CD Sound Quality -- please help educate me

Saru,
there is a lot of info on Wikipedia & other resources so you can dig thru there & get a lot of info re. the exact procedure of CD mastering. It's a fairly involved process & makes an interesting read. To answer your questions:

1) I have CDs that range from 1990 to now - not a lot as there is a big gap in the middle. But just some older and new CDs. How are CD's generally mastered? I have a Joe Pass CD that indicates it was re-mastered in 24 bit. That seems to imply that CD's are generally mastered with something less than 24 bits.

Mastering with 24-bits means that the bit length during the mastering process was 24-bits. This could have occured one of 2 ways - they recorded the session in 24-bits to begin with & then just kept the width 24-bits thru-out the post-processing OR, they resampled the original data & grew the bit width to 24-bits for the mastering process. The idea of using 24-bits is to ensure that any noise created in the system (due to the short-comings of the electronics &/or the various steps in the mastering process, etc) eats away at the least significant bits, which are bits 22, 23 and 24. The 1st 16 bits are not corrupted with noise.
When the data is finally burnt to the silver disk (to send over to packaging for onward passage to the CD store), the bit width is always reduced to 16-bits. This is because the redbook CD standard calls for maximum bit widths of 16-bits - it simply cannot be any longer. So, when the hack the bit width down to 16-bits the intention is to hack away the noise & retain the 1st 16 bits which are (supposedly) pristine with music & the have the full dynamic range of 16-bits.

2) If CDs generally contain less than 24 bits, then how does a a newer CD player with 24 bit DAC help? Or does a newer player with 24 bit DACs only provide a benefit with CD's that contain 24 bits?

If the CD is only 16-bits & the DAC is 24-bits, then, the original 16-b CD data is re-sampled (upsampled or oversampled) to increase the bit width. No new data is created; rather, the original data is zero stuffed & run thru a proprietary smoothing algorithm to generate a 24-b word that the DAC can use. How good this will sound will depend on the proprietary algorithm because the proprietary algorithm is guessing as to what the interpolated digital data should be. For data that is originally 24-b, there is no guessing. The data is pulled off the source & fed directly into the DAC without any re-sampling. Hence it is very likely that original 24-b data will sound better than 16-b data re-sampled but note that this is *not* always for-sure.

3) What sampling is usually used for CDs when mastering? 44KHz? 96KHz? 192KHz? If CD's are generally created using 44 or 96Hz, then does a player supporting 192KHz help? Does it interpolate between points? Or is 192KHz support only useful for CD's that were created using 192KHz sampling?

redbook CDs use 44.1KHz sampling freq. If you do some research then you'll discover that this freq came about from the raster scan freq used in televisions at that time.
Players supporting 192KHz often support several lower frequencies such as 44.1, 48, 96KHz. So, once the player detects a freq other than 192KHz, it'll lock to that freq & light an LED on the front panel indicating which freq it locked to. Then, that CD is played at that locked freq. What you can also do it command that player to re-sample the original CD to another freq of your choice (sometimes you are forced to upsample to 1 freq which is often the max freq supported by the player).

4) And the bottom line: practically speaking, for a non-musician like me who enjoys listening to music but perhaps has neither the expertise nor critical hearing as most of you, is the sound quality / experience between a current player with 24b/192KHz DAC noticeably and obviously better than my Sony? Or would it be relatively subtle, require concentrated listening, and likely be lost on me? This question probably should keep my system in context.

dude, you are talking "audio" here so "better" is an entirely subjective adjective here! '-). Better for one person is not necessarily better for another person in the same room listening to the same track at the same time. It is not true that a newer 24/192 player is necessarily better than an older 44.1/48KHz player. It all depends on the implementation - an older player could have been implemented well such that it brings much listening pleasure even today.
One very good suggestions by a few members was to use your Sony as a transport & buy an inexpensive, more modern DAC. Instead of spending $200 on a more modern 1-box CDP, spend $200-$400 on a more modern DAC + digital cable. This, of course, only if your Sony offers a digital output. Good sound is often not lost on people who are interested in obtaining better sound; some of these people take longer time to acknowlegde better sound & some take shorter time but such people always end up appreciating better sound because they are looking to progress towards it.

Saru,
yes, you have a lot of questions arising/welling up. The more they get answered, the more they arise... ;-)

@bombaywalla -- much thanks for the synopsis. I'll need to read it more closely. But the quick gist I'm taking away is that the number of bits relates to a refinement of the original sound through noise (hopefully noise only, not actual sound - I don't understand how noise is relegated to the least sig bits) reduction and/or interpolation during playback.

glad that I could help. this is a complex subject that is taught in engineering school in digital signal processing & takes time to assimilate. Suffice it ot say that the noise power in the electronics is small compared to the power of the wanted music signal. Thus, noise power & the power of the signal in the least sig bits is of the same order of magnitude. That's why the least significant bits are corrupted by noise.

The benefit of non-44KHz sampling rates in the player still eludes me unless it's related to the interpolation of points from the data on the CD. But even then, it seems it would not necessarily make the sound 'better' only perhaps slightly different. It's filling in the gaps between data points for what may already be a practically smooth curve.

Putting all this together, it seems like, to me, this is all a matter of subtle refinement intended to better replicate the original sound, not necessarily to make larger, obvious improvements.

Some of your statements are true & others written out of naiveity. Yes, the process of resampling does create a smoothing function which can often merely change the sound without making it better/worse but equally often resampling can make the sound much better. It's hard to say which brand & model will make that impact on you - you'll have to try many of them before you declare success for yourself. ANd, these units are not always cheap.
I believe that the premise of this very high degree of resampling is based on the premise that if you resample at high enough frequency you can make the digital bitstream practically analog i.e. if you sample a signal at an infinite frequency, you'll get an analog waveform.

Taking for example the Cambridge DAC Magic box and their 540/640Cv2: Is the design and components in the DAC much better than what is in whole CDP? Or is it basically the same thing less the transport and the laser unit?

I find that a piece of electronics that is dedicated to performing only one function works better than another piece of electronics that is designed to do more than one function. in a 1-box CDP, one needs to pay attention to the laser drive unit, the associated receiver, the buffering of the DAC, the DAC, post DAC processing, the analog power supplies, the digital power supplies. So, a lot of things in 1 chassis. And, they have to built to a certain price point. There is bound to be a compromise or a major compromise in sonic performance. If you buy a dedicated DAC, then all the effort by the designer has gone into making the best DAC at that price point with no other consideration. You'll get a better DAC for the same money.

are transports significant in CD playback?

oh yes! search the digital archives here on A'gon. Use the words in your question as the key words in the search & you should find a thread from yester years that discussed this very topic. The topic ran into 100s of posts & a heated debate. You can learn much from that thread.
Depending upon how well the Sony is built it might suffice as a transport. DACs that asynchronously resample the date before converting tit to analog can undo some of the detrimental effects of weak transports.