Corrected few typos and one major mistake in Discrete Cosine Transform topic.
Updated article is on the same place:
Published first draft with detailed explanation on LPC. Was choosen the shortest way to show why Levinson-Durbin Recursion is right.
See it here:
The corrected mistake could lead to computing only half coofecients on each recursion step. The result could be all the vector computed wrong.
The document was checked more carefully and now direct implementation is computing right values.
See corected document at
TODO: I must check it again.
Described FCT and FST algorithmes of Radix 2 with usage of only "Decimation In Time" approach.
The document is here:
Two issues are resolved:
1. Strict prove validity of Discrete Fourier Transform.
2. Some of ambiquity in Fourier Integral prove is repaired.
All the set must be yet revised for mistakes and clarifications.
Proved Orthogonality of Complex Continuous Functions.
The document is updated and new files are available.
Corrected several typing and logic errors in math documents.
Document with detailed motivation for Complex Multidimentional vectors Orthogonality to be valid.
Open Office: http://arkadianalysis.sourceforge.net/Orthogonality.sxw
Conception "Set of Pairwise Orthogonal functions" is used as it must be.
In previous version two sets of functions where used. It is mathematically right, however it will stop farther development of the subject.
Available new document with detailed prove of Fourier Transform.
The original is writen with Open Office: http://arkadianalysis.sourceforge.net/Fourier.sxw
Else PDF version is available:
And HTML version: