I have been read lot of about digital signal processing like Z transform, Fourier transform, DFT, FFT, Digital filter I always read but I don't know its real use why I am learning all this teacher teach us only theorem but they don't tell their real use in filterations like voice signal If i want to filter my voice signal what dsp theory will be use I want to ask use of dsp in embedded system
question about digital signal processing
- Thread starter vead
- Start date
Start by deciding how you want to filter a voice signal. For instance, say you want a high-pass filter. Then apply what you're learning to that specific application.
There are many types of signal besides voice that use digital filtering. Just about any analog sensor signal in industrial applications requires filtering before being processed and that can often be done digitally by the same processor that is monitoring the signals.
Pick a simple objective and then play with it.
Get a program (write your own or download a free one) that can pull the samples out of a WAV (or other) music file. Then apply some DSP filters to them and do something with the result, such as plot the spectra or put the output samples back into a WAV file and listen to them.
Some things you might try are to make an equalizer where you can amplify or attentuate different frequency bands by different amounts. Another one would be to record the DTMF tones from a touch-tone phone and design filters to identify the numbers.
Lot's of things you can do to explore the concepts you are learning in class.
Get a program (write your own or download a free one) that can pull the samples out of a WAV (or other) music file. Then apply some DSP filters to them and do something with the result, such as plot the spectra or put the output samples back into a WAV file and listen to them.
Some things you might try are to make an equalizer where you can amplify or attentuate different frequency bands by different amounts. Another one would be to record the DTMF tones from a touch-tone phone and design filters to identify the numbers.
Lot's of things you can do to explore the concepts you are learning in class.
It's easier to filter signals in the spectral domain.
Your question is too broad, there are too many answers, that's why people are being so vague.
Read this: http://dea.brunel.ac.uk/cmsp/Home_Saeed_Vaseghi/Chapter13-Speech%20Processing.pdf
Your question is too broad, there are too many answers, that's why people are being so vague.
Read this: http://dea.brunel.ac.uk/cmsp/Home_Saeed_Vaseghi/Chapter13-Speech%20Processing.pdf
That's somewhat like asking someone to tell you the purpose of screws in the auto industry.
The Fourier and Z transforms let you work in the "frequency domain". You have the same information as in the time domain, but it is presented differently which allows you to do certain things much more easily, both from an analysis standpoint and form a synthesis standpoint.
You need to get a lot more specific for me to given any more specific an answer.
