Holy smokes, hitman, what the hell are you talking about?
Bit depth and cutting & pasting and crossfading are all completely seperate topics and barely related.
A soundcard is really little more than a package of analog to digital and digital to analog converters. When your analog signal hits the card, the card converts the signal into its digital representation, and from there of course it can be saved to disk, edited, etc.
Likewise, when you hit the play button in your recording software, a data stream representing your sound is sent to the soundcard, and the soundcard converts the digital signal into an analog signal that you can eventually listen to.
The conversion from analog to digital is especially important to the sound you get when you're recording. An analog (real life) sound is represented by a *continuous* waveform, which is impossible to represent using our digital 1's and 0's. So what happens is that the analog signal is sampled a certain number of times per second. Each sample, comprised of a certain number of bits, represents the amplitude (height, instantaneous voltage) of the analog wave at that specific moment. Now you take enough of these samples per second, and use enough bits to represent an ample number of amplitude fluctuations, and you've got something that approximates the original analog waveform pretty closesly.
"CD Quality" means 16bit, 44.1khz digital audio. That means that 44,100 16bit samples are taken every second. 16bits is enough to represent 2^16= 65,563 amplitude fluctuations. That's pretty good, and *technically* provides 96db worth of room to work with, at frequencies between 1 and 22khz. (maximum frequency = 1/2 the sample rate: search Nyquist)
Now most prosumer & professional soundcards on the moment work with 24bit audio, at sample rates up to 96khz. Why would you need more than CD Quality? Well, the human ear is a lot more sensitive than people give it credit for one thing. Also, using 24 bits pushes the noise floor down and you get a theoretical headroom of over 120db. Plus when it comes to Digital Signal Processing (DSP), like digital reverbs, compressors, etc...the more bits & samples you're working with, the better your end fidelity (in theory, e.g. more precise mathematics), even though you eventually downsample & dither to 16/44 for your final product.
Now, don't get too caught up in the numbers game. Know this: IT IS POSSIBLE TO HAVE 16bit/44.1Khz CONVERTERS THAT SOUND BETTER THAN 24bit/96Khz CONVERTERS!
Why? Building converters is extremely difficult. Consider what's going on. A little clock is ticking away, and 44 thousand times per second, it's telling the system to sample the incoming audio voltage. Not only is it critical that it measure the amplitude preciesly, but it's equally critical that it take these samples at very precise intervals. Think about a simple sine wave being cut up into samples by a converter...what happens if the samples it takes along that wave don't occur at precise intervals? According to the resulting data file, you just have 44 thousand samples that were supposedly taken in 1/44,100th of a second intervals...but they weren't, were they! What you end up with in this case is not a nice smooth sine wave, but a jagged representation of a sine wave. Now this jagged wave hits and equally jittery digital to analog converter, and the sound you end up hearing is recognizable, but pretty shitty compared to what went in!
This is why the guy who says "this CD player sounds better than that CD player" isn't crazy. The "industry" really got off on telling everybody that anything digital sounded great, and that every digital copy is identical to the original. They were right, technically, and they just sort of skipped over the ugly stuff.
Now consider multitrack recording. If you're sticking 24 jittery shit sounding tracks on top of eachother, the sound degredation really starts to add up.
So, in the end, asking "what's the best sounding soundcard" is just as valid as asking "what's the best sounding 1/2" tape machine?"
There is a stupildy easy way to test a converter. Do you have a POD? Run your POD into your soundcard, and monitor directly from the soundcard. Now, what you're monitoring is the direct, undigitized signal (if the signal was digitized, you'd have a very noticable delay). Take a good listen to how the guitar sounds and play around for a bit. Now hit record. Theoretically, the resulting recording should sound pretty much like what you just heard, right? Right. But it won't. Congratulations, you just heard your converters!!!!! This is always an ear opener. (btw, you don't need a POD, I'm sure you can improvise something...run a CD of a a solo guitar into the soundcard and monitor it for a while - or something simple and detailed - and be critical here, yes of course it will still sound like a guitar)
Slackmaster 2000
