# Convolution Reverb

## Abstract
* **Convolution Reverb---Convolution Reverb** The relevant Plug-ins on the market are also called **IR Reverb (Impluse Response Reverb)**. This algorithm uses **Fast Fourier transform Fast Fourier Transform** obtains the audio frequency, and then uses the mathematical operation method of **Convolution** to obtain the frequency distribution of Audio in the IR, and write the result of the calculation.
---
## Usage
The following figure shows the GUI of this algorithm:
* Step 1: **Enter Dry Audio path**
Enter the path of the Wav file to be processed in the Dry Audio field.
* Step 2: **Enter IR path**
Enter the IR Wav file path in the Impulse Response field.
* Step 3: **Press Deliver**
The file path will be passed into the algorithm, and *Please Put Audio File Path In Entry* will become *Please Enter the Process*.
* Step 4: **Press Processor**
At this time, the algorithm will process the audio. This step will take a little longer. When the GUI can operate normally, it will be completed, and *Audio have been Processed* will appear.
* Step 5: **Press Play to play the audio, Dry/Wet Slider to adjust the dry humidity**
Press Play to hear the processed audio, and you can adjust Dry/Wet to adjust the tone you want.
* Step 6: **Output Audio**
After the adjustment, you can output it as a Wav file. Seeing *Wav_File Output* indicates that the output is complete. **The output file will be based on the current position of the Slider. Please be sure to check it clearly**.
* Remarks:
* Dry.wav is unprocessed
* Wet.wav is calculated
* Final.wav is adjusted by Dry/Wet

---
## Technique
### Development tools
* Python Language
* PyCharm
* wave
* struct
* matplotlib.pyplot
* scipy.signal
* tkinter
* pygame
* pydub
#### scipy.signal.convolve

convolve(sample[0] * gain_dry, reverb[0] * gain_wet, method='fft')  

This line of code in the algorithm is the main one, **Fast Fourier Transform FFT**&**Convolution Revolution** is completed here, which is why you need to wait when you press Process.
#### Impulse Response
IR refers to the response of a piece of audio in a space. Usually, the method of recording IR is to puncture balloons or items that can make a loud instantaneous noise. Record the audio and record the response of the sound.
#### Parallel processing
The control of Dry/Wet Slider is to use the idea of parallel compression Parallel Compression on the mix, adding the signal of Dry to the signal of Wet, and then outputting it.

---
## Summarize
In the code of Convolution_Reverb.py, there are labels for detailed purposes. Those who don’t understand can read the labels at the back to help understand the algorithm process. There are many areas where this algorithm can be optimized:
* Increase the calculation speed. This time I use the scipy.signal.convolve function. In the future, you can write a Fourier transformation and convolution process by yourself.
* Add function, you can add Low High pass Filiter or Gain, Pre-Delay and other functions
* Plug-in, you can use MSP programming, written as Plug-in, this is the skill I am learning, and the content of Plug-in will be placed on the website in the future.