Can someone clear some doubts on downsampling..
When we downsample say 2:1 .. then we are taking every 2nd sample right?
Whats the effect on the signal....(say on audio)?
Will the signal only contain lower freq components??
if i play an orig audio signal & downsampled audio sig ..what will be the diff ?
when you downsample by "taking every 2nd sample" you get low freq band sound and also highpass band aliased to lowpass band. To get right sound you should use lowpass filtering first (cut-off high pass band). The simplest way is to sum pairs of samples (CIC filter of the order 1). Or half-band filtering in analog will do
losing high frequencies is a relative issue ,for example if you have a speech signal sampled at 44100Hz, then the new sampling frequency will be 22050 ,it means that you still have all frequencies upto 11025 Hz ,and it is more than enough for the speech signal , dont forget to use antialiasing filter with cut frequency fc=Fs/2
Down sampling or Decimation is a method to decrease the sampling frequency. Aliasing might be one of the problems that will creep in if the new sampling rate is less than the Nyquist rate. Effects of aliasing are distortion of the signal upon reconstruction.
also according to theory, after decimation ur sample will be degrated by 4, hence u should amplify the sample by more than 4 to get back the original effect