Agree, but since we are talking about reclocking USB signal, packet clock jitter is the only thing such a reclocker can potentially fix, not sample jitter.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Review and Measurements of Gustard DAC-X26
- Thread starter amirm
- Start date
Agree, but since we are talking about reclocking USB signal, packet clock jitter is the only thing such a reclocker can potentially fix, not sample jitter.
True. Although I cannot possibly imagine that would be audible. I also don't really think you can fix it with any kind of "USB packet reclocker" - I believe the only way is to change the design so the analog section is less influenced by these kind of digital noise.
(SPDIF) Jitter is not noise and it is not related to analog world. As such it is a time domain issue but of a digital signal - it has nothing to do with time domain of analog signal (phase).
In digital communication jitter is simply a variation in clocking of synchronous data transfer. If the data contained audio signals jitter will manifest itself as a distortion once data is converted into analog form.
Jitter exists in all forms of synchronous data transfer and not all data is audio.
Last edited:
Not sure why there needs to be a distinction since frequency domain is the same as the time domain, over a transformation. Maybe just a question of terminology?
I’ve always thought of noise as some unwanted signal. It can be correlated to the “wanted” signal, uncorrelated, or purely random. It’s noise regardless of what domain you measure it in. At least that’s my definition.
But, if you meant to write phase noise then yes, that's measured in the frequency domain.
Last edited:
Phase noise is a term related to analog signal, not digital.
Jitter we are talking about is SPDIF jitter coming from timing varations during synchronous data transfer. In analog signal it will manifest itself as a distortion (change of waveform).
USB packet noise is not jitter, it is coming from USB control pulses. From analog signal perspective it is a noise being injected in analog signal.
All valid, but I think this discussion is jumping too much between USB, SPDIF, reclocking and noise in digital and analog. All of these are separate topics, and no generalization cuts across all of them.
OP
- Thread Starter
- #209
Jitter can be anything it wants to be. It can be noise in which case it just raises the floor of the DAC, or sine wave, or square wave or in reality, combination of all of these.
I don't agree with that. Is there a link which supports such wide definition of jitter?
Newk Yuler
Active Member
- Joined
- Mar 5, 2019
- Messages
- 155
- Likes
- 252
My issue is addressing the evolution of USB receivers and implementation that makes a USB refining device (for S/PDIF or I2S) unnecessary. You all who will say that asynch USB has no issues are wrong in my experience because it was not a be-all solution when it was introduced as a port for DACs a decade or so years ago. I want to know if modern designs have become the be-all port some of you gentlemen say it is.
Whether or not USB async is the 'ultimate' solution isn't the question. Whether or not it needs external reclocking to improve sound quality -- is. And the answer is: it doesn't.
Newk Yuler
Active Member
- Joined
- Mar 5, 2019
- Messages
- 155
- Likes
- 252
Are you saying specifically that USB data has no use for external reclocking before it reaches an asynch USB receiver? I understand that USB packets are conveyed in an orderly manner and it's just data to be processed. I'm thinking of that which the USB receiver has produced and is sent to the DAC internally or run out through S/PDIF or somebody's configuration of external I2S (like HDMI).
I should be able to take an asynch USB receiver like this and it will perform as well as my Off Ramp 5? If not, why? I'm not trying to be an ass. I want to learn.
A discussion on this and similar issues deserves a unique thread. ASR is a community with a ton of potential for schooling. Seems a new corner of the forum could be created for schooling, although there would probably be just as much arguing there. I almost feel bad or guilty of something bad for asking about this sort of thing (above). I realize it can get complicated and it's probably a task to dumb it down and have it make sense to less technical folks. I have some idea that technically proficient members are here because they're technically proficient and seeking empirical data and not to try to explain technicalities to amateurs. Although many like to argue among themselves anyway. I just wish these guys would admit they enjoy arguing with each other.
In simple terms the receiving end of an async connection controls the rate which the data is sent to it so that there is no buffer over/under flow on the receiver, the data from this buffer is then sent further inside the DAC clocked by its internal stable clock. This is possible because, unlike the SPDIF, USB is a bi-directional protocol and sender and receiver can "talk" and agree on data transmission terms.
OP
- Thread Starter
- #215
What? That is like asking me to give you a link for 1+1 = 2. How much have you studied this to challenge me this way? Even the Wiki tells you this: https://en.wikipedia.org/wiki/Jitter
jhwalker
Active Member
Don't feel bad about asking questions At least you're asking. Most don't want to or are afraid to, instead accepting an answer that is often incorrect, made up, or misleading.
Timing of samples can be an issue with SPDIF (and HDMI audio) protocols because, as others had said, the source is driving the timing of samples output by the DAC. This means the clock signal must be properly transmitted from the source to the DAC, and that's often where jitter occurs. The timing of USB isochronous protocol has no such issue: the timing of sample output is controlled by the DAC and its internal clock. Eliminating jitter in a working USB transmission does not accomplish anything useful, other than maybe making some of the USB eye diagrams look better.
OP
- Thread Starter
- #218
FYI I posted the measurements of the PLL in this DAC in another review: https://www.audiosciencereview.com/...rements-of-empirical-audio-synchro-mesh.7062/
- Joined
- Feb 25, 2019
- Messages
- 33
- Likes
- 52
Yes the other thing to consider is that if you are looking at headphone amp capabilities,the RME ADI-2 DAC has that built in, which the Gustard does not... I also would be willing to bet on a better stateside and euro side warranty and repair infrastructure for RME at this time.
Similar threads
- Poll
