Our life would be much easier if that would be truth.
Unfortunetly it is much more complicated if we desire good quality sound. Windows is the hardest to get good sound. Media player is no good.
You should read Ecdesigns threat, in the last couple of pages he writes about the bottleneck of PC systems and dacs.
"In order to get data into a DAC an interface with non-zero bandwidth is required. This is the same for USB, Toslink, coax, I2S, built-in transport and so on. For 192 KHz Toslink a bandwidth of approx. 13 MHz is required so bandwidth of the source interference spectrum will be limited. With USB interface the bandwidth is much larger, the HF interference can now easily crosstalk by means of very low stray capacitances.
Measurements show that source interference can be found on the ground plane, power supplies, data and timing signals and in the form of EMI.
Non-zero bandwidth allows source interference to reach the DAC, there is nothing that can be done about this. As soon as the (optical) connection is made, source interference will flow into the DAC electronics."
http://www.diyaudio.com/forums/digi...ding-ultimate-nos-dac-using-tda1541a-562.html
Regards Danico
Last edited:
Windows 7 audio architecture is the same as Vista, a 10+ years old technology that has been superseded and vastly overhauled by Microsoft on the latest operating system. This is an overview of the new Windows software audio stack https://msdn.microsoft.com/en-us/library/windows/hardware/dn302026(v=vs.85).aspx , it was introduced in Windows 8 but it really works in Windows 10 only (as most Windows 8 technologies). The new stack does suport offloaded audio streams: a device driver could delegate all the processing to the DAC board. I would say that the combination of Windows 10 and modern audio DAC does a pretty decent job. The Danico observation stands true, to get the pinnacle of audio perfection a dedicated hardware must be used. But before reaching the point of been able to hear the PC limitations, there are many more bigger bottlenecks to remove on the average Hi-Fi setup: speakers, room acoustics, amplifier...
I've had zero problems getting good sound from Windows. Download EAC and foobar, and at that point, the digital part is fine. I would hesitate to use the on-board DAC, but a decent external DAC (or an internal card if you're using a desktop PC) is inexpensive and can be very high performance. Lynx is as good as it gets, and the M-Audio 192 is 99% as good (measurement-wise, sonically equivalent) at 1/5 the price.
Where did you find this 'non-zero bandwidth' nonsense? Zero bandwidth is DC, and I can't imagine someone thinking audio, digital or analog, is DC!
Maybe it was meant to say 'infinite bandwidth'? That is exactly the opposite of 'zero bandwidth' of course.
But at any rate, infinite bandwidth is not required - to someone who understands re-clocking and jitter suppression this would be clear of course.
Interesting statements of source interference on ground planes - it is of course the PURPOSE of a ground plane to get rid of that interference. And ever since Kirchhof we know that any signal current needs a return path; what goes around, comes around.
Jan
Amusing is the link to the NOS DAC thread, a minefield of nonsense. All the worry about digital "distortion" while they plow the analog images into their preamps.
Folks:
I suggest that the 'The Ultimate NOS TDA1541 DAC' thread of John Brown's is not to be dismissed on a too quick reading. It's one of the more lengthy threads here, and John is no dilettante. He is a creative yet detail oriented engineer who has deeply explored many audio DAC design issues.
I believe that some are taking John's non-zero bandwidth quote out of it's full correct context. I believe John was referring to common-mode noise transfer. Ideally, you would want a zero bandwidth common-mode noise interface between source and DAC, except that the signal itself ideally needs an infinite bandwidth channel to help minimize jitter introduction. John uses those facts as a predicate for why he believes an purely optical interface betwen source and DAC is the best digital signal interface solution. He is well aware of the jitter problems which TOSLINK is prone to introducing, and so he has developed a separate solution to suppress clock jitter after clock & data recovery.
If I correctly recall, John doesn't view balanced interfaces as an optimum solution for stopping common-mode noise transfer because parasitic capacitance causes the CMRR of such interfaces to typically degrade as noise frequency increases. There is high frequency noise present on the ground of fast digital circuits, especially in PC based servers, even with the utilization of ground planes. Gound planes provide the best practical solution, yet even they are not without some impedance. Of course, it's debateable as to whether ground plane noise exists at a significant level, but I don't think it debateable that it exist. All of John's concern over minimizing ground noise transfer between boxes is due to his view that such noise introduces clock jitter within the DAC, a view I share based on my own basement experiments.
Whether one agrees or disagrees with John's technical views, he's no clown.
I suggest that the 'The Ultimate NOS TDA1541 DAC' thread of John Brown's is not to be dismissed on a too quick reading. It's one of the more lengthy threads here, and John is no dilettante. He is a creative yet detail oriented engineer who has deeply explored many audio DAC design issues.
I believe that some are taking John's non-zero bandwidth quote out of it's full correct context. I believe John was referring to common-mode noise transfer. Ideally, you would want a zero bandwidth common-mode noise interface between source and DAC, except that the signal itself ideally needs an infinite bandwidth channel to help minimize jitter introduction. John uses those facts as a predicate for why he believes an purely optical interface betwen source and DAC is the best digital signal interface solution. He is well aware of the jitter problems which TOSLINK is prone to introducing, and so he has developed a separate solution to suppress clock jitter after clock & data recovery.
If I correctly recall, John doesn't view balanced interfaces as an optimum solution for stopping common-mode noise transfer because parasitic capacitance causes the CMRR of such interfaces to typically degrade as noise frequency increases. There is high frequency noise present on the ground of fast digital circuits, especially in PC based servers, even with the utilization of ground planes. Gound planes provide the best practical solution, yet even they are not without some impedance. Of course, it's debateable as to whether ground plane noise exists at a significant level, but I don't think it debateable that it exist. All of John's concern over minimizing ground noise transfer between boxes is due to his view that such noise introduces clock jitter within the DAC, a view I share based on my own basement experiments.
Whether one agrees or disagrees with John's technical views, he's no clown.
Last edited:
If that's indeed what he says, there's some serious misunderstanding, whether on his part or those who take that stuff seriously.
I haven't found much interest in technical explanations among the faith-based audio set, especially when there's commercial interests mixed in. Good story-tellers have their audiences.
How DACs work, what actually affects what comes out of them, and how signals are reconstructed are all pretty basic and well-understood issues. That's not good if one is selling solutions to creatively-spun problems.
How DACs work, what actually affects what comes out of them, and how signals are reconstructed are all pretty basic and well-understood issues. That's not good if one is selling solutions to creatively-spun problems.
Never said he is a clown, and your interpretaties makers more sense. Not convinced that the problems you get with optical interface are worth the effort. The issues with galvanic connections are well understood and pretty much solved. Then again, I understand the need to distinguish yourself from the pack.
Jan
No one with some deeper insight feeds a DAC chip with a signal coming directly from any external interface, be it galvanic, opto, homing pigeons and what not. So it *should* be about moving bits unaltered. Then You care for the timing locally - here one can mess up also....
If one chooses to bring in a pice of plastic or a metallic cord into the DAC box is a matter of choice but I prefer the plastic. Yes, I know that the metal version is transformer coupled.
//
If one chooses to bring in a pice of plastic or a metallic cord into the DAC box is a matter of choice but I prefer the plastic. Yes, I know that the metal version is transformer coupled.
//
Last edited:
Thanks so much everyone for your input!
At this point, I can't resist saying "But it goes to Eleven..." Much of your response has gone beyond my level of understanding, but first I would like to ask a very basic question regarding your feedback so far;
The digital interference from ground issues and/or 'clock jitter'; how do these issues affect the analogue signal. My uneducated guess would be that it may be in the form of glitches or pops rather that a subtle introduction of harmonic distortion, background noise (maybe?) or frequency response changes
Please correct me, but unfortunately, my digital knowledge is at the very beginning of my learning curve!
Thanks again, all. Story of my life, I make trouble where ever I go!
At this point, I can't resist saying "But it goes to Eleven..." Much of your response has gone beyond my level of understanding, but first I would like to ask a very basic question regarding your feedback so far;
The digital interference from ground issues and/or 'clock jitter'; how do these issues affect the analogue signal. My uneducated guess would be that it may be in the form of glitches or pops rather that a subtle introduction of harmonic distortion, background noise (maybe?) or frequency response changes
Please correct me, but unfortunately, my digital knowledge is at the very beginning of my learning curve!
Thanks again, all. Story of my life, I make trouble where ever I go!
I didn't intend to imply that anyone was dismissing him as a clown exactly. I would better have wrote that, he should be taken seriously.
I too, am not convinced that optical is worth the effort. In Toslink's favor is the complete and wideband breaking of ground noise transfer between boxes. Toslink's primary drawback is that it tends to perform poorly on introducing interface jitter. John claims that he has designed a novel circuit that greatly suppresses transferred clock jitter.
While I have no way of assessing John's claim, as the circuit in question is proprietary and in a commercial product, I do know John to be a lateral thinker who comes up with rather creative solutions to design problems. If John's jitter suppression circuit performance claim is accurate, then Toslink would indeed provide the better commonly available source to DAC interface. Potentially, even better performing than asynchronous USB because, unless optically coupled, there is still a mechanism for source common-mode noise to corrupt the performance of a low jitter master clock generator that's local to the DAC via USB.
Last edited:
As long as the representation stays PCM, jitter can't be introduced. While in PCM, calculation rounding errors from changing Fs or bit depth will occur - but not jitter. Jitter induced sound quality degradation can't happen on a HD, on a CD, on CD drive, on a Toslink link etc.
"Jitter" is harmful at D/A conversions and is a property of the clock, not the PCM stream. When they meet, as they have to in the D/A process, the clock jitter has an impact.
Toslink is a perfect data conveyor and "isolator", but a poor clock transmitter. Use an other clock!
//
"Jitter" is harmful at D/A conversions and is a property of the clock, not the PCM stream. When they meet, as they have to in the D/A process, the clock jitter has an impact.
Toslink is a perfect data conveyor and "isolator", but a poor clock transmitter. Use an other clock!
//
The true clown in the room would like to once more interject;
Using my laptop as a 'server' if I am using the term correctly, how would clock jitter be ultimately manifested in the audio output signal?
Using my laptop as a 'server' if I am using the term correctly, how would clock jitter be ultimately manifested in the audio output signal?
Arta has a "jitter test" that will show you the extra noise introduced by jitter. So that is one type of noise. Whether it affects the sound in other ways I don't know.
I posted about this before and have this spectrum for my DAC: http://www.diyaudio.com/forums/digi...-software-measuring-jitter-4.html#post4469416
I posted about this before and have this spectrum for my DAC: http://www.diyaudio.com/forums/digi...-software-measuring-jitter-4.html#post4469416
Yes, a J-test comparison broke out on another thread (I don't remember which), but nearly everyone's DAC sailed through it, including the cheapest one I have in house (Focusrite Scarlett 2i2).
One more very basic question;
I was forced to go to wiki to interpret some of the information I have received here. There it was mentioned that SOME CD's use high frequency pre-emphasis to increase dynamic range, which is automatically detected by the CD player and de-emphasized.
Is this being overlooked by my external sound card? Was this something that should have been addressed when the flac file was created?
I was forced to go to wiki to interpret some of the information I have received here. There it was mentioned that SOME CD's use high frequency pre-emphasis to increase dynamic range, which is automatically detected by the CD player and de-emphasized.
Is this being overlooked by my external sound card? Was this something that should have been addressed when the flac file was created?
- Status
- Not open for further replies.