Outline
"Pitch Flattener" is a very simple pitch collection tool. "Pitch Flattener" is specialized for collecting pitch of sounds whose frequency changes with linearlity. Examples of the sounds are EMU motornoises on accelerating/decelerating and engine sounds.
License
"Pitch Flattener" includes the following third-party softwares.
| Software Name | License |
|---|---|
| Math.NET Numerics | Math.NET Numerics License (MIT/X11) |
| Reactive Extensions | MICROSOFT SOFTWARE LICENSE TERMS |
| Visual Studio 2013 Image Library* | MICROSOFT SOFTWARE LICENSE TERMS |
*Installer version only
Install
Pitch Flattener runs on Windows PC which is installed "Microsoft .NET Framework Version 3.5". If the app failed to launch, please try to install the runtime.
If you have downloaded the installer version, extract it and then launch the extracted installer. Otherwise, extract it and then move the extracted folder "Pitch Flattener" to the place where you want.
Uninstall
If you have installed the app using installer, uninstall from "Settings" or "Control Panel". Otherwise, just delete the app folder "Pitch Flattener".
Basic Usages
Load an Audio File
Open an audio file from the menu "File > Open" or file drag and drop. Acceptable audio format is only WAV, which bit depth is 8bit/16bit/24bit/32bit PCM or 32bit/64bit IEEE float and channel is stereo or monaural. There is no no restriction on sampling rate.
Set a Pitch Collection Parameter
In "Pitch Flattener", only one parameter "Octave diffrence" is used for pitch collection. It represents pitch change of the audio in octave. "Pitch Flattener" performs pitch correction on the presupposition that frequency of a sound changes linearly.
This parameter can be configured by changing "Ocrave diff." on the app main window. If pitch of the audio increases(e.g. EMU accelerating motornoise), set positive value there. Otherwise, set negative value. The green curve represents a pitch envelope curve.
Listen Pitch Collected Audio
Click "Listen" button, then starts a pitch collection process. After that, the pitch collected audio playback begins with alternate loop. "Alternate loop" means repeating forward playback and backward playback. This loop method make hear a sliced short continuous audio more smoothly. "Stop" button click stops the playback. Until the pitch is collected completely, repeat "Octave diffrence" value change and "Listen".
Spectrogram for Helping Pitch Collection
You can use spectrogram view from "View > Spectrogram" in the menu. The spectrogram view will help "Octave diffrence" value adjustment.
Save Pitch Collected Audio
You can save the pitch collected audio from the menu "File > Save" or "Save" button. The file format is fixed to "16bit-PCM stereo WAVE".
Checking the option "Lanczos" is recommended at an audio save. It improves pitch collection quality. However, "Lanczos" method is much heavy, so checking "Linear" is recommended during "Octave diffrence" value adjustment.
Advanced Usages
Spectrogram View
Using a spectrogram, you can visually verify whether the audio pitch is collected or not. Click "View > Spectrogram" in the menu to open the spectrogram window. To make the verification easier, the spectrogram visualizes alternate-looped audio waveform. The spectrogram is updated both after opening a audio file and after a pitch collection process.
Currently, the implement of the spectrogram in "Pitch Flattener" is not good, therefore generating spectrogram is a heavy process. Please don't open a large audio file by "Pitch Flattener".
The upper status bar indicates some information: time of the audio, frequency, and note name pointed by the mouse cursor, and FFT options. The lower indicates some instructions.
The spectrogram view is customized by changing options at the top of the window. "FFT" button click shows the FFT option configure window. "Col." button click shows the spectrogram color configure window. "Initialize" button click restore the defaults.
The other options are listed below.
Channel
L
If selected, left channel audio data is used to display.
R
If selected, right channel audio data is used to display.
Spectrogram View Options
Brght
Represents brightness of the spectrogram. The spectrogram gets brighter as the value increases.
Range
Represents dynamic range of the spectrogram. Contrast of the spectrogram gets lower as the value increases.
Gamm
Represents gamma collection value applied to the spectrogram.
Zoom
Zoom setting of the spectrogram.
H
Represents horizontal axis zoom rate.
V
Represents vertical axis zoom rate.
FFT Options
"FFT" button in the spectrogram window click shows this configuration window.
FFT size
Represents the number of wave data applied to a FFT process. The larger this value gets, the higher the frequency axis resolution gets and the lower the time axis resolution gets.
Overlap
Represents data overlapping degree. The larger this value gets, the higher the time axis resolution gets.
Window function
Represents a window function which is applied to audio waveform data before FFT.
Spectrogram Color
"Col." button in the spectrogram window click shows this configuration window.
Dark
Set a color to the dark part of the gradation.
Light
Set a color to the light part of the gradation.
Copy from L/Copy from R
Copy the color setting from the other channel.
Advanced Save Options
Create alternate loop
If checked, alternate-looped waveform is used for save.
Add params
If this and "Auto file naming" are checked, pitch collection parameter "Octave difference" is added to at the end of the filename for save.
e.g. "audio.wav" > "audio_flat_1-50.wav"
Toggle
Toggles states of "Create alternate loop" and "Add params" checkbox.
Disable overwrite warning
If this and "Auto file naming" are checked, file overwriting warning dialog is not shown.
Auto file naming
If checked, no filename setting dialog is shown on save. Instead, the filename for save is created automatically by adding suffix "_flat" to the original.
e.g. "audio.wav" > "audio_flat.wav"
Interpolation Method
Linear
If selected, pitch collection process done in a short time but the quality is not good. It is recommended to use this option during the parameter adjusting.
Lanczos
If selected, you can get high quality pitch collected audio, but the method takes much more time than "Linear". It is recommended to use this option only on audio file save.
Quality
Represents quality of the Lanczos method. The larger this value gets, the higher the result audio quality gets.
Remarks
High Quality Converting
Actually, "Lanczos" method in "Pitch Flattener" provides better quality but not the best. If you want to get the result in much better quality, please try to use the procedures listed below.
- Upsample an audio by another audio wave modification app
- Collect pitch of the audio using "Pitch Flattener" with "Linear" method
- Downsample the audio if necessary