22604e4 100644
--- a/src/main/java/eu/siacs/conversations/ui/RtpSessionActivity.java
+++ b/src/main/java/eu/siacs/conversations/ui/RtpSessionActivity.java
@@ -25,6 +25,7 @@ import org.webrtc.VideoTrack;
 
 import java.lang.ref.WeakReference;
 import java.util.Arrays;
+import java.util.List;
 import java.util.Set;
 
 import eu.siacs.conversations.Config;
@@ -108,7 +109,13 @@ public class RtpSessionActivity extends XmppActivity implements XmppConnectionSe
     }
 
     private void requestPermissionsAndAcceptCall() {
-        if (PermissionUtils.hasPermission(this, ImmutableList.of(Manifest.permission.RECORD_AUDIO), REQUEST_ACCEPT_CALL)) {
+        final List<String> permissions;
+        if (getMedia().contains(Media.VIDEO)) {
+            permissions = ImmutableList.of(Manifest.permission.CAMERA, Manifest.permission.RECORD_AUDIO);
+        } else {
+            permissions = ImmutableList.of(Manifest.permission.RECORD_AUDIO);
+        }
+        if (PermissionUtils.hasPermission(this, permissions, REQUEST_ACCEPT_CALL)) {
             //TODO like wise the propose; we might just wait here for the audio manager to come up
             putScreenInCallMode();
             requireRtpConnection().acceptCall();
@@ -285,6 +292,7 @@ public class RtpSessionActivity extends XmppActivity implements XmppConnectionSe
             putScreenInCallMode();
         }
         binding.with.setText(getWith().getDisplayName());
+        updateVideoViews();
         updateStateDisplay(currentState);
         updateButtonConfiguration(currentState);
     }
@@ -300,26 +308,6 @@ public class RtpSessionActivity extends XmppActivity implements XmppConnectionSe
         setIntent(intent);
     }
 
-    private void updateVideoViews() {
-        final Optional<VideoTrack> localVideoTrack = requireRtpConnection().geLocalVideoTrack();
-        if (localVideoTrack.isPresent()) {
-            ensureSurfaceViewRendererIsSetup(binding.localVideo);
-            //paint local view over remote view
-            binding.localVideo.setZOrderMediaOverlay(true);
-            binding.localVideo.setMirror(true);
-            localVideoTrack.get().addSink(binding.localVideo);
-        } else {
-            binding.localVideo.setVisibility(View.GONE);
-        }
-        final Optional<VideoTrack> remoteVideoTrack = requireRtpConnection().getRemoteVideoTrack();
-        if (remoteVideoTrack.isPresent()) {
-            ensureSurfaceViewRendererIsSetup(binding.remoteVideo);
-            remoteVideoTrack.get().addSink(binding.remoteVideo);
-        } else {
-            binding.remoteVideo.setVisibility(View.GONE);
-        }
-    }
-
     private void ensureSurfaceViewRendererIsSetup(final SurfaceViewRenderer surfaceViewRenderer) {
         surfaceViewRenderer.setVisibility(View.VISIBLE);
         try {
@@ -477,6 +465,26 @@ public class RtpSessionActivity extends XmppActivity implements XmppConnectionSe
         this.binding.inCallActionRight.setVisibility(View.VISIBLE);
     }
 
+    private void updateVideoViews() {
+        final Optional<VideoTrack> localVideoTrack = requireRtpConnection().geLocalVideoTrack();
+        if (localVideoTrack.isPresent()) {
+            ensureSurfaceViewRendererIsSetup(binding.localVideo);
+            //paint local view over remote view
+            binding.localVideo.setZOrderMediaOverlay(true);
+            binding.localVideo.setMirror(true);
+            localVideoTrack.get().addSink(binding.localVideo);
+        } else {
+            binding.localVideo.setVisibility(View.GONE);
+        }
+        final Optional<VideoTrack> remoteVideoTrack = requireRtpConnection().getRemoteVideoTrack();
+        if (remoteVideoTrack.isPresent()) {
+            ensureSurfaceViewRendererIsSetup(binding.remoteVideo);
+            remoteVideoTrack.get().addSink(binding.remoteVideo);
+        } else {
+            binding.remoteVideo.setVisibility(View.GONE);
+        }
+    }
+
     private void disableMicrophone(View view) {
         JingleRtpConnection rtpConnection = requireRtpConnection();
         rtpConnection.setMicrophoneEnabled(false);
diff --git a/src/main/java/eu/siacs/conversations/xmpp/jingle/JingleRtpConnection.java b/src/main/java/eu/siacs/conversations/xmpp/jingle/JingleRtpConnection.java
index b762ab7ec..317dd7cf0 100644
--- a/src/main/java/eu/siacs/conversations/xmpp/jingle/JingleRtpConnection.java
+++ b/src/main/java/eu/siacs/conversations/xmpp/jingle/JingleRtpConnection.java
@@ -820,7 +820,13 @@ public class JingleRtpConnection extends AbstractJingleConnection implements Web
     }
 
     private void setupWebRTC(final Set<Media> media, final List<PeerConnection.IceServer> iceServers) throws WebRTCWrapper.InitializationException {
-        this.webRTCWrapper.setup(this.xmppConnectionService);
+        final AppRTCAudioManager.SpeakerPhonePreference speakerPhonePreference;
+        if (media.contains(Media.VIDEO)) {
+            speakerPhonePreference = AppRTCAudioManager.SpeakerPhonePreference.SPEAKER;
+        } else {
+            speakerPhonePreference = AppRTCAudioManager.SpeakerPhonePreference.EARPIECE;
+        }
+        this.webRTCWrapper.setup(this.xmppConnectionService, speakerPhonePreference);
         this.webRTCWrapper.initializePeerConnection(media, iceServers);
     }
 
diff --git a/src/main/java/eu/siacs/conversations/xmpp/jingle/WebRTCWrapper.java b/src/main/java/eu/siacs/conversations/xmpp/jingle/WebRTCWrapper.java
index 8562f2caa..8e87291cd 100644
--- a/src/main/java/eu/siacs/conversations/xmpp/jingle/WebRTCWrapper.java
+++ b/src/main/java/eu/siacs/conversations/xmpp/jingle/WebRTCWrapper.java
@@ -145,14 +145,14 @@ public class WebRTCWrapper {
         this.eventCallback = eventCallback;
     }
 
-    public void setup(final Context context) {
+    public void setup(final Context context, final AppRTCAudioManager.SpeakerPhonePreference speakerPhonePreference) {
         PeerConnectionFactory.initialize(
                 PeerConnectionFactory.InitializationOptions.builder(context).createInitializationOptions()
         );
         this.eglBase = EglBase.create();
         this.context = context;
         mainHandler.post(() -> {
-            appRTCAudioManager = AppRTCAudioManager.create(context, AppRTCAudioManager.SpeakerPhonePreference.EARPIECE);
+            appRTCAudioManager = AppRTCAudioManager.create(context, speakerPhonePreference);
             appRTCAudioManager.start(audioManagerEvents);
             eventCallback.onAudioDeviceChanged(appRTCAudioManager.getSelectedAudioDevice(), appRTCAudioManager.getAudioDevices());
         });