- If your desired frequency response is symmetric (i.e., it has conjugate symmetry), the resulting filter coefficients will be real. This is typically the case for linear phase filters.
- If your desired frequency response is not symmetric (i.e., not conjugate symmetric), the IFFT will generally produce complex coefficients. This means your filter will be complex valued.
frequency function to FIR by frequency sampling method
6 次查看(过去 30 天)
显示 更早的评论
i have function with respect to frequency. and i want to the function to FIR filter by using frequency sampling method
i make symetric value using that function.
the function is not linear phase. So when i do ifft, i coefficient is not real but complex coefficient
when i use coefficient = real(ifft(function)), the output is somewhat similar
is that right to use real value of complex coefficient???
0 个评论
回答(1 个)
Samhitha
2025-8-18,12:23
When designing an FIR filter by specifying its frequency response and then taking the IFFT, the nature of your frequency response determines whether your filter coefficients will be real or complex:
Taking the real part of the complex coefficients (using real(ifft(...))) is not recommended. Doing so will change the filter’s characteristics and result in a frequency response that does not match your original specification. The imaginary part of the coefficients contains phase information, and discarding it alters the filter’s behavior.
If you need a real-valued FIR filter (for processing real signals), you should ensure that your desired frequency response is conjugated symmetric before applying the IFFT.
For more details look into following documentation:
Hope this helps!
0 个评论
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Digital and Analog Filters 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!