Slide 1

Acoustic Basics and Detector Settings: an overview of recording high quality bat calls and using Sonobat software to identify species Bryce Maxell, Senior Zoologist (406) 444-3655 bmaxell@mt.govbmaxell@mt.gov February 24 th 2015, Helena, Montana http://mtnhp.org Slide 2 Basics of Sonobat 3.05 and Full Spectrum Call Analysis (Slides from Joe Szewczak) Slide 3 Bat echolocation call data begins with vibrations of the bats larynx imparting pressure waves on the air (i.e., sound) that propagate through the air at high velocity (i.e., the speed of sound). Plot of pressure oscillation in air per time measured from a fixed point. A sensor at a fixed point will oscillate sympathetically as these waves pass. Plot of voltage oscillation in detector circuitry per time acquired from microphone. Slide 4 Zero-crossing analysis Divide by 8 count every 8 period oscillations Measure the time to make every 8 oscillations; this corresponds to the average frequency over the interval of those 8 oscillations....... frequency time Then plot those avg. frequencies per time. Slide 5 Period counting is independent of signal strength. frequency time...... Zero-crossing analysis Strong or weak signals with the same frequency content would plot equivalently. Slide 6 Multiple frequency content and zero-crossing analysis Multiple signal sources contribute to real world soundscapes, e.g., cascading water, vehicles on a roadway, wind blowing past vegetation or structures, insects, and perhaps bats. Pressure oscillations from multiple sources interact and combine to form a single signal. This example soundscape has two high frequency sources from bats (blue and green), and a stronger lower frequency source. In the combined signal received by the microphone, the stronger lower frequency signal overpowers the higher frequency signals and controls the zero axis crossing, rendering the bat signals undetectable by zero-crossing. Slide 7 Multiple frequency content and zero-crossing analysis Zero-crossing analysis can only detect the dominant, i.e., strongest, frequency content of any signal. Any other signals in the soundscape remain invisible to zero-crossing analysis. Full-spectrum analysis can access multiple frequency content to reveal the entirety of bat calls even when the signal strength of all or part of the a call falls below other signals in the soundscape. Full-spectrum analysis would reveal the bats in this signal. Slide 8 Full-spectrum analysis amplitude time window sections Full-spectrum analysis assembles these to construct a representation of the entire soundscape by first passing the signal through bandpass frequency filters and repeating this process for each frequency band. Full-spectrum analysis extracts frequency and amplitude content by sampling overlapping snippets (windows) of the waveform. Slide 9 Full-spectrum analysis frequency band time amplitude An example of a full-spectrum rendered soundscape. A sonogram displays a flat plot of this type of data with the amplitude mapped in color Slide 10 The scattered points at the beginning and end result from Z-C interpretation of lower amplitude background noise, which Z-C cannot readily discard because it cannot interpret relative amplitude of the signals that generated these points. Full-spectrum & zero-crossing analysis compared Slide 11 Pipistrellus hesperus & subflavus (pipistrelle) Slide 12 Slide 13 Quantitative call parameters measured Count of Filename Avg of duration Avg of Hi_Freq Max of Hi_Freq2 Avg of Low_Freq Avg of bandwidth Max of bandwidth Avg of Freq___max Avg of max_amp Avg of Lo_Fc Avg of D-Fc Avg of slope Max of slope Avg of up_lo__heel_ Avg of uppr_slope Max of uppr_slope Avg of lower_slope Avg of X75_AmpDur Avg of X50_AmpDur Avg of X25_AmpDur Avg of Total_Harmonic_Distortion Avg of X1st_Quarter_Amp Avg of X2nd_Quarter_Amp Avg of X3rd_Quarter_AmpAverage of X4th_Quarter_Amp Avg of X80__Amp_Bandwidth Avg of X50__Amp_Bandwidth Avg of X30__Amp_Bandwidth Avg of F_Center Avg of Mean_Curve_ Avg of Max_Curve_ Avg of Initial_Curve Avg of mid_curve Avg of end_curve Avg of Upper_Curve Avg of Body_Curve Avg of steepest_slope Max of steepest_slope Avg of Char__Slope Avg of Body_start frequency Avg of Body_Dur Avg of Body_Slope Avg of Body_Top_D_slope Avg of End_Slope Avg of SNR Avg of X2H_Max_Amp Avg of T_2H_Max_Amp Avg of Q1_windowed Avg of Q2_windowed Avg of Q3_windowed Avg of Q4_windowed Avg of F_Peak Avg of Min_F_>_-20dB Avg of Max_F_>_-20dB Avg of Band__15dB_ Avg of Band__40dB_ Avg of F_minSlope Slide 14 Example classifier discriminant functions for Myotis leibii vs. M. septentrionalis showing the relative dependence of frequency and amplitude parameters selected as important for discrimination. SonoBat generates high resolution continuous trends of time-frequency and time- amplitude content that enable robust parameter extraction. Inclusion of amplitude parameters increases classification performance above that achieved by using just time- frequency parameters alone. Slide 15 Confidence ranges? Slide 16 Deploying Detectors to Record High Quality Call Sequences Slide 17 DONT DO THIS! Slide 18 DO THIS! Slide 19 Recording with SM2 and SM2+ Detector/Recorders Slide 20 Headphone jack Reset: Used to abort a recording session Power Source Jumper for external or internal power Memory card slots (x 4): read status during wake/exit and remove/add by pushing until click Right (channel 0) and left (channel 1) microphone connections Internal Power from D-cell batteries. AA batteries on right are for internal clock. External power connection Left and right microphone jumpers banks. These are now switches in the new SM2 + (see settings in manual) Use wood screws in corner holes for mounting detector/recorder Wake/Exit Button: used to power unit up or down between recording sessions to check settings and card memory status. With latest firmware can hold wake/exit down to start recording during a scheduled recording session when power was added after scheduled start time. High Pass Filter Jumpers for right and left microphones: set at 1,000 (both to far right) Gain Jumpers for right and left microphone: set at 48 (staggered-24 on top and bottom) Back, Up, Select, and Down buttons: used for programming internal settings. Can press Up and Down buttons at same time outside of a programmed recording session to force a ~45 minute recording session and can press the Back button to stop this recording. During a forced recording session you can listen with headphones and see the level of sound on the display by pressing the Select button External Sensor Connector Slide 21 SM2 Internal Settings Overview for Nightly Recording Schedule Settings 1.Push Select to get to main menu and push it again to select Schedule 2.Push Up or Down to select advanced and push select 3.Program the following six lines of code by pushing select and then using the Up or Down buttons to select the relevant line of code 01 At SSET-00:00:00 (use this to start recording a specified time before/after sunset) 02 DO 03 Record 00:30:00 (you will need to push Select and Up to program the 30) 04 GOTO LINE 03 0X (e.g., you could record for a total of for 2 hours by specifying 3X) 05 UNTSRIS+00:00:00 (use this to start recording a specified time before/after sunrise) 06 GOTO LINE 01 0X (e.g., you could record for a total of for 7 nights by specifying 6X) 4. Push Back to get back to the main menu Slide 22 SM2 Internal Settings Overview for Nightly Recording Utility Settings 1.Push Select to get to main menu and push Down and Select to get to Utilities 2.Select Go to sleep to ensure that the SM2 is turned off and wont use much power and will wake in the middle of a recording session if Wake/Exit is pushed 3.Select Load a config from A: to load a new recording configuration you have programmed in Song Meter Configuration Utility or that you have saved after programming an SM2. 4.Select Save A:SONGMETR.SET to save the current recording settings into a configuration file. 5.Select Erase all cards to do that. However, probably better to just erase data while hooked to your computer after the data has been backed up. 6.Select Set factory default to return to default recording settings. Upgrading Firmware 1.Load the new firmware onto an SD card from the Wildlife Acoustics website http://www.wildlifeacoustics.com/wa-php/downloads.php http://www.wildlifeacoustics.com/wa-php/downloads.php 2.Hold the Up button down as you hit and release the Reset button 3.Your.SM2 file will show up and use the Up/Down and Select to select it 4.Press Reset to reboot and the screen will tell you it is Upgrading and Complete Slide 23 SM2 Internal Settings Overview for Nightly Recording Time and Date Settings 1.Press Select and then Down and Select to select the Settings menu 2.Press Select to select the Time and Date menu 3.Use Select, Up/Down, and Back to set the correct time (set off of a GPS if you can) 4.Select Solar and use the Up/Down to select the Sunrise/set setting Location Settings 1.Press Select and then Down and Select to select the Settings menu 2.Press Down and Select to select the Location menu 3.Select Prefix and use the Up/Down and Select to enter letters in a prefix that will be included on all resulting sound files (exclude spaces using camel hump format) 4.Use back to get to Location menu then Down to Select and enter North Latitude and West Longitude coordinates rounded to the nearest 2 decimal places. 5.Use back to get to Location menu then Down to Select and enter the Timezone. Montana is -7:00 hours from GMT. Sensors Settings 1.Press Select and then Down and Select to select the Settings menu 2.Press Down and Select to select the Sensors menu 3.Use Select, Up/Down, and Back buttons to select sensor settings for an external temperature, anemometer, barometer, or other sensor callibrated by hand. Slide 24 SM2 Internal Settings Overview for Nightly Recording Audio Settings 1.Press Select and then Down and Select to select the Settings menu 2.Press Down and Select to select the Audio menu and use the Up/Down, Select, and Back buttons to enter the following Audio Settings Sample Rate = 192,000 Channels = Mono-R, Mono-L, or Stereo depending on microphones deployed Compression = WAC0 Gain Left and Right = leave at 0.0 dB. These are only set with jumpers Advanced Settings Dig HPF Left and Right = Off if not in use and fs/16 if in use (192/16 12 kHz as minimum frequency recorded) Dig LPF Left and Right = Off since we dont want remove highest frequencies Trg Lvl Left and Right = Off if not in use and 18SNR if in use (SNR = signal to noise ratio) Trg Win Left and Right = 2.0s so that call sequence wont be recorded within 2 seconds of one another Div Ratio = 16 Slide 25 Other Important SM2 Notes 1.If power is not provided to the detector/recorder until after the beginning of a scheduled recording session or if you are making changes in the settings menu during the begging of a scheduled recording session, the detector will not record that night and will instead start recording the next evening. If this happens, you can hold the Wake/Exit down to force recording to initiate. The display will read Backing up and then Recording. 2.If you are moving the detector around during a recording session, unplug the microphone(s) or have the unit Go to Sleep in the Utilities Menu to avoid recording noise artifacts 3.You can force the SM2 to record outside of a scheduled recording session by pressing the Up and Down buttons at same time outside of a programmed recording session. This will force an ~45 minute recording session and you can press the Back button to stop this recording. During a forced recording session you can listen with headphones and see the level of sound on the display by pressing the Select button. 4.If you are recording a bat of known identity for a reference call, record the time of the recording in your field notebook so that you can match it up with the time stamped recording Slide 26 SM3 Internal Settings Overview for Nightly Recording Audio Settings 01HPF1,000Off 02GainAuto (12)Auto 03FSWAC0 04ZCOff8 05Frqmin7,00016,000 06Frqmax128,000192,000 05Frqmin7,00016,000 06Frqmax128,000192,000 07DminOff002.0 08DmaxOff200.0 09Trglvl+18Auto 10Trgwin2.03.0 11TrgmaxOff15 12Repeat 13Record00:30:00 14Repeat 15Record00:30:00 16Untsunrise00:00:00 17UntcountInf Slide 27 Working with SonoBatch Automated ID Output Files Formulas for parsing file names: Location =LEFT(A2,8) Channel =IF((MID(A2,10,1)= "0"),"Right","Left") Year =MID(A2,12,4) Month =MID(A2,16,2) Day =MID(A2,18,2) Hour =TIME(MID(A2,21,2),0,0) Hour_Minute =TIME(MID(A2,21,2),MID(A2,23,2),0) Date =DATE(MID(A2,12,4),MID(A2,16,2),MID(A2,18,2)) Night_Lump =DATE(D2,E2,IF(H2