ACELP | ADPCM | AMBE | AMR | AMR-NB | CELP | Codebook | DTMF | G.723 | G.726 | G.729 | GSM | Interpolation | LPC | LSF | LSP | MELP | PCM | Perceptual | Pitch | PSOL | QCELP | Quantization | SMV | VAD | Vocoder

Hello friends, I am working on G.729 Speech standards. I have some doubt on this 1. while doing a windowing, speech frame is multiplied with a 30ms assymetric window. window consist of 2 parts. The first part is half a hamming window and, The second part is a quarter of a cosine function cycle. CAN ANYBODY TELL ME THE REASONING FOR ASSYMITRIC WINDOWING. 2. After finding a autocorrelation parameters , a 60Hz bandwidth expansion is applied by multiplying the autocorrelation coefficient with some exponential functions. CAN ANYBODY TELL ME WHY THE BANDWIDTH EXPANSION IS REQUIERED. Kindly solve my doubts. Thanking you Yours faithfully, tilak email: __________________________________________________

(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )

Tilak and others, The window has to be chosen taking into account that the actual synthesis filter is formed for every subframe by interpolating between the filters estimated in the current and the previous frame. So, the window should be centered closer to the end of the current frame to assure the performance in interpolation. On the other hand, the window must fit into the available data segment comprising the samples until the end of the lookahead. (Lookahead is the portion known from the subsequent frame.) A symmetric window should be very short to meet these two requirements. As a conclusion, an asymmetric window gives the best compromise between the performance and the resulting delay. Bandwidth expansion essentially decreases the sensitivity of the LP filter to quantization errors. This trick brings the poles of the filter slightly towards the origin in the complex plane. Thus the magnitude spectrum of the filter maintains its basic shape, but the sensitivity of the filter decreases. Bandwidth expansion has been discussed in [1]. Hope this helps. Vesa Ruoppila [1] Y. Tohmura, F. Itakura and S. Hashimoto, "Spectral smoothing technique in PARCOR speech analysis-synthesis," IEEE Trans. on Acoustics, Speech, and Signal Processing, pp. 587-596, No. 6, Vol. 26, 1978. ---------- From: tilak shetty [mailto:] Sent: Monday, May 15, 2000 07:53 To: Subject: [speechcoding] some doubts Hello friends, I am working on G.729 Speech standards. I have some doubt on this 1. while doing a windowing, speech frame is multiplied with a 30ms assymetric window. window consist of 2 parts. The first part is half a hamming window and, The second part is a quarter of a cosine function cycle. CAN ANYBODY TELL ME THE REASONING FOR ASSYMITRIC WINDOWING. 2. After finding a autocorrelation parameters , a 60Hz bandwidth expansion is applied by multiplying the autocorrelation coefficient with some exponential functions. CAN ANYBODY TELL ME WHY THE BANDWIDTH EXPANSION IS REQUIERED. Kindly solve my doubts. Thanking you Yours faithfully, tilak email:

______________________________
New Code Sharing Section now Live on DSPRelated.com. Learn about the Reward Program for Contributors here.



(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )