SAFIRE 4G Audio Tool Box

1
SAFIRE 4G Audio Tool Box

• Concept - Architecture - Building blocks Demo
  applications

2
Introduction

• Target of the AudioToolBox demo bundle is to
• Present SAFIRE-4G framework and IDE/Y development
  environment.
• Present features of the new Y language.
• Present some under development Y features, such
  as Dialog Magic.
• Since both tools/Y language and demo are under
  development, current limitations are
• Demo is based on the IDE0.55/Y0.8 tool
  version
• Only one instance of different AudioToolBox
  building blocks (FSMs) can be used per
  application.

3
Getting started

• Install IDE0.55, Y0.8 and DbgView
• Install AudioToolBox demo
• It will create new Y_atb user
• Open IDE as Y_atb user
• Open workspace Y_Atb and build it, selecting from
  the menu build-gtall
• Run demo applications (right click-gtRun)
• There are two application types (for details see
  next slides)
• Simple applications, without GUI interface
• Simple dialog applications, with basic GUI
  interface
• Run demo applications and enjoy!

4
Audio Tool Box demo

• Workspace organization

5
Basic projects

• Y_Common folder contains definitions of the
• Configuration structure types (ConfigATB_TP)
• Data interface signal list (Data_SL)
• Management interface signal list (MgtExt_SL)
• Y_FSMs folder contains
• Implementation of the different AudioToolBox
  building blocks, such as AudioIn, Volume, etc.

6
Simple applications projects

• Y_MgtFSMs folder contains implementation of the
• Management FSMs
• Y_App folder contains
• Implementation of the different AudioToolBox
  simple demo applications
• Names of the demo applications are self
  explanatory
• For example, App_AudInDelVolOut is
  AudioIn-gtDelay-gtVolume-gtAudioOut

7
Simple dialog applications projects

• Y_DlgFSMs folder contains implementation of the
• Management FSMs containing simple GUI dialogs
• Y_AppDlg folder contains
• Implementation of the different AudioToolBox
  simple demo applications using GUI dialogs
• Names of the demo applications are self
  explanatory
• For example, App_DtmfPiano demonstrates playing
  DTMF tones

8
Audio Tool Box demo

• Concept Architecture

9
Concept Architecture

• FSM based framework
• FSMs implemented in Y language, with built in
  functions in C
• Architecture with 4 building block FSM levels
• Management FSMs (configuration and application
  control)
• Top Level FSMs (audio data generators such as
  AudioInput)
• Middle Level FSMs (audio data processing such
  as Volume)
• Bottom Level FSMs (audio data receivers such as
  WavWriter)
• Bidirectional SIGNAL type interfaces (Data_SL)
  with handshake for inter-FSM data communication
• Bidirectional SIGNAL type interfaces (MgtExt_SL)
  for management inter-FSM communication
• STRING type interfaces for internal FSM
  communication (from the other threads)
• Signaling concept and signal flow are described
  in more details and depicted on the next slides

10
Interface types signal lists

• MgtExt_SL
• SIGNAL Configure( ANY Configuration)
• SIGNAL Start()
• SIGNAL Stop()
• SIGNAL Request( ANY RequestParam)
• SIGNAL Indication( ANY IndicationParam)
• Data_SL
• SIGNAL Act_Req()
• SIGNAL Act_Ind()
• SIGNAL Deact_Req()
• SIGNAL Deact_Ind()
• SIGNAL Data_Req( STRING PDU)
• SIGNAL Data_Ind( STRING PDU)
• STRING
• Data string

11
Signaling MgtExt_SL concept

• Management Level Audio Tool Box FSMs
• Generate Start(), Stop(), Configure(
  CONFIGURATION) and Request(REQUESTPARAM) signals
• Receive Indication( INDICATIONPARAM) signal
• Top Level Audio Tool Box FSMs
• Receive Configure( CONFIGURATION) signal and
  configure itself according to the received
  configuration parameters
• Receive Start() signal, open itself and generates
  Act_Req() signal on the Data_SL interface (to
  open/start data flow)
• Receive Request( REQUESTPARAM) signal and
  performed required action (for example, send
  specified DTMF digit)
• Receive Stop() signal, close itself and generates
  Deact_Req() signal on the Data_SL interface (to
  close/stop data flow)
• Middle Level Audio Tool Box FSMs
• Receive Configure( CONFIGURATION) signal and
  configure itself according to the received
  configuration parameters
• Bottom Level Audio Tool Box FSMs
• Receive Configure( CONFIGURATION) signal and
  configure itself according to the received
  configuration parameters
• Generate Indication( INIDICATIONPARAM) signal, as
  a result of some audio detection (for example
  DTMF digit or silence is detected)

12
Signaling Data_SL concept

• Top Level Audio Tool Box FSMs
• Generate Act_Req() and Deact_Req() signals
• Receive Act_Ind() and Deact_Ind() signals
• Generate Data_Req( PDU) signal containing PCM16
  audio data in PDU
• Signal Data_ind( PDU) currently not used (since
  audio flow is unidirectional
• Middle Level Audio Tool Box FSMs
• Directly forward Act_Req(), Deact_Req(),
  Act_Ind() and Deact_Ind() signal between IN/OUT
  DATA_SL gates
• Receive Data_Req( PDU) signal from the IN gate,
  process PDU and send signal with new PDU to the
  OUT gate
• Signal Data_ind( PDU) currently not used (since
  audio flow is unidirectional)
• Bottom Level Audio Tool Box FSMs
• Receive Act_Req() and Deact_Req() signals
• Generate Act_Ind() and Deact_Ind() signals
• Receive Data_Req( PDU) signal containing PCM16
  audio data in PDU
• Signal Data_ind( PDU) currently not used (since
  audio flow is unidirectional)

13
Signaling sequence (example 1)
14
Signaling sequence (example 2)
15
Implementation

• FSMs implemented in Y
• Additional functionality using C Built In
  functions
• For every FSM test pair FSM is implemented for
  testing purposes

• Connected back-to-back on all FSM interfaces
• Provides minimal features for testing of the
  basic FSM functionality
• Send signals to the FSM under test
• Accept signals from the FSM under test and
  displays them in DbgView window

16
Audio Tool Box demo

• Basic building blocks

17
Main building blocks

• Top Level FSMs
• Input blocks (Data_SL 1 OUT, MgtExt_SL 1
  IN)
• Audio In
• Wav Reader
• DTMF Generator
• Bottom Level FSMs
• Output blocks (Data_SL 1 IN, MgtExt_SL 1
  IN)
• Audio Out
• Wav Writer
• Detector blocks (Data_SL 1 IN, MgtExt_SL 1
  IN)
• DTMF Detector
• Silence Detector
• Middle level FSMs
• Processing blocks (Data_SL 1 IN and 1 OUT, 1
  MgtExt_SL IN)
• Delay
• Volume
• Branch blocks (Data_SL N INs, Data_SL - M
  OUTs)

18
Audio In FSM

• Manage Audio input (microphone)
• Uses PCM 16-bit audio format for PDU
• Pumps audio data packet (Data_Req signal) every
  20ms
• Configuration info
• SEQUENCE
• NUMBER SamplesPerSec
• Config_Aud_T

19
Wav Reader FSM

• Reading from input Wav file (A-Law, 8000 Hz,
  64kbps, mono)
• Pumps audio data packet (Data_Req signal) every
  20ms
• Format of the audio data output (PDU) should be
  configured (ALAW or PCM16)
• Configuration info
• SEQUENCE
• STRING FileName
• STRING Format
• Config_Wav_T

20
DTMF Generator

• Use DTMF tone buffers
• 12 DTMF digits
• Configurable Sample Rate
• Configurable DTMF tone length (in 20ms
  increments)
• Generate audio data output (Data_Req signal)
• 16-bit PCM format for PDU
• Send DTMF Tone (N packets, on 20ms)
• Configuration info
• SEQUENCE
• NUMBER SamplesPerSec
• NUMBER Len20msIncr
• Config_DtmfGen_T

21
Delay

• Receive (from IN gate) and send (to OUT gate)
  Data_Req signals with PDU in PCM 16-bit format
• Delay in 20ms intervals
• for DelayN, N 20ms actual delay
• Incoming audio packet (PDU of the Data_Req
  signal) processing
• Save received Mth packet (PDU) in Packets
  buffer
• Find (M-N)th packet (PDU) stored in Packets
  buffer and send it
• Configuration info
• SEQUENCE
• NUMBER Len20msIncr
• Config_Del_T

22
Volume

• Receive (from IN gate) and send (to OUT gate)
  Data_Req signals with PDU in PCM 16-bit format
• Volume defined as MULTIPLIER / DIVIDER
• Incoming audio packet (PDU of the Data_Req
  signal) processing
• Every sample of the incoming packet (PDU)
  multiplied by volume and stored as output sample
• Send prepared output packet (PDU)
• Configuration info
• SEQUENCE
• NUMBER Multiplier
• NUMBER Divider
• Config_Vol_T

23
Splitter

• Provide Data_SL interface (Data_In_G) branching
  into two Data_SL interfaces (Data_OutA_G and
  Data_OutB_G)
• Receive any Data_SL signal input from Data_In_G
  and forward it to two outputs (A and B)
• Data_OutA_G
• Data_OutB_G
• Receive any Data_SL signal from one of the two
  outputs (A and B) and forward it to
• Data_In_G

24
Mixer 1/2

• Provide merging of two Data_SL interfaces
  (Data_InA_G, Data_InB_G) into one Data_SL
  interface (Data_Out_G)
• Expect audio packets (PDU of the Data_Req signal)
  from both inputs (A and B) to be in PCM 16-bit
  format and of the same size
• A input (Data_InA_G) is master and any signal
  received on it is directly forwarded to the
  output Data_Out_G, and vice versa, with the
  following exception
• PDU of the Data_Req received on Data_InA_G is
  added (sample by sample) to the stored PDU
  (received from B, before) before forwarding to
  the Data_Out_G

25
Mixer 2/2

• B input (Data_InB_G) is slave and signals are not
  forwarded to the Data_Out_G, in order to avoid
  endless loops in case where it is used for
  feedbacks such as Echo application)
• PDU of the received Data_Req is just stored
  (saved)
• Received Act_Req() and Deact_Req() signals are
  just remembered (mark is set)
• Signals received for Data_Out_G are forwarded to
  the slave input (Data_InB_G) only in some
  situations
• Received Act_Ind() or Deact_Ind() signal is
  forwarded only when signal Act_Req() or
  Deact_Ind() signal is received on the Data_InB_G
  and marked, before

26
Audio Out FSM

• Manage Audio output (sound card)
• Receives Data_Req signal, with PDU in audio
  format PCM 16-bit
• Configuration info
• SEQUENCE
• NUMBER SamplesPerSec
• Config_Aud_T

27
Wav Writer FSM

• Writing to output Wav file (A-Law, 8000 Hz,
  64kbps, mono)
• Receives Data_Req signal with PDU in audio format
  that should be configured (ALAW or PCM16)
• Configuration info
• SEQUENCE
• STRING FileName
• STRING Format
• Config_Wav_T

28
DTMF Detector

• Recive Data_Req signal, with PDU in PCM 16-bit
  format
• Process incoming audio packets (PDU of the
  Data_Req signal) to detect DTMF digit using one
  of the available functions (configured in compile
  time)
• Slow (using floating point)
• Fast (using integer math)
• Faster (using integer math and inline assembler)
• Generate Mgt_G signal gt Indication( DTMF digit)
• Configuration info
• SEQUENCE
• NUMBER SamplesPerSec
• NUMBER Threshold
• Config_DtmfDet_T

29
Silence detector

• Receive Data_Req signal, with PDU in PCM 16-bit
  audio format
• Process incoming audio data packets (PDU of the
  Data_Req signal) to detect silence
• Generate Mgt_G signal gt Indication( Silence or
  not)
• Configuration info
• SEQUENCE
• NUMBER AllowedErrors
• NUMBER SilenceLevel
• Config_SilDet_T

30
Audio Tool Box demo

• Simple demo applications

31
Simple Applications - concept

• No GUI interface
• Use management FSMs to implement application
  level. (start/stop application and other actions
  on internal FSM timers)
• Configuration parameters should be changed in
  management FSM Y code
• Use DbgView window to display/check information
• Run DbgView with Print filter in order to
  remove unnecessary debug info

32
Simple Applications - Management

• Initiates actions using internal timers
• According to the initiated action, appropriate
  management signal (MgtExt_SL) is generated and
  sent to the Audio Tool Box FSM(s).
• On the received management signal (MgtExt_SL)
  from some Audio Tool Box FSM, result is displayed
  in the DbgView window

33
Simple Applications (1/3)

• App_AudInOut
• Uses MgtATB_1_FSM for management
• AudioIn -----gt AudioOut
• App_AudInVolOut
• Uses MgtATB_2_FSM for management
• AudioIn -----gt Volume -----gt AudioOut
• App_AudInDelVolOut
• Uses MgtATB_2_FSM for management
• AudioIn -----gt Delay -----gt Volume -----gt
  AudioOut
• App_WavRdWr
• Uses MgtATB_1_FSM for management
• WavRd -----gt WavWr
• App_AudInSplitAudOutWavWr
• Uses MgtATB_1_FSM for management
• AudioIn -----gt Splitter -----gt AudioOut

34
Simple Applications (2/3)

• App_GenDtmfAudOut
• Uses MgtATB_1_FSM for management
• GenDtmf -----gt AudioOut
• App_GenDtmfDetDtmf
• Uses MgtATB_1_FSM for management
• GenDtmf -----gt DetDtmf
• App_GenDtmfDetSil
• Uses MgtATB_1_FSM for management
• GenDtmf -----gt DetSilence
• App_GenDtmfSplitDetDtmfSil
• Uses MgtATB_1_FSM for management
• DtmfGen -----gt Splitter -----gt DetSilence
• -----gt
  DetDtmf

35
Simple Applications (3/3)

• App_AudFunnyEcho
• Uses MgtATB_1_FSM for management
• AudioIn -----gt Mixer -----gt Splitter -----gt
  AudioOut
• -----gt
  -----gt
• lt---- Volume lt---- Delay
  lt-----
• lt
• App_DtmfSong
• Uses MgtATB_2_FSM for management
• GenDtmf -----gt AudioOut

36
Audio Tool Box demo

• Simple dialog demo applications

37
Simple Dialogs

• Implemented using Y Dialog Magic features
• Simple MENU
• About, about SAFIRE 4G
• Help, provides specific application details and
  explanations
• Exit, exit application

38
Dialog Applications - Management

• Management FSMs incorporates required DIALOGs.
• User initiates actions using dialog buttons/menu
• According to the initiated action, appropriate
  management signal (MgtExt_SL) is generated and
  sent to the Audio Tool Box FSM(s).
• On the received management signal (MgtExt_SL)
  from some Audio Tool Box FSM, result is displayed
  in the dialog

39
Simple Dialog Applications

• SPEAKER (App_DlgSpeaker)
• Play with your voice, change volume and delay
• PLAYER (App_DlgPlayer)
• Play WAV file songs from our small collection
• RECORDER (App_DlgRecorder)
• Record your voice in WAV file
• ECHO (App_DlgEcho)
• Play with echo effect, change delay and volume of
  the audio feedback
• KARAOKE (App_DlgKaroke)
• Sing with music matrix, change volume of your
  voice.
• DTMF PIANO (App_DlgPiano)
• Play your own music using DTMF tones.
• Architecture (used Audio Tool Box building
  blocks, connections) of every application is
  depicted on the next slides.

40
1. SPEAKER (App_DlgSpeaker)
41
2. PLAYER (App_DlgPlayer)
42
3. RECORDER (App_DlgRecorder)
43
4. ECHO (App_DlgEcho)
44
5. KARAOKE (App_DlgKaraoke)
45
6. DTMF PIANO (App_DtmfPiano)