how can digital be different

Hmm, as Mo says, wouldn't have thought it'd be pricey. After all they build them into 50 quid portables. Just strikes me that if this is really an issue then spending a few quid per unit on such a system might be well worth it.

 

I imagine the cost would be that of adding a buffer chip, in the world of makeing money that would be considered to much

 

Some manufacturers spend loads just making there kit look as beefy as possible. Would have thought this would be a drop in ocean in comparison (and who knows might even be more effective) - that is assuming its really an issue of course

 

All of it. Those IAR articles are unfortunately full of BS

Michael.

 

For instance...

 

there may also be mechanical considerations too. if you are spinning the disc faster and possibly intermittently the lifespan of the drive would be shorter, not good in a multi '000 pound cd player. also you've got noise considerations too - cd-roms aren;t the most silent transports ever are they. also the addition of buffer memory may have adverse effects on jitter performance. not an issue on a 50 quid portable but on a costly player if it's got high jitter it MUST be crap (if there are any americans reading that was sarcasm btw ).
other than that i couldn;t guess.
cheers


julian

 

Not true AFAIK - the error correction is sufficiently robust that the data is always recovered bit-perfect. The question of error correction is largely irrelevant to the transport argument which concerns jitter, ie. timing errors. If there is an error in reading the data, there's a loud click or pop of some sort or the CD skips or repeats, but this actually constitutes an improvement in the sound quality as it makes it more vinyl-like

 

If transports are so perfect how come things like the green pen make such a difference?

 

this is going to rumble on me I prefer a sanyo cd drive & russel hobbs leads trashes my wadias big style

 

Or a DVD player, the image will distract you from all those problems... :guiness:

 

If the green pen makes such a difference, how come you cannot prove it?

Seriously, we can repeatedly rip a CD on a variety of CDRs and get the same results each time. And this can happen at greater than real time speeds. So there's no intrinsic problem feeding a DAC with accurate data. 'jitter' then becomes totally the responsibility of the DAC.

If transports sound different I think you have to question the engineering ability of the transport and DAC makers. And why people are joining together components in a way that makes trouble for themselves and increases expense...

Paul

 

.............which is why I now have a one box player.

Funnily enough, the green pen works with that to. Does that mean my one box player is crap too?

 

urghhhhhhh pplllleeeeeeeeeaaaaaaaaasssssseeeeeee not another green pen debate

:banghead:


I'd even rather talk about Mana

 

Just stirring the pot a little!

Gimp suits at the ready

 

peteh,
if what you say were true there would be no need for drives to re-read data which has been incorrectly read. i can assure you that most 'drives' be they cd-rom or hard do indeed have to re-read data when incorrectly read at some point or another.
i'm not sure of the exact specifications of the rs coder used for cd-a but as a system rs is better at correcting 'burst' errors (errors which occurr one after the other) than randomly distributed errors. obviously this will deal with physical defects on the disk better than random electrical noise induced data errors.
an example would be if you were using an rs system where for every 255 bits of data you used 32 bits as parity information, and had 8 bit words then you'd be able to correct up to 16 individual errors so one 50 bit 'burst' of errors would be fine but if those 50 incorrect bits were randomly distributed through the 232 bits of real data then there would likely need to be a re-read. also if the errors occur during the reading of the parity bits then you also run into problems.
another thing is that you've got to remember that cd's are 20+ years old. the amount of processor power necessary to impliment rs decoding is not trivial so red book probably uses a low number of parity bits and so won;t be able to rescue many 'words'.... i'm winging it now so i'll shut up.
cheers


julian

 

Feel free to bring the green pen debate here.

 

even linear tape drives re-read. Imagine how much more work is involved in that. Infact they can make a fair amount of re-read before a fail condition. Even though i dont work on optical or cylinder drives, i imagine, just from the mechanics of them, it is a hell of a lot easier to re-read on that media

 

Me too I don't believe it to be the case though that the error correction starts making approximations to the original data if it can't quite retrieve all the data - you can't make approximations in the digital domain. Either the data is read bit-perfect and your CD plays, or it's not and you typically get loud pops or cracks - the whole point of the digital storage is that there's nothing in between. As I say though I'm no expert so I'm quite prepared to be proven wrong.

 

Julian, you are right in part...but wrong in as much. If the bursts you talk about are short enough then you are right about being able to correct more bits that way, but you are more likely to be able to correct a given (low) error incidence if it's spread randomly across multiple frames than if they are all localised within a single frame - hence the levels of interleaving put into the system (i.e. deliberately scattering burst errors across many different data words). If anyone wants to find out about how the error correction is done etc people should read something like this well-written article. And as adb has said repeatedly in the past, the error-correction doesn't "kick in" and "work harder", resulting in some analoguy sag or distortion - if works at the same rate continuously while the player is playing.

I would be very surprised if audio players do re-read data - misreads IME just result in a click. Re-reading would also not help with damaged media, only external influences (as in e.g. anti-jog)

OK..just a couple of examples

1) These "sketchy clues" are the data bits being read off the disc which are are read off essentially perfectly in all but the worst cases. While CD-ROMs in your PC have a little more circuitry to do an even better job, think what would happen to your PC if if couldn't. The whole point of digital transmission is that the exact value of the signal amplitude has NO SIGNIFICANCE. The only important thing is whether it is greater or lower than a certain value - if it exceeds it then there it is IRRELEVANT by how much. 2) That figure of 2kHz is complete crap - maybe they meant 20kHz. 44.1 kHz can distinguish fully (to a resolution dependent on the bitdepth of the digitisation) any frequency below 22.05kHz.

This assumption is what's called a standard! If the data is not stored at 44.1kHz then all bets are off. But this is not a difficult standard to meet - crystals are avaiable for less than a couple of quid that provide standard frequencies to better than 1 part in 10e7. That's many orders of magnitude better than you'll get off the best TT. (And to accommodate drift or physical tolerances on the disc it appears there's 3 sync bytes in every 36 on the disc)

With this amount of misinformation, distortion and obfuscation how can we work out what is useful information (which it looks as if there may be scattered amongst the garbage) and what is more crap?

 

I'm not getting involved in the transport debate...

But If you think it's all just a matter of 1s and 0s then I suggest you get your mate to look at exactly how the receiving circuitry in a DAC decides whether the signal comming from the cable is a 1 or 0 (GrahamN as already given the answer).. then get him to research into the affect of Capacitance on a squarewave signal.. and he should (if he has any capacity to reason) realise how the cable can corrupt the signal in such a way as to induce errors (both bit and timing) in the input circuitry of the DAC. Once these errors have occurred.. they CAN NOT be rectified. The DAC has NO cirtuitry to enable it to do so.. it also has NO IDEA of what the signal SHOULD look like.

GTM