quantization_tests.cpp shows how to optimize the coefficients of a FIR filter synthesize by fquantizer.
Is there a way to optimize the coefficients of a FIR filter synthesize by another program?
Unfortunately fquantizer fails to synthesize the filter specification I'm working with, however GNU Octave fir2 is able to successfully synthesize the filter. So I have the specification (meaning frequency bands and magnitudes) as well the the floating point coefficients generated by GNU Octave fir2 .. I wish to generate an optimized set of integer coefficients.
Can this be done using fquantizer?
Looking it testLatticeBasedQuantization I see outputD.x is being used ... it's not clear to me how to provide that information.
quantization_tests.cpp shows how to optimize the coefficients of a FIR filter synthesize by fquantizer.
Is there a way to optimize the coefficients of a FIR filter synthesize by another program?
Unfortunately fquantizer fails to synthesize the filter specification I'm working with, however GNU Octave fir2 is able to successfully synthesize the filter. So I have the specification (meaning frequency bands and magnitudes) as well the the floating point coefficients generated by GNU Octave fir2 .. I wish to generate an optimized set of integer coefficients.
Can this be done using fquantizer?
Looking it testLatticeBasedQuantization I see outputD.x is being used ... it's not clear to me how to provide that information.