Quantizers: tQuantOpt - File Exchange

Code covered by the BSD License

Highlights from
Quantizers

PDFFn (PDFType, Arg1, Arg2) Return pointers to functions to calculate functions of the PDF
Quant (x, Xq, QType) Quantize a vector of samples
QuantALawTables() This routine returns quantization tables for a 256 level segmented A-law
QuantEntropy (Xq, Farea) Calculate the entropy for a quantizer
QuantLloyd (Nlev, FPDF, QSym) Iterate to find the output levels for a minimum mean square error
QuantMSE (Yq, Xq, FPDF) Calculate the mean-square quantization error
QuantMuLawTables() This routine returns quantization tables for a 256 level segmented mu-law
QuantOpt (Nlev, FPDF, Sym) Find a non-uniform minimum mean square error quantizer.
QuantRefine (Yq, FPDF, QSym) Iterate to find the output levels for a MMSE quantizer.
QuantSNR (Yq, Xq, FPDF, Scale... Calculate the SNR (dB) for a quantizer defined by a table for a
QuantUnif (Nlev, FPDF, Sym) Find a uniform minimum mean square error quantizer.
TestPDF Use two different methods to calculate each of the following
p=lin2pcma(x,m,s)
tLogQuant (enc) Test Quant for A or mu-law quantizer tables
tQuantOpt
tQuantUnif
x=pcma2lin(p,m,s)
View all files

from Quantizers by Peter Kabal
Routines to design/evaluate MMSE scalar quantizers, and an efficient quantizer routine.

tQuantOpt

function tQuantOpt

FPDF = PDFFn('gamma');

Sym = true;
%Sym = false;
NlevA = [2 4 8 16 32];
for (Nlev = NlevA)
  [Yq, Xq, MSE, Entropy, SNRdB] = QuantOpt(Nlev, FPDF, Sym);
%  Xq
%  Yq
  fprintf('Nlev:%4d, MSE = %.3g, SNR =%6.2f dB, Entropy =%4.1f bits\n', ...
          Nlev, MSE, SNRdB, Entropy); 
end

return

Highlights from Quantizers

Highlights from
Quantizers