Speech recognition has become an essential and considered as the exciting field in natural language processing, where the accuracy and acceptance of speech recognition has come a long way in the last few years. Speech recognition is the method; machine identifies word or phases in spoken language. Speech recognition focus on phonetics is a branch of linguistics. Phonetics is the study of phoneme or sound system of any language and it has three different aspects such as Articulator phonetics, Acoustic phonetics, and auditory phonetics. This paper presents acoustic modeling algorithms for sound identification along with the results particularly for Kannada, English and Telugu.