KiRAT - Kiel Real-time Application Toolkit

If you can only hear the noise during the real-time playback, but not in the recorded wav files, the noise is probably not code related. Try changing the hardware setup or rebooting all devices.

The "logo_objspeechana_hoarseness_mod.ini" inherits all parameters depending on the manipulations. If the parameter "bDefinedModulationTime" is set to false, the manipulation lasts during the whole speech activation time. Otherwise the parameters "fLowLimitInSec" and "fHighLimitInSec" define a time period for a randomized start time in seconds. If the modulation is to be an exact amount of time, "fModPhonationTime" can be set to this value, if the parameter is set to zero, the modulation is until the end. The parameter "fRampingTimeSec" should not be set to zero to avoid artefacts.

• If you cannot hear anything over the headphones, the first step is to observe the output signal in the GUI. Use the plotter in the lower "Visualizations" part of the GUI and select KiRAT Main Signals -> Output Signals. Check the checkboxes of the signals and select the output signals.

• If you cannot see a signal, choose IO Manager -> Signal Mapping Output in the upper "Algorithmic Structures" part of the GUI. Try to find the right channel of your soundcard device until you get an output signal in the plotter.

• You may also have a look in the TotalMix console to find the right channel. If you see a signal but do not hear anything, check if you see level changes in Total Mix in a specific channel that matches your audio input and still cannot hear anything, try checking your hardware by switching to other devices or restarting all hardware devices.

• You can also check to see if Windows is using the correct devices for audio input and output. If not, try switching to the correct ones (your devices should be audio devices!). If this is not possible, the reason could be the USB settings of your sound card: Check if the WDM Devices parameter is set correctly, the number should match the number of inputs/outputs of your sound card.

Unlike direct playback during recording, the volume of the audio tags depends on the Windows settings. It is therefore important to always leave these at 100% volume and to use these settings to parameterize all other volume settings.

All specific settings for the modulation of the signal in terms of hoarseness can be selected in the logo_objspeechana_hoarseness_mod.ini file.

• The piModModeOrder parameter specifies the index for selecting the mode. This also requires a table that describes the properties of the respective modes. Only 20 values are to be entered per bracket []. For more values, the number of additional arrays required must be entered in the iExpandModesby parameter. The total number of values entered here should match the number of tests specified in logo_objspeechana.ini.
• You can also decide whether the modulation should take place over a defined time: If the parameter bDefinedModulationTime is set to false, the signal is modulated for the entire time during which speech is recognized. Otherwise, the time limits fLowLimitInSec and fHighLimitInSec can be used to define a period in which the modification starts randomly. The parameter fModPhonationTime can then be used to select how long the modification is continued. If the parameter is set to 0, the modification is carried out until no more speech is detected and the trial is therefore completed.
• The parameter fRampingTimeSec determines how much time is used to increase the modulation level from zero to the level defined via the code and modulation index. This value should be as small as possible but not less than 5 ms, otherwise artifacts may occur.
• bHeadphoneHint can be used to activate and deactivate whether a reminder should be displayed if headphones have been connected for the test. All other parameters are algorithm-specific and should generally not be changed.