Channel Title: Digital Signal Processing Channel Description: These are video clips of the Digital Signal Processing courses at the School of Engineering / University of Glasgow. I'm Dr Bernd Porr, lecturer in Electronics and Electrical Engineering + Biomedical Engineering / AI at the University of Glasgow. This course covers: - Signal processing with the Fourier Transform (non - causal signal processing) - FIR filters (causal signal processing) - IIR filters (causal signal processing) I use handwritten notes, design/simulate in Python and then provide you with the real implementation in C/C++. Check out github for the C/C++ implementations of FIR and IIR filters. Testimonial: "This was very well taught course. The video lectures were a fantastic idea. The content is very well explained and the clips with scripted explanations spare the time wasted in the classroom on handwaving and mistakes. There is a lot of chances to interact with the lecturer and ask about additional explanations and feedback during the laboratories and tutorials." <video> Adaptive FIR LMS filters: deriving the learning rule Description: With the help of the gradient descent we derive the learning rule how the coefficients are changed so that the average error is minimised. </video> <video> Fixed point IIR filters (Py): Summary (0004) Description: Fixed point IIR filters can be implemented by creating a chain of 2nd order filters which are implemented as Direct Form I filters. </video> <video> Adaptive FIR LMS filters: Introduction Description: Adaptive FIR filters are filters where the coefficients are changing while the filter is operating. It's aim is to minimise the average error between a desired signal and its output. </video> <video> How to get the perfect sound for a zoom session or lecture? 20cm rule! Description: I've compared a few microphones / techniques when recording lectures. Microphones: BPHS1 (pro headset), AKG451 (pro radio moderator mic), PC310 (gaming headset), Clip on mic (as used in lecture theatres) and the dreaded mic in a laptop. The single most important rule is to keep the distance between microphone and speaker below 20cm. Ideally less. Anything else are myths! Happy recording. </video> <video> Sampled signals: Finite length effects of the Discrete Fourier Transform Description: In practical applications the number of samples is finite so you'll have a limited number of samples at hand. This leads to certain limitations which are discussed. </video> <video> Fixed point IIR filters (Py): Calculating the coefficients (0002) Description: The coefficients are calculated with the help of Python high level commands. They all directly generate 2nd order pole/zero pairs so that these can be copy/pasted into the filter C++ class. </video> <video> High level IIR filter design in Python: Introduction (0000) Description: Scipy has a large selection of high level filter commands. </video> <video> Fixed point IIR filters (Py): C++ implementation of the Direct Form I filter (0001) Description: https://github.com/berndporr/iir_fixed_point The C++ implementation of the Direct Form Type I filter is completely kept in a header file so that no library is generated and the C++ compiler can in the best case inline the filter function to speed up calculations. </video> <video> Fixed point IIR filters (Py): C++ filter demo (0003) Description: https://github.com/berndporr/iir_fixed_point The C++ demo simulates a realtime process in reading integer ECG data from a file and then processing it sample by sample by our two biquad filters. </video> <video> Adaptive FIR LMS filters: 50 Hz removal as an example in Python Description: Here, the standard FIR Python filter class is extended by a method which changes the coefficients (= the learning rule). Then 50Hz is removed from an ECG. See the adaptive filter playlist: https://www.youtube.com/watch?v=C0qUQNwP6SY&list=PLxWwb-b9LnpAtV3arjkyl3ltE2sMyTzEl </video> <video> High level IIR filter design in Python: Butterworth filter (0001) Description: The most common IIR filter is the Butterworth filter. </video> <video> IIR high level design commands: Summary (0003) Description: </video> <video> How to record online lectures with zoom (Basic) Description: This howto shows you how to record with zoom by doing a zoom meeting with yourself, enabling screen sharing and then off you go! Bernd Porr. </video> <video> Using the visualiser to create an online lecture with zoom Description: In this clip I describe how I've made the online lectures of this DSP class. I've been using an EPSON ELPDC06 visualiser for this purpose. More info here: http://berndporr.blogspot.com/2018/04/blended-learning.html Bernd Porr </video> <video> Fixed point IIR filters (Py): Introduction 0000 Description: Fixed point IIR filters just use integers for all operations. The coefficients are scaled up so that they use the full dynamic range of the integer coefficients but won't create overflows in the accumulator. </video>

Contacts and Links

Description

These are video clips of the Digital Signal Processing courses at the School of Engineering / University of Glasgow.

I'm Dr Bernd Porr, lecturer in Electronics and Electrical Engineering + Biomedical Engineering / AI at the University of Glasgow.

This course covers:
- Signal processing with the Fourier Transform (non - causal signal processing)
- FIR filters (causal signal processing)
- IIR filters (causal signal processing)

I use handwritten notes, design/simulate in Python and then provide you with the real implementation in C/C++. Check out github for the C/C++ implementations of FIR and IIR filters.

Testimonial:
"This was very well taught course. The video lectures were a fantastic idea. The content is very well explained and the clips with scripted explanations spare the time wasted in the classroom on handwaving and mistakes. There is a lot of chances to interact with the lecturer and ask about additional explanations and feedback during the laboratories and tutorials."

Videos from channel