Ok,
Take a car w/ awesome handling (your choice -'any' car), and send it an analog signal (you turning the steering wheel) to turn a 90 degree angle (and say you didn't flip the car while doing that- heh)...
The sharpest handling car in the world will make a sharp, but still 'curved' right angle.
The driver's 'analog' input can't ever be 'perfect' sending the car a 'perfect' right angle signal.
Now, say you send that car a 'digital' signal telling it to do EXACTLY a perfect 90 degree right angle.
A TRON bike can do it (heh), but a 'real life' car will still only manage the exact same flawed 'curved' turn and in the 'analog' test.
And all for the best right 'cuz a 'perfect' right angle would be very jarring and probably toss you out the side window. heh
Now relate that to a sound system....
Some might say first -"How do you send a car a DIGITAL signal!?!?!?!"
And my answer... well... right... yeah... you can't.
But try to send an analog loudspeaker (your choice which) a digital signal, and again... really... you can't. Right?
There must by definition of 'analog' be a conversion to analog to make a loudspeaker move.
So say you send a loudspeaker a digital 16 bit signal (as unfiltered as possible from whatever DAC and through whatever amp output. Again... your choice)...
Try as it might the loudspeaker can't recreate that 'stair step' looking digital signal that can only be created and measured in the digital domain (like the perfect 90 degree angle a TRON bike that can turn).
Of uncompressed 'digital' format's CD's the worst, so let's look at that one's purest 'stair step-looking' signal...
You aren't using 16bits at all times. How many bits you use is a matter of how loud or quiet the music is right?
If you flattened out a recording so that it was 16bits at all times, wouldn't that sound like a test tone and no longer a musical signal?
If something is recorded quietly then it's using much less bits than if it's loud.
Many CD recordings are so compressed that they don't have full 16 bit resolution. Danny Richie I think posted some info w/ measurments about this subject a while ago. Maybe it was on HDforum though. I'll have to re-look that up.
I've personally seen on my bro-in-law's computer CD's that are pretty compressed using less than 16bits of res., and some aren't normalized either to have anything recorded at -0db which in itself lowers the actual bit level used.
1bit is 6db of dynamic range right? Making CD's 'potential' 96db louder than the noise floor -though lots of recordings have noise floor higher than that -which again... lowers the dynamic range/bit level right?
Now live music can be a lot louder than 96db, so I'm not saying CD is perfect.
DVD-A has the 'potential' to be 144db above the noise floor which is a LOOOOOOOTTTT more bits, BUT....
because of the noise in every actual electronic 'thing' that can output DVD-A, the best the 'best' can do is ~18-19bit (~114db) resolution.
Plus if say an actual recording isn't louder than 96db above it's own noise floor (or is compressed) in this case aren't all those extra bits DVD-A's 24bit resolution has actually filled w/ 'NO' information?
You can also look at the freq. of PCM...
CD is 44.1kHz which can record up to a 'brick wall' of 22kHz.
Some maybe call that 'jarring' to just stop the signal like that, but that 'brick wall' is actually 'filtered' by your ears which are totally rolling off the signal at that freq. (if you can hear it at all). Isn't that relevent?
So DVD-A in 2-chan can record up to 96kHz, and 48kHz in 5.1.
Great, but 'at best'... I can't hear either of those, and 'at worst' ('if' you have a speaker that can output those signals which only 'supertweeter' type speakers can), that inaudible signal can only interact w/ the audible one that you're actually listening to. no?
Some say that you need some headroom over the 22kHz 'brick wall' of CD because something in a DAC/CDP/Processor's output makes the 'brick wall' effect the audible band below it.
I'm not sure what they say on that though?
SACD' DSD is a whole 'nother thing of course.
The editor of S'Phile has measured and said many times now that up to 2kHz DSD is about equal to 24bit PCM, but from there it's resolution lowers, and in fact above 6kHz it becomes 'worse' than 16bit (CD) PCM. At 22kHz it's MUCH lower res. than CD.
He most recently measured this on a $5K universal player that they pretty much called 2nd in ref. quality only to some really high 5 figure system.
He does say that he still thinks both DVD-A and SACD sounds better to him though, but he can't say why, or if it's due to other variables that he hasn't been able to factor out.
I know people rave about the amazing quality of the new Dark Side SACD (which I have), but S'Phile clearly showed that all three versions are different from eachother.
Even the 2-chan SACD layer taken from the same source as the 2-chan CD layer are balanced diff. and the CD layer has ~300+ clipped samples (though it shouldn't 'have' to), and the SACD layer has zero.
Anyway... that's fine if we just talk CD's and ignore DVD-A, SACD, and Vinyl only to keep the issue as low a 'distorion level' as possible. heh (damn, what a 'geek' joke! A certified dork wouldn't forgive that one! heh).
I'm not sure how to wrap this all up and make any totally clear points or conclusions 'cuz I'm really posting this to see if I can learn more on (moron?) the subject or have corrected any flaws in what I'm spouting.
Please if you respond try to keep it as down to earth as possible (though you could say I might not have I guess?) so the most people can understand and benefit from all this crap.
BTW... I'm sorry, but TRON was a dumb movie. heh
