Rauliruegas...I don't know why you CAN'T have more than 24 bits in a word...the missile guidance computer I used to work with used 32 bit words and much of the data and calculations were "double precision" 64 bit words. But as you say, HOW you use the bits is most important. Our music is "fixed point" data. When you play a quiet section of music through 24 bit digital converters you are actually only using perhaps 8 to 10 of the lowest bits. Not that great resolution. I have recently discovered the importance of keeping the analog signal up before the A/D of my Behringer DEQ2496 equalizer, although this requires me to attenuate the output going to my power amps. A better way to use the available bits, whether 16 or 24, would be "floating point". (A few bits are used to indicate the scaling of the remaining bits of data). This format provides the best resolution for signals of all amplitudes. In the DEQ2496 after the A/D input , a 32 bit floating point digital signal processing (DSP) module performs the equalization and other functions, so we know it can be done. Now we need a disc (or other digital media) that employs floating point.