Thank you so much for your explanations and suggestions. It is really a great help and inspiration for me to interact with the experts in the field. I tried doing a similar experiment today but instead of using the earlier vocoding approach (band pass filtering, envelope extraction and modulation) that I had performed, I used the auditory chimera software developed by Shen, Smith, Oxenham, Delgutte (i.e the envelope of speech and fine structure of random noise with same power spectrum as of speech) using various bands (8-24). On analysing the pitch I found relatively lesser mismatch (15-20 Hz) in Fo/pitch of the original and vocoded stimuli from 8-19 bands and almost similar pitch (no mismatch) from 20-24 bands (i.e. within 5-10 Hz), which probably may indicate a technical fault (maybe filtering/normalization) and/or difference (envelope extraction method) in the vocoding that I had performed earlier. The main question still on my mind is how or in what form the pitch gets encoded in a complex temporal envelope of speech.