I am reading the LAME's fast mdct implement (newmdct.c). I know the mdct was unrolled by hand here and using symmetry of the trig factors involved in the calculation , but I can't understand its details.... I can't combine the mdct theory with this implement in the codes .

Anyone who have known all the details of this codes can help me ? Maybe it's not
suitable for here ,but I can't find any valuable info after a long time googling ....

Thanks ahead.



--------------------------------
Everything is a filter ......