Wav to pulse height histogram converter

Wav2phh is a Qt frontend based audio/wav to pulse height histogram converter. It may be used to postprocess and generate a gamma spectrum from a scintillation counter wav record.

Screenshot

Software and system requirements

Wav2phh can be build for linux and windows. Prerequisites:

• Qt API (Version 5)
• g++ - Gnu Compiler Collection (Linux platform)

Building

• On Linux:
  • Install the packages qt5-qtbase-devel and qt5-qtmultimedia-devel
  • Execute /usr/bin/qmake-qt5 wav2phh.pro
  • Type make
• On Windows:
  • Install Qt 5.6.2 for Windows 32-bit (MinGW 4.9.2, 1.0 GB) qt-opensource-windows-x86-mingw492-5.6.2.exe
  • Building from within qtcreator:
    open wav2phh.pro
    press the "play button"
  • Building from cmd line:
    Set the environment set PATH=C:\Qt\Qt5.6.2\5.6\mingw49_32\bin;C:\Qt\Qt5.6.2\Tools\mingw492_32\bin;C:\Qt\Qt5.6.2\Tools\QtCreator\bin;C:32;C:
    Type qmake.exe -spec win32-g++ "CONFIG+=debug" "CONFIG+=qml_debug" -o Makefile wav2phh.pro
    Type mingw32-make -f Makefile.Debug

Recommendations on sampling rate

Depending on the shaper output a very rough estimation can be made as follows: Assume an input pulse shape u(t)~exp(-(a*t)^2). It's fourier transform equals to U(w)~exp(-(w/2a)^2). U is mainly located within w=-6a..6a (+-3 standard deviations) therefore U has a bandwith of approximately B=6a. To avoid aliasing during sampling the Nyquist Shannon theorem recommends a sampling frequency ws>2B. With Ts=2pi/ws one will get Ts=1/2a and therefore the sampling points at the positions n*Ts are given as u(n*Ts)=exp(-(n/2)^2). u is approximately located within n=-6 .. 6 and therefore minimum 12 sampling points per pulse should be considered for sampling rate calculations. Lower sampling rates are possible if the aliasing error made is allowable.

Notes on shaping hardware

I have obtained good results using a variation of the semi-gaussian filter of second order.

The License

LGPLv2.1+