This is probably a stupid question, but I was not able to come up with a satisfactory answer...
My question is about the transport. If I am not mistaken, the job of the transport is to get the data from CD's (0's and 1's) to the DAC. And therefore the better (less vibrations...etc.) the transport the less errors are going to be made. I have seen A LOT of efforts made by manufacturers to ensure the stability of transports, some costly ridiculous amounts of money. But here is my question, my CD-ROM drive in my computer, which presumably uses the cheapest possible kind of transport, does not seem to have any problems in achieving 100% accuracy in data transfer (afterall, all my documents have loaded up 100% correct, and my program CD's, like Windows XP, have installed perfectly...).
So why does the transport matter? In what ways is it important???
a very good question, I am not sure I know all the answers.
One thing that I know is that in CD playback, the data stream is essentially synchronous, else you won't get continous music playback. So any error correction, or re-reading of the disc must be done in real time. (E.g. UK Meridian uses CD-ROM drives in its current line of CD player, the data is read off the disc multiple times, and then buffered before being fed to the DAC.).
Another thing that I know, is that in the simplest case if transport and DAC are separate, the DAC unit recovers the clock from the digital SPDIF input line from the transport.
(Now this is not always the case, some more sophisticated DAC units make use of buffering and have an independant master clock for D to A conversion). But in many cases the master clock in the 'CD Player system' as a whole resides in the transport. Hence this makes the engineering of the clocking more critical.
So from my limited understanding, it is the ability to perform disc re-reading / error correction in real time, and the possibly the need to act as the master clock source in a CD playback system, that make the engineering more challenging in a high quality CD transport as opposed to a PC CD ROM unit. Also there may be mechanical aspects, like the need to provide mechanical and acoustic isolation of the CD drive's laser pickup mechanism.
Now we have hit the limit of what little knowledge I have.
cheers Rav
P.S. If you posed this question in a "let's look under the hood" forum like diyaudio.com you may get some more intelligent answers....
one further thought, I did not explain veryt well.
Most CD players do not make use of a computer CD ROM mechanism (unlike some newer machines), instead it is a dedicated audio CD drive, which spins at 1X or 'normal' speed. So that means that the PCM data should be read from the disc hopefully correctly, the first time round.
Thanks for your thoughts! It's great informaton! I didn't know very much about how the transport mechanisms, clocking and data buffering works, so that helps a lot. Also it is true that the acoustic isolation is very significant, as I recall CD-ROM drives could themselves be very noisy, which significantly distracts from music listening.
I still am not clear about the need of the buffering and error correction though... Again, because the CD-ROM mechanism has demonstrated that it could retrieve data 100% accurately (that could be objectively demonstrated since if a program or a document loads without mistakes, the CD-ROM obviously is 100% accurate, as opposed to CD players, in which the accuracy can only be subjectively heard/evaluated). But does that mean a CD-ROM mechanism (which is 100% accurate) is as good as the best possible transport mechanism out there (can't beat 100% accuracy)? Which totally does not make sense to me. Afterall, transports can be ridiculously expensive and can involve very elaborate constructions...what's the point???
There must be something I'm missing, please help out.
Hey, Roger. In fact, yes a cd-rom will have 100% copy, as proven by a working program or file. The difference is the bit-rate transfer along with file type archetecture. Files and programs do not matter in regards to transfer rates. The litle egg timer shows that all the time, as compared to clocked programs(like Mozart) which must be timed and, in worst case, corrected, according to the clock rates and errors in clocking(jitter). Computer clocking only sounds decent because you have not heard the proper timing, only the original error prone recording or file.
The transport and DAC also use seperate power supplies, further seperating possible interference. In fact some DAC's come to mind with seperate torroidal supplies for each channel, further isolating nasties.
you make some very good points!!! And Nuck does too!
Yes the cheap CD-ROM can retrieve data from a CD-ROM with 100% accuracy (but when it comes to Microsoft I sometimes wonder!! (JOKE)).
Also when the CD-ROM drive cannot read the disk for whatever reason, the hardware and software combined will tell the end user that the media or the file itself is corrupted. But when it comes to music playback from CD, we don't want to do that, error correction is applied that I guess tries to approximate to the original data, and hopefully sounds musical too!! This area is a complete void to me, other than being aware that CD transports have in built hardware error correction mechanisms for music CD playback.
I think the key is the real time nature of music playback, as Nuck states.
After all when we retrieve a file from the CDROM, if it is a data file, it is not really too significant if the transfer takes a matter of milliseconds or a few seconds. This just depends on the level of performance of the computer we are using.
Also as Nuck states, electrical isolation is also very important. We want to isolate the DAC from noisy micro-computers that deal with user I/O and operation logic, and transport servo's and motor circuitry.
We need to isolate the laser pickup mechanism from external vibrations, just like the cartridge in a vinyl record player.
Also the lines are getting very blurry these days. Some of the current hifi CD transport units, also perform upsampling of the data prior to sending it out on a digital interface to the DAC. One machine that I know which does this is the Chord Blu transport (very expensive and nice too!). So this means even a standalone transport unit may contain on-board digital processing and multiple power supplies, so this adds to the cost.
Also on some super high end stuff, there may be provision for the exchange of word-clock info, between the transport and DAC, and may make use of proprietary mechanisms. This is one area which may require costly solutions. Also clocking at a microscopic level is crucial to digital music playback, it is a problem of 2 dimensions, amplitude and time. So the clock in any CD playback system is crucial, whether it is an integrated CD player or transport unit.
All of this type of detailed engineering adds to cost, and for niche products you the customer are paying for the R&D cost, since there probably is not a payback in the lifespan of the product.
Also when you strive for perfection, I guess you enter the land of diminishing returns, hence the extreme cost of some transport units e.g. Teac Esoteric comes to mind for some reason.
These are my thoughts, they are by no means definitive.
Also, I think I can extrapolate where you are going with this. A logical question is why can't we just use a computer CD-ROM drive as our CD transport and have done with it...... I guess this is possible, but we would need to be aware that the SPDIF output from such a drive is probably very poorly clocked for high quality music reproduction. So any DAC employed would need to buffer the incoming data and then re-clock it very precisely before feeding it to the DAC chip itself. I would guess that this is a reasonable (but not optimal) approach, but perhaps would not make many audiophiles happy, who would prefer see an CNC machined ingot of aluminium sitting on their rack space.
Wow, good points guys. Now I see the difference and am convinced. I may have to save up for the Rega Planet 2000 (or whatever the new version is called) after all. :>
You're right, Nuck, you've pointed out the crucial difference: In CD-ROM it really doesn't matter how long it takes to load the stuff. Also the sources of interferences that come into play.
Rav, thanks for you clear explanation. You were right on when you said in your very first post that "One thing that I know is that in CD playback, the data stream is essentially synchronous, else you won't get continous music playback. So any error correction, or re-reading of the disc must be done in real time." It just took Nuck's further explanation to drive the point home.
Also, Rav, you read my mind. A computer CD-ROM transport and an "audiophile", external (ie. out of the computer case) sound card is what I was getting at (wishful thinking?). Your idea about buffering the incoming data and re-clock before feeding to DAC seems interesting, and I wanted to go further with it. So here are my questions:
1. In a computer, where is the clock?? (CD-ROM drive? Motherboard? Soundcard? All of the above? If there are multiple clocks, which one dominates?)
2. Can a good soundcard (eg. Roland (Edriol), M-audio or Soundblaster (some of them uses Wolfson DAC's) achieve what Rav suggested? (re-clock the incoming data)
3. Anyone aware of CD-ROM drives that are designed for audiophile use (or mid-fi use, at least?), those can be put in quiet pc cases, which have some anti-shock, noise reduction mechanism for its drive bays (albiet relatively primitive).
Or am I likely getting no-where and should start saving for a good CD player?
before "trying" to answer your questions one further point, that may help you:-
Don't try to use a cheap DVD player as a transport. The chipsets in all these DVD players, convert the 44.1Khz sample rate data from a CD, to "up-sampled" 48Khz data stream to match the DVD standard for audio on DVD discs. Now this is an asynchronous upsampling conversion since 48Khz is not a direct multiple of 44.1Khz. This introduces the potential for jitter in the purely digital domain, as opposed to jitter rearing its ugly head at the time of conversion to analogue.
Please dont ask me to explain the above, my brain understands it, but my brain is not so good with words. Basically you can be sure that the DSP filter algorithm employed will assume that the clock is perfect, when it is calculating the interpolated sample values. And of course in a cheap $60USD DVD player the clock circuit will not be perfect!!! Only use this type of system if you like the sound of jitter!!!!
One way to go is to use a good old 2nd hand CD player as transport, one which puts out the CD data at 44.1Khz on its SPDIF interface. Then use a nice DAC like the Benchmark DAC1, which has buffering, and reclocks the data leaving the buffer prior to conversion to analogue.
If you are using a PC based solution I am no expert. But here is what I know.
1) Most CD-ROM drives are a cheap CD player, since they have analogue line outs on the rear connector. So there is a clock in the drive for sure. But I guess you would not use the analogue out from a cheap CD ROM drive. The soundcard has a clock for the digital audio, and when using the soundcard, this one takes precedence of course. There is a clock on the motherboard, but this drives the PC's CPU, so needed of course, but has no bearing on playing music back.
I believe some of the best soundcards out there have respectable circuitry, in terms of the chips used. The only problem is if the sound card is in the PC chassis, and you use the analogue outputs of the soundcard, I am sure it will be 'fine and okayish', but no match for a dedicated audiophile CD player. I don't think the power supplies and noise environment in a PC are the best for high quality analogue audio circuitry.
I would think a good solution is to use a good sound card and connect it to a good DAC like the Benchmark DAC1. This will give you the convenience of PC mass storage.
But, myself I prefer using a Slim Devices squeezebox for this type of system JMHO. That way I can keep my ugly PC away from my nice hifi system.
If you want the convenience of a music server, and a nice user interface a PC and a Slim Devices SB3 is a good way to go. This type of system can be very good using the internal SB3 DAC, or can be just as good as the external DAC you use, provided the DAC does buffering and re-clocking.
Oh I forgot your question 2) I am sure that a good sound card will re-clock the data from the CD ROM drive prior to conversion, but I dont know of the level of buffering employed. Please bear in mind most soundcards are really serving the home studio, aspiring musician customer base!!
But ultimately it is horses for courses, myself I am highly enamoured with the Slim Devices Squeezebox, for convenience and high quality. Also usinmg the SPDIF digital out of the soundcard with a good DAC is a good solution I feel.
But using the analogue out of the soundcard in a PC would 'I guess' be beaten by a respectable dedicated CD player. My comment here is somewhat loose, it is based on listening to the recordings of some friends, who have musical 'talent', and whom feel that there creative soundcard is the best thing since sliced bread. I had to be a little polite here. I am guessing since I did not do a side by side, but my guess that a Rega Planet, NAD 542, or Rotel 1072 will outperform the analogue out of the soundcard.
But for the ultimate sound, you will not beat a high end dedicated CD player, but then again most of us cannot afford such dream machines.
Again though it all depends on your personal biases, not everyone wants to boot up a PC or Mac just to hear some Mozart, or P.Diddy (godamnit). So maybe a respectable CD player like the new Rega Apollo is a nice simple elegant solution after all.....!!!
Roger,I will likely have all 3 together in a couple of weeks. Classe DAC,Classe transport, Rotel 1072 cdp and Creative labs X-1 soundcard with multiple digital outs. I hope to have them all working in short order, a review might help you out.
My thinking is that the digital files sent out to DAC(after eq'ing to my tastes, if desired) should match the best cd burn of those same files. I'll burn the files as well, and play in both external units for a direct comparo.
Thanks Rav, once again for the info. I think your explanations are very clear! Oh yeah, and I will remember not to use a cheap dvd player as transport.
I will eagerly await your review of your new system/comparisons, Nuck, thanks!
This is getting highly addictive! Never thought participating in forums is so much fun!
the email notifications of postings are coming through and distracting me from my work. Its 3:12 AM here in Melb.
Tomorrow I face my Waterloo, and this should be my last all nighter fuelled by Red Bull.
So no more late nights for me for a few weeks to come, and plenty of R&R. I have pulled the plug on a work project that had totally gotten out of hand.
I await to see what the repurcussions are since, I kind of under-estimated the workload involved. But hey! no other bozos were standing up to make themselves heard. This will be a little mini A380-saga, quite humbling for me.
I notice both you guys are in Canada, if I tell you that I work for a Canadian telecommunications company, and I'm sure you can figure it out!
In the best case my rep is in tatters, worst case could be "employee status terminated", but this would be rather harsh, and I may have grounds to litigate.....
But back to hifi, best of luck Roger, I hope that you are able to figure out what you want to achieve. Also I recommend diyaudio.com, since there are people there far greater and in depth knowledge that myself. Its a good resource.
And hey, if in the coming weeks, I get fired, I can take a trip to the USA, buy a bulk lot of Kinber 8TC speaker cable, and start my own aussie EBay store, cos speaker cable sure is expensive down here.
I am a follower not a leader, but yeah sometimes you really wonder where the heck you are being led!!!!!
In any case I think I am about due for a major-ish karmic re-adjustment.
Anyhows, all the Classe' stuff sounds very good, and in fact I should finally get the time to finish my mods to my ancient but strangely wonderful JVC XL-Z1050TN CD player. I have high hopes for this machine, since when you lift the lid (unlike my XA-5ES), this one is full of components that no one wants to own up to!!! So I think it is an understatement to say that this machine has a lot of potential for improvement.
Also, look forward to hearing how your Classe' gear works out. I imagine it will take some time to get to grips with it all. Not least, the small matter of re-EQing and burning some of your disks. (You could give up the day job and open a mastering suite.....)
But Rav, giving up my day (week) job would mean missing out on Iowa this week and Halifax the following. I am a weekend listener at home, that was a bit of a snag to get through, but the stuff will be around forever, so when I win the lottery, I can enjoy it more.
Dan, thanks for checking in. I have actually gone with a Goertz cable, much better than the SonicLink. The link was a bit sloppy sounding, just a bit loose in the bass rolloff, the Goertz is much sharper sounding. Please do check the other link and look into the Lehmann DAC+, I am trying to get some feel for this unit.
Yeah, the DAC-1 is a legacy piece for them now. I get support from them, but just one guy there.
The CDT-1/DAC-1 were a pair, and not the best, so far. Don't get me wrong, the kit is well rounded, capable and shows moments of absolute brilliance! However, I use the Rotel 1072 for a lot of music. Floyd plays superbly on the Classe source.
Nuck, do you have any experience with the Benchmark DAC-1? I've been curious for quite a while what a decent yet inexpensive CD player and the DAC-1 could do together...
I don't know if it's available in town or not, Nuck. The Benchmark site lists Edmonton as the nearest retailer. I have some interest, but not enough to make a special trip to investigate ;)
