[Phat Phreddy]

Malcolm, Well its nice to have a con to my Pro !!!

I have listened to some high end equipment but never had it in my home to test in the same signal chain and play these games...

I would express surprise that A) there was such a level of coloration as this indicates that there is a large amount of error correction differences in both these machines and B) that mere mortals (spouses and children) can readily hear them... If that read sarcastically it is not intended to AT ALL... However I find once equipment is above a certain baseline.. It takes a pretty good ear to discern small changes like cabling etc.. I would have belived that the level of difference and error correction involved in a Black CDR against a vannila CD would not be dramatic enough to become obvious to a casual listener...

Dont think I am questioning you... I enjoy any valid discussion and belive your results but wonder what else can effect them...    

I am guessing that the DAC's in the CD player are what you are using ?? I am guessing analogue out of both machines ??

[Phat Phreddy]

Felipe, I guess you read the shootouts done by Branxx I linked above... The RME card seems pretty fine and well ahead of the 1010...

There are rumurs that M-Audio will be releasing a 1010 replacement with new better DAC's (already pretty damn fine) and the ease of use of the revo or 410 drivers... That would give them CSII, speaker delay, bass managekment etc in a high end rackmount external to the PC... I am curious about this offering...

[Larry]

I did not see on what you disagreed with me. I think we both have no dispution on that EAC is good at its task.

Sorry Larry it was probably my reading.... I took this post

SamL wrote:
You can skip step one with the correct CD reading software. Try EAC - Exact Audio Copy.
http://www.exactaudiocopy.de/


This CD reading software does a very different job for a different purpose from the step one of checking reading quality and it cannot tell what the step one tries to tell.

This type of the software tries to correct the erorrs while reading. That's why the author found different software sounds differently.

The step one tries to check the hardware reading quality before any touchs on the music by software. If to check CDROM drive's quality, do step one and avoid using any software like this, otherwise the results touched by this type of software have to be interpreted differently. The errors reported by this type of software may not be caused by the reading CDROM drives, in the other hand, errors caused by reading may not be reported, so that it does not tell how well the CDROM drives or CD writers do the reading.

I still am a touch confused but perhaps you were refering to the alternatives to EAC when stating "avoid any software like this"... Anyway I have no need to argue... It was me that probably missread it...  

I did say "avoid any software like this" (EAC) in checking hardware reading quality. I also said "use EAC in actual coping". Do you see the difference here?

Again though EAC does not perform error correction...

How did you know?

If the drives firmware does not perform error correction (and it is my understanding that it should not but...)

The drive does perform error corrections and it should.

then EAC reports all faults as it attempts to reread multiple times until it is sure that any suspect sectors are what it believes...

I am not sure what EAC actually check on. There are layers of error correction coding on CDs. Which layer the EAC is checking?

What it believes is what is correct? What about errors while CD pressing, which will be read by any drives as the identical data in multiple reading?

[Larry]

Hi, Felipe,

My audio experience of PC is limited with using my Creative Extigy for normal non-hifi duty, though I am interested in how to directly read data from digital input on a soundcard and send it back to digital output after filtering in real time, but I have not found a way to do so yet, which could open a new world for me to indulge in. I am dreaming of using my own algorithms to do room correction, auto-Q, digital crossover, whatever you can think of before DAC. You can see that I am not quite having the peace of mind with analog on a soundcard.

Using 1/10 of cost to get an equal preformance would be a great achievement. I think I need to catch up with what everybody is doing in the direction of HTPC.

[Phat Phreddy]

I did say "avoid any software like this" in checking hardware reading quality. I also say "use EAC in actual coping". Do you see the difference here?

Sure... As I said its not a biggie...

How did you know?

If you mean how do I prove'ably know, I dont... I am relying on the info I have read about EAC (by its coder and people who admin the EAC forums)... Also the purpose of EAC as a ripping app is to rip and report what it finds as errors and suspect sectors..  The whole point of EAC is to perform as accurate a rip as it can and when it reaches a suspicious sector will reread again and again until it has (IIRC) 8 identical bits out of 16 tries... If it does not get 8 identical bits it retrys 5 times... So questionable sectors get 82 reads and must make a greater than >50% all the same in any one iteration...

Of course it also performs jitter detection and turns off any audio cacheing etc but thats not needed to say I think...  

The drive does perform error corrections and it should.

Well this I am not certain, it is my understanding that when a drive has had all its cacheing / buffering / etc turned off the drive should function in RAW mode... When used this way should the drive not report each and every bit as it read it ???

What it believes is what is correct? What about errors while CD pressing, which will be read by any drives as the identical data in multiple reading?

The way I understand it EAC does not rely on any error correction system provided to it by the drive... It reads in RAW mode (as I used the term above) when set up in secure and reads every sector at least twice... This way it provides it own check system... When the initial bit matches the second read attempt it is happy... If there is a discrepancy it starts the above listed 16 checks and does 1 3 or 5 iterations until it gets a >50% exactly the same...

Errors in CD pressing will still be read and stored as errors... Whats on the CD is what you get ripped to HD in every test I have done (with mostly clean unabused CD's)...

[Phat Phreddy]

Larry,

You may be interested in the work of Don Maurer

http://home.pacbell.net/donwm  

Realtime digital speaker correction... The same principals apply to room correction (and this is where the PC could start to turn a lot of things in its favour... Source, Pre / Pro, Speaker and room filters, either outputted to the DAC or using the soundcards DAC's all in the digital domain...

Unfortunately I cant get him to release the source but he would at least have clues for you..

[Malcolm Fear]

Phat Phreddy

>>However I find once equipment is above a certain baseline.. It takes a pretty good ear to discern small changes like cabling etc..

I find the opposite. Once you get beyond a certain level of equipment, the changes (cabling etc) are easier to discern.

>>I would express surprise that A) there was such a level of coloration as this indicates that there is a large amount of error correction differences in both these machines

We were surprised by the difference that the support system made to a cd player. I brought home my shiny new Rega Planet 2000 cd player at the same time a friend brought over his Meridian (that sounds good, but actually, I ran the Rega heavily for a couple of weeks to run it in). We just plonked them on the floor, used ordinary cables and listened. The Rega was so bad, my friends suggested I return it and get my money back. I then placed the Rega on a wooden shelf, with an innertube under it. I put in some special cables. It was difficult to tell them apart. We then did the same with the Meridian, and it improved to the point where the Meridian was a little better than the Rega. I was happy. I kept the Rega, I can't afford the Meridian.
We then bought Corian shelving (we went to a kitchen benchtop manufacturer and bought the sink cutouts and had them resized). We still use the innertubes. We both noticed an improvement in clarity, air and sound stage.
We both scratched our heads and just accepted. Error correction again? I don't know.
The black cd did the same thing.

>>I would have belived that the level of difference and error correction involved in a Black CDR against a vannila CD would not be dramatic enough to become obvious to a casual listener...

I agree. But the platform a cd player sits on having an audible effect surprised me more.

>>I am guessing that the DAC's in the CD player are what you are using ?? I am guessing analogue out of both machines ??

Correct.
I have tried a friends outboard DAC. It is better than the Rega.
Hugh is releasing an AKSA DAC at some stage (I hope). I will wait for it.

[Phat Phreddy]

I find the opposite. Once you get beyond a certain level of equipment, the changes (cabling etc) are easier to discern

Well if I rephrased it... Without a certain level of equipment it is nearly impossible to detect the changes of minor tweaks (agree with you) but still feel that most of the tweaks tend to introduce subtle change in the sound... Such that in my experience most people simply dont focus and listen (or care) enough to discern...

The only thing I can contribute from a logical point of view is that the inner tube is providing damping to vibration that the CD players drives need.  If they are getting a better extraction with this solution then they must be getting a more accurate read when this occurs... In turn this means that they must be performing a lot of error correction when they are undamped and facing many read errors... If this is the case again I am surprised.

I would find it extreemly hard to accept that the vibration damping (of this level) has an effect on the electronic output stages (but maintain an open mind)... I know people put amps on inner tubes but IMHO this is where we start down the road that ends with us asking for a frontal lobotomy and 10 rolls of rubber wallpaper for christmas !! I dont mean any offense by that, but I am firmly in the camp that thinks 'magic' stones rare earth compounds and the like do only pshychological good...

[AKSA]

Mal,

Yes, eventually Aspen will release an AKSADAC, but it will take some time, and I would not expect something before end 1Q 2004.     This is a long, demanding project and the big problem will be reducing the complexity of the pcb, which must be double sided at least.  

Let me give some views on this, and remember, I'm not a digital expert, this is Ben William's domain and I'm just the supporting cast!

There are six fundamental issues with DACs as we see it:

1.  Control of jitter.  This mandates very accurate clocks, of the order of 2 parts per million, and for best performance necessitates slaving the transport to the DAC clock.
2.  Power supply topology and implementation.  This is a major failing of many DACs;  the digital current streams are abrupt and fairly high current, and the power supply demands are multiple.  I have a few clues on power supply design, and you can be sure this will be addressed very thoroughly.  I do not favor battery supplies.  They are costly and impractical, in precisely the same way that electric cars are a PITA.
3.  Choice of chip.  Most are compromised;  we will be using the PCM1704 from Burr-Brown.  This is pro-audio, one chip per channel, and expensive.
4.  I-V conversion.  This is a major cause of inaccuracy on present systems due to the non-linear operational mode of the active conversion element, often an IC.  We will be using an all discrete high voltage converter with 300MHz transistors to avoid slewing of the HF artefacts before passive filtering (multibit, 8 times oversampling), and the design is already complete for this element.
5.  Analog current amplification.  This is often performed badly, using high feedback opamps, with all their failings.  We have solved this problem.
6.  Format selection.  We will be using a microprocessor to control format selection, reducing pcb complexity very considerably.


Many decisions remain.  Do we offer a full kitset, or part.  Who is the target purchaser?  What are his/her tastes and requirements?  What is the price?  What do we assemble here in Melbourne?  How many layers on the pcb;  how much do we spend on the circuit elements, and chips?    What is manufacturer cost?  How is it to be marketed?    Do we offer SACD and DTS?  And so it goes on.........

One thing is certain.  Since I am geared towards both excellence and simplicity, and this is a highly competitive area, I don't feel I have any choice but to aim for the ultra high end, as good as say a top-of-the-line Meridian, essentially studio quality.  This is no problem for me or for Ben;  it's a huge mental challenge, and Ben is a veritable genius from whom ideas blossom constantly.  My input on power supply, I/V conversion and pcb design are most complementary to his, and I'm confident the product would be one of the best of its type in the world.  But naturally it would not be cheap, and I really have no idea until I do the sums on how much it would all cost.  Suffice to say that development of this product will be rather expensive, and this will inevitably reflect to some degree in the retail cost, but hell, there's no free lunch.......

So, in closing, I'm interested in feedback.  Many of you have huge expertise in digital electronics, and I'm interested in feedback.  (I find negative feedback particularly useful as it really helps with the marketing angles........ )

TFTBW,

Hugh

[SamL]

Well,
Don't know much about DAC so this will be a silly question.
This is how I understand how DAC fit in for 2 chn music. Transport->DAC->preamp.
How is it look like for a 5.1 HT system, where is the HT decoder sit? Do I need multiple DAC?

Thanks,
Sam

[AKSA]

Sam,

Yes!

Hugh

[Larry]

If you mean how do I prove'ably know, I dont... I am relying on the info I have read about EAC (by its coder and people who admin the EAC forums)... Also the purpose of EAC as a ripping app is to rip and report what it finds as errors and suspect sectors..  The whole point of EAC is to perform as accurate a rip as it can and when it reaches a suspicious sector will reread again and again until it has (IIRC) 8 identical bits out of 16 tries... If it does not get 8 identical bits it retrys 5 times... So questionable sectors get 82 reads and must make a greater than >50% all the same in any one iteration...

This is a error correction strategy, called automatic-repeat-request (ARQ) scheme with a majority-logic, in this case repeated reading. With this scheme in place, it can effectively clear most of hardware misreadings caused by vibrations, defective laser head etc or marginal pressing errors on the CD.

So to check your hardware quality, don't use it as it mask misreadings of a CDROM drive, unless you can force it read only once. But it's good to use while actual copying to eliminate the effect of random laser pick-up errors.

The drive does perform error corrections and it should.

Well this I am not certain, it is my understanding that when a drive has had all its cacheing / buffering / etc turned off the drive should function in RAW mode... When used this way should the drive not report each and every bit as it read it ???

The drive is supposed to report the errors it can not correct but no standard enforces firmware programmers to do so.

The drive reads data in blocks of bits, not bit by bit, and the error correction coding also works on blocks of bits so the drive reports errors for each block.

In music CD drive, I don't think the average CD players will have complex algorithms to do a good error corrections so that the chance is that all corrections are dependant on the drive. Open a CD player then you will know that there is no computing chips there are able to do so. Therefore, errors will be passed to DAC and result in distortions.

What it believes is what is correct? What about errors while CD pressing, which will be read by any drives as the identical data in multiple reading?

The way I understand it EAC does not rely on any error correction system provided to it by the drive... It reads in RAW mode (as I used the term above) when set up in secure and reads every sector at least twice... This way it provides it own check system... When the initial bit matches the second read attempt it is happy... If there is a discrepancy it starts the above listed 16 checks and does 1 3 or 5 iterations until it gets a >50% exactly the same...

Errors in CD pressing will still be read and stored as errors... Whats on the CD is what you get ripped to HD in every test I have done (with mostly clean unabused CD's)...

According to what you described, it is true that it does not make use of the error correcting codes embedded in the data, however, as said above, it implments an ARQ scheme on top of that.

In my opinion, after that, the software should  further correct  errors caused by CD pressing, scratchs etc, making use of the embbeded error correcting codes on the CDs before it burn into a new CD. This way, the copied CD will be cleaner, which will benifit the average CD players that do not do sophisticated error correction.

[Larry]

Larry,

You may be interested in the work of Don Maurer

http://home.pacbell.net/donwm  

Realtime digital speaker correction... The same principals apply to room correction (and this is where the PC could start to turn a lot of things in its favour... Source, Pre / Pro, Speaker and room filters, either outputted to the DAC or using the soundcards DAC's all in the digital domain...

Unfortunately I cant get him to release the source but he would at least have clues for you..

Thank you, Phat Phreddy, for pointing me to this link. It is evry intersting.

I quickly went through his system setup, he is using the DA/AD on the sound card. In this case, the data can be easily accessed as a block through Windows APIs.

I have no worries to allow computers to touch my digital siginals but no way to touch my analog signals. The digital signals can be easily cleaned and decoupled from the computer however the analog siginal can not be cleaned without compromising the sound quality. If there are any analog chips on the card touching the signal, it spoils all the fun of doing digital filtering. I think in his application, it's probably fine as he is trying the idea of digital crossovers.

My idea is getting the data straight away from the input byte by byte to the memory, which needs to crack the soundcard drivers to do so. But, these days no drivers are willing to yield the control to software. There might be some soundcard drivers do so, but I am not aware of.

What I really like to have is a card with only digital input and outputs on it and all the rest of the chips on the card is stripped off. But I don't know who would produce such a card.

[Larry]

>>However I find once equipment is above a certain baseline.. It takes a pretty good ear to discern small changes like cabling etc..

I find the opposite. Once you get beyond a certain level of equipment, the changes (cabling etc) are easier to discern.

It should be so from principles.

We were surprised by the difference that the support system made to a cd player.

I think the CD drive in your CD player is sensitive to the vibrations. This is not much to do with the CD media. The vibration causes laser pickup errors that the dirve is unable to correct. It does not need a big vibration to cause troubles to the laser pickup if the vibration frequences are right, as we are talking about 0.8um displacement. If you have no worries about warrant, open the case and use enough BluTack to damp the CD drive and chasis. It may improve further if your Corian did not fully eliminate the effect. Most CD players including some expensive players use very cheap CD drives, in which
a lot of important parts that should be made of metals are all made of plastic.

If you have a DAC, you can try DVD player as a transport. DVD drives use different generaton of lasers that reads with much more accurancy and better resistance to vibrations. Use a reasonably good DVD player otherwise the noise can find a way into your DAC.

[Larry]

I would find it extreemly hard to accept that the vibration damping (of this level) has an effect on the electronic output stages (but maintain an open mind)... I know people put amps on inner tubes but IMHO this is where we start down the road that ends with us asking for a frontal lobotomy and 10 rolls of rubber wallpaper for christmas !! I dont mean any offense by that, but I am firmly in the camp that thinks 'magic' stones rare earth compounds and the like do only pshychological good...

We are agree the major effect of vibration will be to the CD laser pickups. However, if you have loose wires hanging around in the chasis that are prone to vibrations, you might have chances to hear that if some conditions are met. Vibrations to some other components like Quartz can cause signal noises too.

[Larry]

Hi, Hugh, you are eventually pulled into this thread  

I have great confidence in your incredible genius in audio electronics. I really like to see you come up with a DAC.

I am using a modded ART DIO with great satisfaction.  The system driven by similarly modded DIO won the award of Best of Show in CES. But one thing makes me, as a purist, a title you confered, sleepless is that there are a couple of opamps on the way of signals. I know it just takes a cup of coffee's time for you to come up with a solution to kick the opamps out of the way. It's every easy to redirect the signal from the pins of the DAC chip. So all mods are not basically parts upgrading, except class A biasing omamps. I believe if you touch its output stage, the quality of AKSA-DIO will jump to an unbeatable level. If you would like to do this first, I could place my advance order today.  

If you do a DAC from scratch, I will wait and see how much you sell it before I put my order. My DIO only costs me just few hundreds of Oz dollars including shipping from overseas and all parts for mods. The latest news is that DIO drops the price to US$97.50 stock, a price very difficult to beat, considering that it is a AD/DA with a tube. DIYers do not usually spend a lot on gears. Only those riches spend but in this case you cannot offer it as a kit to them and in this case again they may not pay for the sound but are very much willing to spend for a brand.  

[SamL]

Well, a mad inmate is building a tube DAC from Sheldons D Stokes. Design is free. http://quadesl.com/dac.shtml

Sam

[Carlman]

Sheldon is a very bright guy.  He likes to share his ideas which I think is very cool.  He posts to the Miataforum and some of his ideas have ended up being archived and considered 'the best way' to tackle certain jobs.

Go Sheldon!  He's been working on that DAC a long time.

-Carl

[Phat Phreddy]

OK I am not coming at these in order as one I want to get on later but...

We are agree the major effect of vibration will be to the CD laser pickups. However, if you have loose wires hanging around in the chasis that are prone to vibrations, you might have chances to hear that if some conditions are met. Vibrations to some other components like Quartz can cause signal noises too.

Hmmm good point on the quartz... Am I right in assuming ALL timing clocks are going to be using crystal oscilators ?? When it comes to the 'loose wires' and vibration susceptibility (at this level of vibration) permit me to smile ruefully ... I tend to find that ABX'ing (correctly done) tends to bring many of these theories down but thats simply my perception.

Most CD players including some expensive players use very cheap CD drives, in which a lot of important parts that should be made of metals are all made of plastic.

To my mind that is about the only conclusion that I can come to also... Doesn't that itself raise mental questions about megabuck equipment to you though... I was pointed to a snippet a while back that one of the megabuck transports (slipped from memory now) was pimping that they had a superior drive extraction technology... Turns out they had incorporated a standard PC drive within a different housing... Was it an improved technology ?? Probably... The fact that it was a $40 drive didn't make it to the paperwork though !!!  

[Phat Phreddy]

I quickly went through his system setup, he is using the DA/AD on the sound card. In this case, the data can be easily accessed as a block through Windows APIs.

I have no worries to allow computers to touch my digital siginals but no way to touch my analog signals. The digital signals can be easily cleaned and decoupled from the computer however the analog siginal can not be cleaned without compromising the sound quality. If there are any analog chips on the card touching the signal, it spoils all the fun of doing digital filtering. I think in his application, it's probably fine as he is trying the idea of digital crossovers.

Well first point I can think to mention is that the 1010 is an external soundcard with its own power supply etc and is decoupled from the PC... This same way of appoaching the problem is tacked by Ego-Sys in thier WamiRack solution... The 1010 is starting to be eclipsed as better DAC technology seems to appear by the month in this field...

I have heard rumours that we can expect a 1010 replacement 'this year' from the M-Audio consumer dept so that it will (should ??)  have the bass management and time delay / CSII / whatever features of the other consumer cards from M-audio (I actually quite like CSII but that may make many purists cringe)...

But external issues seems to be over rated... Everything I have read seems to indicate that the fears I used to have concerning stability of power supply / EMI / RFI / etc are not nearly as much of an issue as I would have believed... I wont say I am fully sold but I am a lot more open to the idea than I was 12 months ago...  Much has been made of the differences between higher end cards and reviews etc but today I found this little ABX test that makes fun of that... A cheap $120 soundcard (a couple of generations old now too, the REVO should be much better at output in theory) with multiple iterations of play / record and no conclusive ABX indifying them !!  

http://www.kikeg.arrakis.es/stest/

Now that clearly contradicts much of what I have read and believe concerning ABX identifying better soundcards (even via SPDIF) or how eg RME is more 'musical' than LynxTWO which may be more 'accurate' etc etc... If a cheap internal card played and recorded multiple times cannot be detected from the origional then ??!!?? As such I want to reread and see if I can find any flaws in the methodology and try this when I get time in the next few days (as I say I was only alerted to this page this morning)... Unfortunately I dont have any black CDR's to burn these WAV's onto  

I have some more I want to go into concerning the PC as a transport v the PC as a analogue output but I gotta get some work done... Back later...