New DSP FAQ Section - please suggest topics
Started by 7 years ago●30 replies●latest reply 7 years ago●179 viewsHello!
This forum has proven to be full of very knowledgeable DSP people and I would like to abuse your collective brain by asking your help in coming up with a list of topics that could eventually be part of a new FAQ (Frequently Asked Questions) section that would get lots of visibility on DSPRelated.
Here is the plan.
You are invited to use this thread here to suggest one or more topics (one topic per post - please refrain from posting multiple topics in one post) that you think would make sense as part of a 'Frequently Asked Questions' section.
So for example, let's suppose that you would suggest the topic 'What are the differences between the DFT and the FFT' and maybe add a few words in your post to explain why you believe the topic would make sense to address (optional!). At this point, other members could 'thumbs up' your topic if they think it is a good one and should be part of the FAQ section. They could also reply to your post and suggest an alternate way to phrase the question/topic or maybe suggest a completely different way to address the topic in question if they feel strongly about it.
As the list of topics grows and the thumbs-ups accumulate (please use the Thumbs-Up buttons at will!), I will get a sense of which topics you guys feel should be part of the #FAQ section. I then intend to continue to abuse your collective brain and post the suggested questions/topics one at a time as individual threads, maybe one per week or so. If you guys play along and the vision materializes, we will end up with a very informative and useful DSP FAQ section on the site, thanks to your contributions.
Let's see where this goes and thanks so much in advance for your time and input(s)! Do not hesitate to ask questions if you are not clear about what I am trying to do here.
Stephane
why do we need windowing
And when we don't. And what it means.
Thanks a lot Jason for breaking the ice.
I made three posts out of your three suggestions so that people can Thumbs-up the topics individually. Thanks again.
What is the difference between FIR and IIR fiters?
How do I make an FIR filter have gain of 1 at f = 0?
does the bilinear transform really matter? or can I just use Forward or Backward Euler approximation?
I'm not sure I understand what you're getting at here. I view the bilinear xform as a tool. Does a hammer matter?
regards,
Neil
Edited question. The analogy is more like "does it matter if I use a claw hammer? or will any hammer suffice?"
Efficient usage/understanding of DFT/FFT algorithms. It seems to be trivial but once you really understand it then you can use it efficiently. As an example, the Matlab fft function has many underlying tricks, if you don't read the help carefully, you will be lost and the results interpretation is off.
What is zero padding?
You could also include links to blog articles that cover the topic of the question as well.
Ced
Hey Ced,
I like your idea and agree with you that most FAQs are already in the archives. For instance, if I search the comp.dsp archives for 'understanding convolution', one of the first results that comes up is this thread with some pretty good content. Ideally though, the title would be different, the equations would be in Mathjax and maybe some images would be included to help the reader understand. I think we can do better with the new forum interface (not so new anymore), which allows equations, images, youtube videos, thumbs-ups, etc. But of course, my vision won't go very far without the support of users of this forum like you.
I agree that if there are already threads in this forum that could be considered as FAQ, I should mark them as such so we do not address them twice (please feel free to point any).
Thanks for the suggestion Ced.
Stephane
How to get from the math to the real world?
Why is the FFT so important? Can I use it to solve my problem?
Pretty much everything addressed in Sampling: What Nyquist Didn't Say, and What to Do About It. It all came from questions that were asked frequently, so it should fit nicely in a FAQ.
Oops -- just noticed you want separate topics. To follow.
How can people get away with sampling radio signals at less than the carrier frequency?
Why can't I set my anti-aliasing filter cutoff equal to the Nyquist frequency? Why must it be lower? How much lower?
What anti-aliasing filter should I use in my control loop?
If I'm sampling a power line at 60 (or 50) Hertz, why is 120 (or 100) Hertz just about the worst choice of sampling rates?
Why does my anti-aliasing filter make my EKG response look worse? (Because it's ringing!)
IIR filter is unstable filter even though many are using still IIR filter in their applications so How can we use IIR filter to get stable output and linear?
How can I implement filter X in environment Y and does filter X even make sense in environment Y?
How to best get hands-on with DSP?
What is fixed-point arithmetic, why does it matter, why is it awesome to master and how work with it?
What are the caveats of using floating-point arithmetic by default?
(Why?) Should I use an OS on my DSP? (Or why not?)
Should I use a fancy library to build up my DSP application and which library--or should I develop everything from scratch and why (or why not)?
How can I denoise my signal? How can I detect XZY in whatever signal?
Can Tensorflow or other Artificial Intelligence program obviate the need for FFT or any other DSP?