CDP Question

[tim_j_thomas]

First, I'm relatively new to audio -- so please excuse me if this question is "out of line."

But I don't understand why someone would want to pay thousands of dollars for a cd player. For example, if you have $100 cdp with a digital out, why would you pay hundreds or thousands more for another cdp with a digital out.

By definition, neither can sound better than the other as all they are "outputing" are 1's and 0's.

Am I missing something.

Thanks in advance, I hope I can learn something with this post.

 

[jimholt]

A lot of factors here the digital out is for an external dac to improve sound further. The internal dac and power supply would have better components. Jitter comes in there too the better the clock less jitter. Case construction more solid less external vibration. The rest on here will come up with more info no doubt. You get what you pay for is the old adage. Jim.

 

[lini]

For one part, you should probably read up a bit on jitter, which describes the range of deviation of the digital signal timing from the ideal. One doesn't have to pay thousands of bucks for a player with very low (100 -150 ps) jitter, though. There are sub-$1000-players that can do that, too. For more money, you can exspect tank-like build quality, expensive high quality parts, oversized power supplies (usually separate ones for the digital and analogue sections), maybe some nifty extra features like several digital filter algorithms to choose from, switchable upsampling options et cetera.

Whether that's all worth it, is another question, though. Let your ears and wallet decide, I'd say. For myself, I'd rather mod the hell out of some decent sub-$1000-player - however, I might think different, if I was rich enough to buy something like an Accuphase DP-67 with my pocket money.

Greetings from Hannover!

Manfred / lini

 

[Garbz]

search for Jitter on this board. There have been some very detailed discussions of exactly how and why cdplayers sound different, what companies have tried to do to combat it, and why most of the technologies today simply can not eliminate it.

 

[tim_j_thomas]

Perhaps, this is getting off topic, but it was my understanding that jitter happens where there is a problem with the oscillator on the device accurately generating the correct time frequency -- and that this is an incredibly rare event and that the use of ram buffer with a local oscillator can clock out the data in a reliable manner.

Seems to me that just about any cd player over $50 would implement this.

Still trying to learn and figure all of this out.

 

[Garbz]

Jitter is actually a few things. Ocillators not being accurate, but rather it is measured as the difference in time the data arrives and the time the clock beats. It can be induced by anything. Poor layout in a large FPGA chip, poor layout of the data, most predominatly the digital S/PDIF interface induces a LOT of jitter as a result of combining the data and clock into the same signal, sending it down the line with incorrect termination and then recovering the clock at the other end.

So now you have a poor clock useless for reclocking. Your buffer option will not work. No two clocks beat at exactly the same rate. Pretty soon your buffer will either fill or underrun, and what then. Enter Asyncronous Sample Rate Conversion. Some say its good, some say its bad. My little admittedly poorly implimented experiment on my DAC made me pull it right out again. There are graphs I have seen to show it attenuates jitter, and graphs to show the negative inpact of now having 2 sources of jitter in the same datastream.

In the realm of S/PDIF there is no workable solution to this problem, and bear in mind that we are talking about trying to get the jitter down below 10s of picoseconds. Many $50 players will be in the order of 1000 times worse then that

 

[tim_j_thomas]

Thanks for the info. Can the human ear really detect errors at that small of a magnitude?

 

[replytoken]

Backing up from the whole issue of jitter, there is a more basic issue of digital to analog conversion. Yes the data is 1's and 0's. But they are converted to analog, and there (the conversion process) is one simple area to understand why CDP's can sound different. How much different CDP's can sound is up for discussion, but the conversion process is often a good place to start the discussion. Save jitter for when you really want to chew on something esoteric!

 

[lini]

replytoken: Hmmm - I don't think the whole jitter affair is so hard to grasp on the logical side, actually. If you imagine PCM conversion as hitting points (or probably more exactly: lines) within a very exact grid as exactly as possible, it's pretty clear that both not hitting the exact level and not keeping the exact intervals is detrimental and will produce artifacts/alter the result.

I guess, what makes matters so complicated is that there are so many influences on sound in a cd player - power supply and its filtering & buffering, the transport, the dac section, analogue output stage, chassis construction & isolation, possibly some extra dsp functions for upsampling, different digital filter algorithms and so on. So, when one compares different cdps, there usually is a whole bunch of variables, of which jitter is only one (and, even worse, one that, again, is influenced by a number of variables...). Less people, as diy modders or while visiting a modder's workshop, e.g. ever had the opportunity to compare two identical players only differing in the clock and thus could say "The higher jitter of the stock clock made the sound more nervous and the imaging more fuzzy.", i.e. had the opportunity to actually isolate jitter effects from other variables.

Greetings from Hannover!

Manfred / lini

 

[replytoken]

Quote:

Originally Posted by lini replytoken: Hmmm - I don't think the whole jitter affair is so hard to grasp on the logical side, actually.

You are correct, its not that hard on the logical side. But as you further state, there are many other factors to consider when understanding jitter. And that is where it can be a bit much, especially if your starting point is that all data is composed of 1's and 0's, and therfore sources that read this data should most likely sound the same. I personally like to start at an overview level and work my way down to the details, like jitter. I am guessing, but I believe that if a person does not generally accept variances in sound reproduction involving the DAC process, then they are probably not going to have much faith in the impact of jitter on sound reproduction. And, my comments above are directed to a general audiance, and not specifically to the OP. Of course everybody's mileage may vary!

 

[Thaddy]

Here is a good write-up that describes some differences between cheap $100 players and more expensive multi-thousand dollar players. It centers around jitter.

 

[PhilS]

Quote:

Originally Posted by replytoken Backing up from the whole issue of jitter, there is a more basic issue of digital to analog conversion. Yes the data is 1's and 0's. But they are converted to analog, and there (the conversion process) is one simple area to understand why CDP's can sound different. How much different CDP's can sound is up for discussion, but the conversion process is often a good place to start the discussion. Save jitter for when you really want to chew on something esoteric!

What he said. We don't listen to 1's and 0's. We listen to music, which is analogue. How the $50 player gets the 1's and 0's off the CD and translates it to the music we hear, and how the $2000 player does it, can make all the difference in the world.

 

[Awk.Pine]

The OP asked about using a cheap CD player as a transport -- so the DAC argument doesn't apply.

As far as why someone would want to (or, more accurately, be willing to) spend thousands of dollars: go have a listen. I don't care how it does it, all I know is that they sound different. (Admittedly, I am curious and enjoy learning about it, but the question isn't "how do they sound different" or "why do they sound different" but "do they sound different?" The answer is, yes, they sound different. And that difference is what people pay for.)

I'm sure there's a hi-end stereo shop in the Detroit area. Do not go into a store that sells appliances or cell phones (because they won't have the good stuff). Try to avoid stores that sell home theater equipment (because, in my experience, they specialize in distancing people from their money). Find a store that sells electronics that you think are priced absurdly (because, in my experience, people at those shops love the equipment they sell and would enjoy teaching you about it). Or find a hi-fi group in your area. Ask them to show you the beauty of a top-notch system. And listen. That's the gravy, in listening.

Best of luck!

 

[lini]

Quote:

Originally Posted by replytoken (...) And that is where it can be a bit much, especially if your starting point is that all data is composed of 1's and 0's, and therfore sources that read this data should most likely sound the same.

Well, maybe that's a flaw to begin with - i.e. to start one's thinking at the point of logical abstraction. It's not fine clean digital 0's and 1's where it starts in practice - it's more or less precisely pressed pits and lands on cd, which are read out by what in the end is still an analogue electronic circuitry after all.

Greetings from Hannover!

Manfred / lini