diff --git a/player/audio.c b/player/audio.c
index 8422e109f1..598e145147 100644
--- a/player/audio.c
+++ b/player/audio.c
@@ -1225,10 +1225,8 @@ void fill_audio_out_buffers(struct MPContext *mpctx)
     // Even if we're done decoding and syncing, let video start first - this is
     // required, because sending audio to the AO already starts playback.
     if (mpctx->audio_status == STATUS_READY) {
-        if (mpctx->vo_chain && !mpctx->vo_chain->is_coverart &&
-            mpctx->video_status <= STATUS_READY)
-            return;
-        MP_VERBOSE(mpctx, "starting audio playback\n");
+        // Warning: relies on handle_playback_restart() being called afterwards.
+        return;
     }
 
     bool audio_eof = status == AD_EOF;
diff --git a/player/playloop.c b/player/playloop.c
index 2e714823f9..3db5818773 100644
--- a/player/playloop.c
+++ b/player/playloop.c
@@ -976,7 +976,15 @@ static void handle_playback_restart(struct MPContext *mpctx)
             return;
         }
 
+        // Video needed, but not started yet -> wait.
+        if (mpctx->vo_chain && !mpctx->vo_chain->is_coverart &&
+            mpctx->video_status <= STATUS_READY)
+            return;
+
+        MP_VERBOSE(mpctx, "starting audio playback\n");
+        mpctx->audio_status = STATUS_PLAYING;
         fill_audio_out_buffers(mpctx); // actually play prepared buffer
+        mp_wakeup_core(mpctx);
     }
 
     if (!mpctx->restart_complete) {