I have been trying to follow the example in this post.
Since I am not trying to implement this in a service but rather in a standard activity, I haven’t experienced the problems described in the aforementioned post.
I keep getting, however, “No voice results” – as implemented in that post, when getStringArrayList(RecognizerIntent.EXTRA_RESULTS) returns null.
Clearly, I am missing someting in what needs to be done in addition to
recognizer.setRecognitionListener(listener);
recognizer.startListening(intent);
What am I missing?
Is it possible that in addition to startListening() I also need to startActivityForResult()? If so, I tried this already but it invoked the full Google’s Voice Search activity (which is what I am trying to avoid, just as @vladimir.vivien wrote here). This creates even more problems due to 2 recognizers running at the same time…
At first I thought that what’s missing is actual submission to Google’s servers, but when I examine the LogCat output from the start of the speech recognition session till its end (see below), I see that it actually creates a TCP session with http://www.google.com/m/voice-search .
So the obvious question is what am I missing?
04-18 07:02:17.770: INFO/RecognitionController(623): startRecognition(#Intent;action=android.speech.action.RECOGNIZE_SPEECH;S.android.speech.extra.LANGUAGE_MODEL=free_form;S.android.speech.extra.PROMPT=LEARNSR;S.calling_package=com.example.learnsr.SrActivity;end)
04-18 07:02:17.770: INFO/RecognitionController(623): State change: STARTING -> STARTING
04-18 07:02:17.780: INFO/AudioHardwareQSD(121): Routing audio to Speakerphone
04-18 07:02:17.780: DEBUG/AudioHardwareQSD(121): Switching audio device to
04-18 07:02:17.780: DEBUG/AudioHardwareQSD(121): Speakerphone
04-18 07:02:17.780: INFO/AudioHardwareQSD(121): AudioHardware PCM record is going to standby.
04-18 07:02:17.780: INFO/AudioHardwareQSD(121): Routing audio to Speakerphone
04-18 07:02:17.780: DEBUG/AudioHardwareQSD(121): Switching audio device to
04-18 07:02:17.780: DEBUG/AudioHardwareQSD(121): Speakerphone
04-18 07:02:17.780: INFO/AudioHardwareQSD(121): AudioHardware PCM record is going to standby.
04-18 07:02:17.780: INFO/AudioService(164): AudioFocus requestAudioFocus() from android.media.AudioManager@46036948
04-18 07:02:17.780: DEBUG/AudioFlinger(121): setParameters(): io 3, keyvalue routing=262144;vr_mode=1, tid 155, calling tid 121
04-18 07:02:17.790: INFO/AudioHardwareQSD(121): Routing audio to Speakerphone
04-18 07:02:17.790: INFO/AudioHardwareQSD(121): do input routing device 40000
04-18 07:02:17.790: INFO/AudioHardwareQSD(121): Routing audio to Speakerphone
04-18 07:02:17.790: INFO/RecognitionController(623): State change: STARTING -> RECOGNIZING
04-18 07:02:17.790: INFO/ServerConnectorImpl(623): Starting TCP session, url=http://www.google.com/m/voice-search
04-18 07:02:17.930: DEBUG/ServerConnectorImpl(623): Created session a7918495c042db1746d3e09514baf621
04-18 07:02:17.930: INFO/ServerConnectorImpl(623): Creating TCP connection to 74.125.115.126:19294
04-18 07:02:17.980: DEBUG/AudioHardwareQSD(121): Switching audio device to
04-18 07:02:17.980: DEBUG/AudioHardwareQSD(121): Speakerphone
04-18 07:02:18.070: INFO/ServerConnectorImpl(623): startRecognize RecognitionParameters{session=a7918495c042db1746d3e09514baf621,request=1}
04-18 07:02:18.390: INFO/RecognitionController(623): onReadyForSpeech, noise level:10.29969, snr:-0.42756215
04-18 07:02:19.760: DEBUG/dalvikvm(659): GC_EXPLICIT freed 5907 objects / 353648 bytes in 67ms
04-18 07:02:21.030: INFO/AudioHardwareQSD(121): AudioHardware pcm playback is going to standby.
04-18 07:02:24.090: INFO/RecognitionController(623): onBeginningOfSpeech
04-18 07:02:24.760: DEBUG/dalvikvm(669): GC_EXPLICIT freed 1141 objects / 74296 bytes in 48ms
04-18 07:02:25.080: INFO/RecognitionController(623): onEndOfSpeech
04-18 07:02:25.080: INFO/AudioService(164): AudioFocus abandonAudioFocus() from android.media.AudioManager@46036948
04-18 07:02:25.140: INFO/AudioHardwareQSD(121): Routing audio to Speakerphone
04-18 07:02:25.200: INFO/RecognitionController(623): State change: RECOGNIZING -> RECOGNIZED
04-18 07:02:25.200: INFO/RecognitionController(623): Final state: RECOGNIZED
04-18 07:02:25.260: INFO/ServerConnectorImpl(623): ClientReport{session_id=a7918495c042db1746d3e09514baf621,request_id=1,application_id=intent-speech-api,client_perceived_request_status=0,request_ack_latency_ms=118,total_latency_ms=7122,user_perceived_latency_ms=116,network_type=1,endpoint_trigger_type=3,}
04-18 07:02:25.260: INFO/AudioService(164): AudioFocus abandonAudioFocus() from android.media.AudioManager@46036948
04-18 07:02:25.270: DEBUG/AudioHardwareQSD(121): Switching audio device to
04-18 07:02:25.270: DEBUG/AudioHardwareQSD(121): Speakerphone
04-18 07:02:25.270: INFO/RecognitionController(623): State change: RECOGNIZED -> PAUSED
04-18 07:02:25.270: INFO/AudioService(164): AudioFocus abandonAudioFocus() from android.media.AudioManager@46036948
04-18 07:02:25.270: INFO/ClientReportSender(623): Sending 1 client reports over HTTP
04-18 07:02:25.280: INFO/AudioHardwareQSD(121): AudioHardware PCM record is going to standby.
04-18 07:02:25.280: DEBUG/AudioFlinger(121): setParameters(): io 3, keyvalue routing=0, tid 155, calling tid 121
04-18 07:02:25.280: INFO/AudioHardwareQSD(121): Routing audio to Speakerphone
04-18 07:02:25.280: INFO/AudioHardwareQSD(121): Routing audio to Speakerphone
04-18 07:02:25.280: DEBUG/AudioHardwareQSD(121): Switching audio device to
04-18 07:02:25.280: DEBUG/AudioHardwareQSD(121): Speakerphone
04-18 07:02:25.280: INFO/AudioHardwareQSD(121): AudioHardware PCM record is going to standby.
According to the documentation of the listener you need to request the results with SpeechRecognizer.RESULTS_RECOGNITION from the bundle given to onResults(). Have you tried that?
RecognizerIntent.EXTRA_RESULTS is to be used when using the RECOGNIZE_SPEECH intent.