Digital Signal Processing :(

[Emc2]

Ok, first off, this DSP First book really sucks. It's so cryptic, and you can tell the guys who wrote it are very egotistic. Needless to say, 95% of the class is completely lost in this subject, and we're all failing the course.

I want to understand DSP. We're all endeavering to become computer or electrical engineers, and DSP is the first major class we have come to (sophomore year).

I need an online tutorial. The prof. doesn't explain any of my questions in plain english (probably b/c English is his third language). The book is ssooooo lame.

Where can I find a place to learn this stuff from? I'm so confused. I sort of know how to do the math in it, but have no idea what I'm doing...

[gino X]

have you tried Googling for an answer yet?

[pbharris]

hello,
DSP is not a real easy subjest to understand, and there are many variations of it. i came quite close to dropping it when i had it (it was called digital filters) but am glad i stuck it out. a fellow at work ahs his PHd in DSP, i can try to ask him some particular questions of you like. i personally only took the class while i was in school and don't do any DSP stuff now (well, not in the algorythm). mosts DSP that i encounter are hardware based, they have MACs that can do 24 bit operations really quick.

i guess if i had one thing to say is learn how to do an fft and you will be way ahead. i have not seen on heavy duty calcualtion not use an fft or something similar (dct dwt) , image processing, audio processing even statistcal crap. here is decent start.
http://www.intersrv.com/~dcross/fft.html#source

[Xeroid]

The prof. doesn't explain any of my questions in plain english (probably b/c English is his third language).

I feel your pain dude. I had an English professor back when I was in college who was from Lithuania. It was the only time during my college career that I received a C for a grade. This guy was trying to teach us English and no one could understand him.

Mike

[DanU]

Originally posted by Emc2
The prof. doesn't explain any of my questions in plain english (probably b/c English is his third language).

Lol. My DSP prof was named Chandrasekaran or something like that. You could tell he was a sharp guy, but i could hardly understand a word he was saying.

[Emc2]

We've sorta skipped over the Fourier Series until the end of the semester...We doing FIR (Finite Impulse Response) filters right now, and IIR (Infinite Impulse Response) filters are right around the corner.

I just learned more about signaling in physics....You know, Physics 2 should be a PRErequisite, not a COrequisite...Phasor addition is so much easier if you understand what it's trying to do...the book just said "take e^j*pi/2 and convert it to cartesian, then add it to the cartesian form of blah blah blah"...All we needed to know was to treat the phasors like vectors and add them together......The prof. was no help either...

A google search brought up Rice University as the first result...I could look up the prof's that teach it there and then go to their webpages to see if they have any notes or anything...

[Emc2]

http://www.dsptutor.freeuk.com/index.htm

found this site, and it has a few links to another site with some intro. info...

I was hoping to see more examples of the math more than the concept...like, real-life examples...I can't connect the concepts with the math in these filters...

[screen]

There are lots of lectures that are related to DSP at this on-line publications site of a Stanford professor.

http://www-ccrma.stanford.edu/~jos/pubs.html

The most interesting and fundamental topics are
- Mathematics of the Discrete Fourier Transform (DFT)
- Introduction to Digital Filters

The first topic clearly tells why "e^j*pi/2" is used in DSP?, what does it mean?, etc. Ofcourse, these topics expect you to know the mathematical fundamentals like "Taylor series", "complex number theory", differentiation, integration, etc.

I hope these on-line publications help you to understand DSP.

One thing I can tell u about DSP is that once you understand the fundamental concepts about signals, complex exponentials, sampling theory, DFT and DTFT, etc., you are all set. FFT comes very late into the picture (it is just clever mathematical manipulation of DFT to make computation faster).

-Screen

[Emc2]

I understand how to do Taylor series, that 2cos(2pi*f*t+pi/2)=2(e^jpi/2)(e^j2pi*f*t), and Calc. II is a prerequisit for this course...

Thanks a bunch for that site. No I don't know why e^j*pi/2 is used, but I do know what it means, now that I've gone through the physics of it...current lagging voltage in an inducter and leading voltage through a capacitor in an AC circuit, and thru the 'phasor addition', aka 'vector addition', You can find impedance, total resistance, etc....IMHO, this was all necessary to know before even attempting this course. I'd have much rather been put on the 5 year plan than be this lost.

[screen]

Those articles take you through the process of understanding the basis for transformation from time domain to frequency domain using the Discrete Fourier Transform. The complex exponential is the key link here because e^j*pi/2 is a complex function of sinusoids, where sinusoids are characterized by the frequency, blah, blah, blah......