Those are the three phases to ideas that are correct. Most ideas aren't correct.
Merry Xmas, John. I wish I were there today, I'd be a lot warmer.
Merry Xmas, John. I wish I were there today, I'd be a lot warmer.
scott wurcer said:
It should be but it will still be interesting to do a test with different cables.
scott wurcer said:
Well, my DAC is what it is (or, rather, my future DAC will be what it will be). What I can do is see whether different SPDIF coaxial cables make a difference in sound quality and which one will give the best sound.
The fundamental problem with jitter was addressed over a decade ago, when we (the serious audio researchers) looked for the problems of CD reproduction. The major problem was the embedded carrier and having to pull it out of the data stream with a phase lock loop. If you study it, it is amazing it works in the first place. NOW, everyone can quote FIFO and RECLOCKING are relatively NEW (hi end) ideas, when "Perfect sound forever" was found to be wanting by the 'golden ears'. I remember talking about it to a digital engineer in Silicon Valley more that 15 years ago, before it was brought to the surface. NOW, it's obvious.
John, the understanding I have is that the problem first really come to the attention of the "industry" in duplication plants. What happened is that some transfers were found to sound like dogpoopy... the problem was eventually traced to a jitter problem in the reclocking... probably the complaints that triggered it did come from "consumers" who were listening to the CDs and noticed rather significant differences between different "pressings" of the same material.
There is/was a detailed article on this event in one of the trades many years back... and I think the original problem(s) were closer to 20+ years ago now...
_-_-bear
alansawyer, the effect of different cables vs. SPDIF comes at least in part with capacitance/inductance that alters the shape of the waveform, and the impedance/termination that causes changes to the waveform shape and introduces reflections that occur as signals adding or subtracting at the "wrong time". Keep in mind that the thing works on transitions between low and high, not on absolute "level", as in reaching some specific logic voltage like +5v or " 0v "... so you can see that under this method, the slope of the waveform affects "timing"...
There is/was a detailed article on this event in one of the trades many years back... and I think the original problem(s) were closer to 20+ years ago now...
_-_-bear
alansawyer, the effect of different cables vs. SPDIF comes at least in part with capacitance/inductance that alters the shape of the waveform, and the impedance/termination that causes changes to the waveform shape and introduces reflections that occur as signals adding or subtracting at the "wrong time". Keep in mind that the thing works on transitions between low and high, not on absolute "level", as in reaching some specific logic voltage like +5v or " 0v "... so you can see that under this method, the slope of the waveform affects "timing"...
Joshua_G said:
The good thing with digital signals
is that they are a lot more easy to test than analog signals.
And to test the accuracy of digital Signals we should not resort to subjective listening tests.
Today there are so very good Digital storage oscilloscopes / Logic analysors.
If data in = data out, then there is no problem.
And if Scott Wurcer ( or elso Kwak ) suggest a better Data (re-)clocking
then we can use this for remove the jitter-factor.
Sorry, Listening tests to determine digital quality seems a bit stupid, to me 🙂
As there are better ways.
I am quite sure there are test reports that have measured digital data qualities
in a proper way
with different setups. Cables, connectors, DAC and Clocks.
For example
now when we surf internet, is often used verification of the data accuracy sent/received.
We have downloaded sometimes files, that has got an extra small file.
This is the checksum file.
If the checksum is the same for data file then the data is 100% same as the originally sent data.
In a data streams, often is used some form of DATA VERIFY.
This can be some extra bits.
If the data is corrupt, for some reason, it will be sent again.
http://en.wikipedia.org/wiki/Checksum
now when we surf internet, is often used verification of the data accuracy sent/received.
We have downloaded sometimes files, that has got an extra small file.
This is the checksum file.
If the checksum is the same for data file then the data is 100% same as the originally sent data.
In a data streams, often is used some form of DATA VERIFY.
This can be some extra bits.
If the data is corrupt, for some reason, it will be sent again.
http://en.wikipedia.org/wiki/Checksum
John, it is common practice to embed clock in data for serial data communication of many types, including professional AES/EBU systems. The problems are not unique to SP/DIF at all, and have been studied by many of us over many years.
If you study it, you come to understand how it works. As you are an engineer I would have expected you to understand the operation rather than be amazed by it.
lineup said:
Well, I neither own Digital storage oscilloscope / Logic analyzer, nor do I have access to one – they are too expansive for me to own and I'm retired, so I don't have access.
What I do have is a pair of ears, which so far are serving me well enough. As strange as my ears may seem in the context of digital data stream, so far they proved to be the best measuring tool …
Also, better think of the data stream as a high frequency square wave, rather than series on ones and zeroes.
Anyhow, I'm interested in what works better in my setup – not in research for its' own sake.
Hello Joshua.
Sorry all Golden Ears.
... but
Listening tests 😱 to determine digital cable quality seems a bit stupid, to me.
With these cables, either they
- are good enough to transfer digital data correct
- or they are not suited for this
... and most all 'digital cables' sold will do their job correctly
Sorry all Golden Ears.
... but
Listening tests 😱 to determine digital cable quality seems a bit stupid, to me.
With these cables, either they
- are good enough to transfer digital data correct
- or they are not suited for this
... and most all 'digital cables' sold will do their job correctly
lineup said:
Need to look at the miles of Cat5 running through buildings for 1Mbps+ LAN. Doesn't get much more rudimentary.
You can not compare a SPDIF signal to a digital signal, except perhaps as Scott mentioned when you use FIFO and reclocking. The problem is not when you keep data in the digital domain, only when you convert to analogue.
A digital storage scope would not be of much help in that.
A digital storage scope would not be of much help in that.
This is like a group of teenagers, when we talk about digital. Lots of opinion, without substance. About 15 years ago, I went to a listening session in this area, precisely concerning digital cables. We could hear every one of them. It was easy. That is just a fact, to me.
Sophomoric opinions as to how this is 'impossible' just confuse the issue.
There has been a lot of study on how to make digital better with those who really care, over the decades. One should get up to date with it, IF you are going to present an opinion.
And Bear, you are right, it was about 20 years ago when I discussed the jitter solution with that silicon valley engineer. Time flies, but it was relatively unknown to the CD world just then.
Sophomoric opinions as to how this is 'impossible' just confuse the issue.
There has been a lot of study on how to make digital better with those who really care, over the decades. One should get up to date with it, IF you are going to present an opinion.
And Bear, you are right, it was about 20 years ago when I discussed the jitter solution with that silicon valley engineer. Time flies, but it was relatively unknown to the CD world just then.
I don't do what may seem as clever, or stupid, to others.
I do whatever may produce better SQ from my existing setup.
Once I'll have a good DAC, I'll test various SPDIF cables and see if they make a difference. If they do, I'll choose the one that will sound best. If they don't, I'll choose the one that costs less.
Yes, I read this theory many times.
I read also the theory that well engineered analog interconnects cannot make a difference.
However, people whose hearing I trust reported differences between SPDIF cables. So, I'm going to test it on my setup.
Andre, SP/DIF IS a digital signal whatever you care to think. At the low level, just as with the signal sent over UTP CAT5 cables the pulses sent can be viewed as a voltage at a time and these pulses have rise times so ONLY in that respect the digital pulses themsels have an analogue, but the signal being sent IS digital because it is a sequense of pulses representing digital o's and 1's with their embedded clock. There are standards for this and for all the other digital communication protocols in use, and if you need to read them a quick use of google will find most of them.
Even Wikipedia might help you
Note I think you referred to SP/DIF previoulsy as different than optical, but you will see that copper and optical are just options for the physical layer.
Depends on shape of rising and falling edge and on rise time and fall time as well. Everyone who ever adjusted CD laser knows.
alansawyer said:
Alan, I'm very aware of the SPDIF format and its origin.
If you care to read my previous posts again, I believe I have explained thoroughly enough for anyone who wants to understand, why I believe the format is flawed for use between a transport and DAC.
Regarding optical and coax feeds, with optical there are two more conversions added in the transfer of the 'digital' signal, just adding even more problems. It is generally accepted by most that coax will sound better than optical.
- Status
- Not open for further replies.