Edit sound with Python NumPy: Improve code performance 1000x

Increase code performance 1000x times in Python NumPy by managing well big arrays & vectors in a sound editing program

4.80 (17 reviews)
Udemy
platform
English
language
Engineering
category
Edit sound with Python NumPy: Improve code performance 1000x
626
students
7 hours
content
Mar 2023
last update
$74.99
regular price

What you will learn

Code optimization in Python using the NumPy library

Sound processing in Python using the MoviePy library

Fundamentals of digital images

Applying code optimization to binarize digital images

Why take this course?

Programming is one of the most flexible fields I know of. You can create a program that achieves a certain task in so many ways. However, that does not mean that all ways are equal. Some are better than others.

That is especially visible when your program has to work with big data. Working with big data means working with gigantic arrays and matrices.

You can create a program that achieves the same task like the other one, but it does so 1000 times faster. It all depends on how you code and which coding practices you use.

And this is what you will learn here. You will learn the good and the bad coding practices, so that you would learn to code the right way when dealing with big data.

In this 100% project based course, we will use Python, the Numpy and the Moviepy library to create a fully functional sound processing program.

This program will import your videos in sequence, extract their audio, automatically identify the silent intervals in that audio, and then cut them out while still keeping some silence on the edges to preserve a bit of pause in between sentences.

Sound processing naturally deals with millions and millions array elements and so it really matters how we write that program. We will do it in a bad way and in a good way, because I want you to see both sides of the coin.

In the end, you will see that the last version of your Python Numpy code will be more than 1000 times faster than the first version, and so, you will see how to code and how definitely not to code.

Finally, I really want you to see that this knowledge is universal and can be applied in other fields as well, not only audio processing. And therefore, in the last section, there will be an assignment in computer vision.

Digital images are in fact, gigantic matrices, and so, it really matters how you handle them in the code. We will build a small program that can binarize these images and we will also do it in a good and in a bad way.

We will use the Python image processing library called Pillow to process all this big data inside the image matrices.

After this course, you will know how to approach programming in the right way from the beginning. Take a look at some of my free preview videos and if you like what you see, then, ENROLL NOW and let's get started! I'll see you inside.

Screenshots

Edit sound with Python NumPy: Improve code performance 1000x - Screenshot_01Edit sound with Python NumPy: Improve code performance 1000x - Screenshot_02Edit sound with Python NumPy: Improve code performance 1000x - Screenshot_03Edit sound with Python NumPy: Improve code performance 1000x - Screenshot_04

Reviews

Jay
January 1, 2023
The course does what it sets out to do. It demonstrates perhaps the most important property of the numpy library -- vectorization of array and matrix operations. This is used in the creation of a python program that processes mp4 files (using the moviepy library) to remove silent portions. You learn a fair amount about processing mp4 movie files using python and the moviepy library. There is also a section which covers the python PIL library for processing images and shows how to create a custom image conversion program that creates a stylized black-and-white image from any input JPG or PNG image. The program that is developed works well and performs well. I think those things make the course worth what you pay for it. The issue that I have is that the way the python programs are written. In this course, they are written in a monolithic way with no function definitions and no class definitions. This is a demonstration of bad programming practice. Employer would be unhappy with code like this -- it is very difficult and expensive to modify and maintain over time.
Yaser
November 16, 2022
I would like to thank Mark here too for this nice course! I learned very good things for my future work and how I can save a lot of time . to be honest, I wasn't very happy at the start because I knew the most things and I saw directly a lot of things which I could have optimized by myself, BUT: even with all my own optimization i would have save maybe 10x of the time, but not 1000x of the time in the sections we can control like Mark did. So I still learned new and importand things from this course and this will definitly improve my programing and engineering life. I am happy I enrolled for this course best regards Yaser
Angel
February 8, 2022
Good course for beginners: It talks about a great concept, vectorization. Very practical course that develop a usefull tool and Mark it's a great instructor, clear and structed expanations. If you area experienced developer used to numpy that course maybe it's not for you. Resume: for beginners 5 stars: talk about a important concept, practical, concise, good duration and structure. For advanced 3 stars, but as I think that it's for beginners, so 5 stars.

Charts

Price

Edit sound with Python NumPy: Improve code performance 1000x - Price chart

Rating

Edit sound with Python NumPy: Improve code performance 1000x - Ratings chart

Enrollment distribution

Edit sound with Python NumPy: Improve code performance 1000x - Distribution chart

Related Topics

3808084
udemy ID
1/28/2021
course created date
6/27/2021
course indexed date
Bot
course submited by