There was a simple question how digital audio could be influenced.
As simple as that.
Cheers!
Do solid state amplifiers sound the same ?
- Thread starter BS5711
- Start date
As simple as that.
Cheers!
bigshot
Headphoneus Supremus
You’ve got that backwards. The studies were intended to establish the thresholds of audibility of jitter. They weren’t studies based on consumer equipment. Jitter levels in consumer equipment is measured all the time, and I’ve never seen any measurement that came close to being audible. As I said before. Measured jitter in home audio components generally falls far below the established thresholds… usually by an order of magnitude. If you know of some product with audible levels of jitter, let me know.
If you’d like to do some research on your own instead of arguing without having any knowledge, here is somewhere to start.
Just Detectable Threshold in Music 20ns, perhaps higher than 200ns (ref. the work of Dr. Ashihara)
http://www.aes.org/e-lib/browse.cfm?elib=8354 (needs subscription)
http://www.nanophon.com/audio/1394_sampling_jitter.pdf (cited in section 2.2)
Last edited:
Its not broken, but its, as mentioned in the measurement, not good either.
Its not even flat from 20Hz to 20kHz, for an product of that price, that is absolutely unacceptable.
I am probably annoying the hell out of you, but look how my 50$ dongle performs ruler flat beyond 40kHz with white noise. It has an rolloff of ±1db at 82kHz. This thing has an rolloff of ~4db at 20kHz
How can an product of that price point fail at this? And this is an audible difference you will be able to hear. So this DAC is not transparent, it is not capable in reproducing the original sound.
Of course that is unrelated to the Jitter and the timing issue but this device seems to be not good in general. Its not broken, no, its just not good.
Thanks for that link!
Lots of interesting details. Had a brief 10 minutes read and will continue on the rainy Sunday...
Thanks again!
This is a valid point of view
No, not at all.
And I also don't want to annoy people over here in the science forum...
... but my very personal, non-ABX-tested
history in high(er) fidelity audio during the last 3 years has brought me quite some experiences.And, after getting a little into the topic, reading many articles, talking to people, listening to dozends of systems at dealers, friends, getting to know also THEIR experiences:
There's a lot to be discovered if one is open-minded.
4 years ago I thought "bits are bits", and that probably the cheapest available streamer would be good enough to feed my DAC. Because its input is buffered and reclocked anyway... blablabla.
Long story short: Boy was I wrong.
By the rules of this forum (which I do respect contrary to common belief of some) there is simply an enormous empirical evidence that in real-world applications there's a lot of "funny" effects happening in digital audio.
Two big contributing factors are noise and jitter (somehow interrelated) on the digital signal line. And no, also USB is not bullet-proof, even with non-defective devices.
It’s been known for a very long time, back in the 1950’s in fact and actively attenuated since around 1952. And incidentally, you’ve effectively cited marketing from 30 years ago, not long after the falsehood of jitter being an audible problem in consumer DACs was first invented by audiophile marketers.
It is of course impossible to have perfect timing, we can measure timing down to thousandth’s of a trillionth of a second, so at that level we can always measure some timing difference in pretty much anything, even in the speed of light over relatively short distances. So it should be obvious that is not the issue, the actual issue is, as it always is; does this measured difference even have any effect on the sound being reproduced and only if the answer is “yes” can we even ask the question; does it reach the threshold of audibility?
Again, jitter has been known about and researched for over 70 years. The first study published publicly that I’m aware of was done by the BBC and published in 1974, half a century ago. Its conclusion that jitter below about 200ns (billionths of a second) is inaudible with music recordings has been confirmed by numerous subsequent studies. To my knowledge, there has never been any consumer DACs with jitter anywhere near that bad, even going back to the first generation of CD players 40 years ago. By about 35 years ago, the cheapest, poorest quality DACs available on the market had jitter less than 2ns and by the mid 1990’s, even cheap OEM CD drives typically had jitter in the region of 200ps (trillionths of a second), so around 1,000 times below audibility!
Can you honestly “not rule out the possibility” that an expensive audiophile or even an equally cheap DAC today performs more than 1,000 times worse than a cheap DAC from 30 years ago?
Yes, again we can measure jitter down to stupendously tiny levels and the example you’ve cited is a good one. The jitter artefacts are around -140dB to -160dB and to put that into the perspective which you don’t appear to appreciate, that’s roughly 33 to over 300 times (respectively) below the threshold that sound can even exist in air (assuming a peak playback level of 85dBSPL). So obviously, any question of audibility is moot, because regardless of the hearing threshold you cannot hear a sound that cannot be reproduced and therefore does not exist!
G
Last edited:
castleofargh
Sound Science Forum Moderator
- Joined
- Jul 2, 2011
- Posts
- 10,980
- Likes
- 6,826
Why jitter and not literary anything else? Everything influences anything else, pushes some molecules, grabs one electron, then the dog pukes on the carpet(the butter knife effect or something, I'm not a meteorologist).
What sets the limit for legitimate concern?
Is it how much a brand advertised the problem it claims to fix?
Is it really set by how audible the impact is, when almost no owner will verify either audibility or causality before deciding to know more than enough anyway(the wonders of heuristics)?
Is there like an Anna Wintour of audio that declares something problematic and gets everybody frantically running about to fix it because now is the year of jitter, again?
I'm genuinely curious. Should I sacrifice a goat for better soundstage? What's the best time to do it? What criteria will firmly have you decide that entertaining the idea isn't being open-minded any longer, and that you have officially crossed into the twilight zone?
Still an amplifier thread, but who cares.
Hm... noise and jitter. But from a technical point of view: Good question! What else (besides the bad "j"-thing) could detoriate a DAC's sound quality.
Naa... spare the poor creature's life! There's better ways to improve soundstage.
But the example of jitter you gave could not “deteriorate” or in anyway affect a DAC’s sound quality, not unless you change the laws of physics!
G
bigshot
Headphoneus Supremus
Thanks for that link!
Lots of interesting details. Had a brief 10 minutes read and will continue on the rainy Sunday...
Thanks again!
These things aren’t hard to figure out with some focused googling. I really shouldn’t be expected to do your research for you. You could have found that out yourself with web searches. You just have to be careful that you’re not cherry picking to suit your pre decided agenda, and get information from trusted sources, which means NOT equipment manufacturers who are trying to sell you something.
Most of all LISTEN to people like Gregorio who know more than you do. Every one of his posts provides solid bread crumbs that you can follow. If you argue with him, you waste a terrific source of information. Feel free to disagree with him, but keep it to yourself until you’ve done the work to research it so you know the topic at least as well as he does. You haven’t been good at this up to now.
Last edited:
bigshot
Headphoneus Supremus
You have absolutely no evidence to back this up except other people’s misinformation and your own guesses. This is a gold plated example of expressing things as fact that you have no clue about. Don’t express opinions like this until you’ve done the research to know what jitter is, what the threshold of audibility is, the typical measured jitter in home audio equipment and how digital audio and USB works. People who express opinions about things they don’t understand run the risk of looking like they’re talking out their ass to people who know the subject inside out ( like Gregorio). You don’t want that because it can shut off a valuable source of information.
Basically there is only one thing an DAC can do wrong, not outputting the sound wave we want.
And sound waves have two properties, an frequency and an volume. So the DAC is doing things wrong when he is playing the wrong frequency/an frequency we did not request (Harmonic Distortion, Noise, Jitter) or when he is playing it at the wrong volume.
That is why the most important tests are 1kHz (To see the harmonic distortion and the SNR), 12kHz (To see Jitter) and Multitone (to see how it behaves, when we play multiple frequencies at once and playing an sine sweep + white noise (to check if these are played at the volume we request).
You can add more tests like comparing the input/output of white noise and/or music and stuff like that. You can test for group delay, phase issues, clock drift, gain drift and so on.
But when 1kHz, 12kHz, Multitone and the sine sweep+white noise look all fine and without issues, it is usually very unlikely to find big errors in any other aspect. Especially when the DAC is from ESS/AKM/Cirruss, they seem to know what they do and produce excellent DACs. Just give these DACs what the maker tells what they need to perform best and they will do exactly that.
The best thing a DAC can do is playing the sound wave we want, undistorted, at the correct volume. And if he is making an error doing that, the most important thing is, that the error is small enough, that nobody is ever able to hear it.
Last edited:
The tl;dr Version^^
VNandor
500+ Head-Fier
- Joined
- Oct 17, 2014
- Posts
- 956
- Likes
- 582
Not just on an abstract level. You cut out the most important part of my message which is that the timing of the conversion is controlled by the DAC's clock and not by the arrival of the data packet received from the USB! Obviously the clock will have some jitter so the timing of the conversion will also have some jitter. This is completely separate from how and when the data arrives to the DAC's buffer. Again, the DAC will not wait for the samples to arrive to the buffer, it will convert the next sample on the (typically) next rising edge of its clock signal. The measurement shows the jitter caused by the clock and not the jitter caused by the asynchronous data transfer.
The review you apparently did not even read mentions that the performance of the DAC stays the same even when the source is under heavy load, which definitely does add plenty of jitter and noise to the USB data sent out!
Users who are viewing this thread
