Disclaimer: I did not attend the lectures in-person so I can't comment on that aspect much. The audio recordings I found to be quite soft or hard to hear as the professor supposedly does not use the on-person mic, so I gave up on trying to listen to the lectures.
Digital signal processing (DSP) extends from the pre-requisite ELEC3004, but focusing on the digital side of things. However, there is a big difference in the way the professors deliver their course. This course has only 4 assignments and a final exam. The assignments do not change over the years, and neither does the final exam much. So if you're looking for a course that basically has the same structure as previous years, Professor Brian Lovell's courses fall under that category.
There is no prescribed textbook and the lecturer states that he thinks you only need the lecture slides. However, without the audio sometimes the lecture slides will not really make sense. This is true, but you also will be using Google quite a bit. Tutors are often posting links to other sites that explain things, on the official Facebook group for the course.
There is also plenty of MATLAB code that the lecturer shows off during the lectures, to illustrate a certain point generally with graphs and figures. This code can be used as a base or template for assignments, which is quite handy. Some of the commands are quite old and have been removed though. However, you should be using it as a base rather than merely editing it anyway.
The course is fairly relaxed in terms of workload. Learning is done through MATLAB assignments and the lectures. You have only one lecture and a prac each week, and that prac is basically the discrete Fourier transform (DFT) in some way. The tutors are excellent and provide you with lots of tips to get started. They go to each person asking how they are going and offer extensive help. The only problem is that it can take a while for the tutor to finally get to you, as they can spend a long time with individuals.
The Facebook page is also quite active, with tutors responding to questions asked.
As a textbook person, you might not like this lack of prescribed textbook though.
There's not much posted on blackboard aside from assignments and lecture slides. It's a bit unorganised with TurnItIn links everywhere, but that's fine given how simple the resource layout is. Announcements are generally only made for assignment releases and extensions (literally every assignment gets extended). All lecture notes are available from the beginning, and they don't tend to vary much throughout the years. Grades tend to get released around 2 weeks after being submitted, but you will get notified about that by tutors on the Facebook group.
ELEC4620 basically is: the DFT and fast Fourier transform (FFT), window functions (Hamming, Hanning, etc, the stuff vaguely touched upon in ELEC3004) and filters. The mathematics involves mainly summation and some matrices. The course presents information that is definitely relevant to the real world and the electrical engineering field. The theory in terms of real world applications become more known towards the later weeks of the semester, where computational efficiency is shown to be an overarching theme in this course.
The key application shown is that of sampling rate and CD players. Although CD players are probably a bit old, they are still used often and are a product of DSP.
There is no major "heading" or "chapters" throughout the course content, you just suddenly stop on one topic and move to the next. In regards to assignment structuring, there is no clear marking criteria for assignments or prose-questions on the exam, though. The assignments are basically reports where you have to justify everything you do in the MATLAB code, relating to the lecture theory. It's not like a straightforward math question problem where you just show your working. So it's a good way of learning the course content as you have to know what you are talking about to get the marks.
I only e-mailed the professor once, but I got a reply fairly quickly. Most of the course contact was related to assignments and was done through tutors instead, via Blackboard. I can't comment on availability for consultation, as I never used it.
Overall, this course presents a fairly light workload compared to other 3rd and 4th year courses. The theory is relatively intuitive and the assignments are straightforward and you are handheld a lot through them. This is in addition to template code being provided. The final exam is mostly multiple choice and the questions don't vary much in the years. The course should be easy to pass, provided you make a decent effort at the assignments (as they are easy marks).
Semester 2 - 2016
Is lecture attendance necessary?
Is the textbook necessary?
Tutors are engaging and helpful (online and offline)
I thought Pr Lovell did a pretty good job at teaching this course and I enjoyed it.
To be upfront I found the content of the lectures is really useful, but quite boring to listen to and I think Brian understands this because he has put in a lot of preparation to create really good demonstrations in matlab code that illustrate concepts or comparisons. He also frequently refers to how he or others have used filters, brings in some interesting applications of DSP both from his work, and from guest lectures.
Brian also has a really fantastic collection of matlab implementations of all the concepts that he posts on blackboard for students to download. This was extremely useful.
The lecture slides are really good, and for the assignments I could get most of the information I needed by referring back to them. The structure of the course is also well done, Brian goes through different digital filters, multirate filters, FFT algorithms and for each topic he explains different methodologies and how to implement them pretty comprehensively.
The assignments for this course are pretty good, each assignment has you implementing the filters you've been learning about with matlab. The tutors were really helpful, and really knowledgeable and go around the room in the tutorial sessions to each student to see how they are going. Additionally they are very active on the course facebook page in responding to student questions and posting course announcements.
The biggest drawback for this course would be that Brian doesn't wear a microphone and wanders around the floor while lecturing, which is fine for if you are attending the lectures, but if you are listening to them on lectopia it goes really soft when he is away from the lectern.