Difference between revisions of "Fast Fourier Transform"

From Hydrogenaudio Knowledgebase
Jump to: navigation, search
 
(3 intermediate revisions by 3 users not shown)
Line 1: Line 1:
{{stub}}
+
'''Fast Fourier transform''' ('''FFT''') is an efficient algorithm for calculating the [[DFT|discrete Fourier transform]] (DFT). The FFT produces the same results as a DFT but it reduces the execution time by hundreds in some cases. Whereas DFT takes an order of <math>O(n^2)\,</math> computations, FFT takes an order of <math>O(n\,\log\,n)</math>, and is definitely the preferred algorithm to be used in all applications in terms of computational complexity. The FFT in most implementations consistent of samples that are exactly a power of 2, this is commonly known as a ''FFT Radix 2'' algorithm where <math> n = 64,128,256,512,1024,2048</math> etc.
  
 +
Unlike the DFT, the frequency spacing nor window size per-bin can't be adjusted because FFT bins aren't calculated individually.
  
Fast Fourier Transform is an efficient algorithm for calculating the discrete fourier transform ([[DFT]]). It reduces the execution time by hundreds in some cases. Whereas [[DFT]] takes an order of <math>O(n^2)\,</math> computations, FFT takes an order of <math>O(n\,\log\,n)</math>, and is definitely the preferred algorithm to used in all applications in terms of computational complexity. The FFT in most implementations consistent of samples that are exactly a power of 2, this is commonly known as a ''FFT Radix 2'' algorithm where <math> n = 64,128,256,512,1024,2048</math> etc. 
+
==External links==
 +
* {{wikipedia|Fast Fourier transform}}
  
 
[[Category:Signal Processing]]
 
[[Category:Signal Processing]]
 +
[[Category:Technical]]

Latest revision as of 13:01, 22 March 2022

Fast Fourier transform (FFT) is an efficient algorithm for calculating the discrete Fourier transform (DFT). The FFT produces the same results as a DFT but it reduces the execution time by hundreds in some cases. Whereas DFT takes an order of O(n^2)\, computations, FFT takes an order of O(n\,\log\,n), and is definitely the preferred algorithm to be used in all applications in terms of computational complexity. The FFT in most implementations consistent of samples that are exactly a power of 2, this is commonly known as a FFT Radix 2 algorithm where  n = 64,128,256,512,1024,2048 etc.

Unlike the DFT, the frequency spacing nor window size per-bin can't be adjusted because FFT bins aren't calculated individually.

External links[edit]