HDMI vs. Coaxial Digital Interconnects

fas42

Addicted To Best
Jan 8, 2011
3,973
3
0
NSW Australia
We can pray that good and precise timing was used in creation of music and that which lives in our digital format represents that. Our job then is to match that timing as closely as we can.
If this comes across a little aggressively forgive me, I'm in that sort of mood ...:)

Getting rid of jitter is all about using an electrical equivalent of a massive flywheel; I think everyone understands how that works! The flywheel spins ALMOST at a perfect rate even if, say, the engine driving it is wobbling, or trying to, all over the place. That is what PLL, phase locked loops, which are used most of the time are all about. The more massive the flywheel, the better the PLL implementation, the less jitter presented to the DAC; anywhere else, jitter could be occurring at a ferocious level and it don't matter!

There are a number of ways of creating massive electronic flywheels, but ultimately it doesn't matter WHAT you use, it is how WELL it's been done. Naim, for example, in its latest DAC uses a number of completely independent flywheels running at different rates, and it picks the right flywheel to use on the fly -- a perfectly good way of doing things!

So, excellent or good enough jitter we now have. Next problem, is the digital input causing interference to the D/A conversion. For people to get a better handle on this, this is just another version of crosstalk! Everyone sees measured figures of crosstalk starting to drop off at 20kHz, just imagine the crosstalk then if there's a noise signal running around inside the case at 20MHz! And the poor old DAC is supposed to ignore this, I have seen pretty dreadful figures of PSRR, power supply noise rejection, for DAC's, no wonder they spew out bad sound at times!

Okay, 'nuff said ...

Frank
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
If I remember correctly this 1.5 m had everything to do with the rate of the SPDIF signal.
If this is correct, you need a different SPDIF cable for each sample rate…..
The dependency is not on the frequency but rise time of the signal. There is an excellent write up on this article I post a while back in another thread: http://www.positive-feedback.com/Issue14/spdif.htm

"If the rise-time is 25 nanoseconds and the cable length is 3 feet, then the propagation time is about 6 nanoseconds. Once the transition has arrived at the receiver, the reflection propagates back to the driver (6 nanoseconds) and then the driver reflects this back to the receiver (6 nanoseconds) = 12 nanoseconds. So, as seen at the receiver, 12 nanoseconds after the 25 nanosecond transition started, we have a reflection superimposing on the transition. This is right about the time that the receiver will try to sample the transition, right around 0 volts DC. Not good. Now if the cable had been 1.5 meters, the reflection would have arrived 18 nanoseconds after the 25 nanosecond transition started at the receiver. This is much better because the receiver has likely already sampled the transition by this time."

Earlier in the article he talks about where the 25 ns came from.

Anyway, an interesting discussion about this subject between a couple of engineers:
http://recforums.prosoundweb.com/index.php/mv/msg/11678/0/32/0/
I have had a few of these :). The more hardcore an engineer, the more they are shocked that there is a real logic here. It usually takes me a dozen back and forth to explain that this is not voodoo but real engineering principals of how data can arrive incorrect even though it is digital.
 

RBFC

WBF Founding Member
Apr 20, 2010
5,158
45
1,225
Albuquerque, NM
www.fightingconcepts.com
If I remember correctly this 1.5 m had everything to do with the rate of the SPDIF signal.
If this is correct, you need a different SPDIF cable for each sample rate…..

Anyway, an interesting discussion about this subject between a couple of engineers:
http://recforums.prosoundweb.com/index.php/mv/msg/11678/0/32/0/

If you go back to page 1 of that link, Dan Lavry states that shorter S/PDIF cables work better. Longer cables work better for the salesman.

So, what's the truth? I'm going to borrow a 2 meter digital cable from my dealer and compare it.

Lee
 

amirm

Banned
Apr 2, 2010
15,813
37
0
Seattle, WA
I read Dan's response. It says nothing other than it is not true and go read about skin effect on my web site.

The recommendation I mentioned came from one of the principals at Berkeley Audio and came unprompted as we were talking about merits of various connections on their DAC. I actually didn't realize that the article I linked to had the same length specified until I went to search for it now. Since Berkeley people don't sell cable, and have more knowledge about DACs than 90% of the world out there, I tend to believe it especially since it makes sense to me.

Now, if Dan really wanted to challenge the guy, he could have said that the formula is only true if the rise time is as slow as assumed. Unfortunately, without being able to measure it, it is hard to say. Maybe when I have some time to kill, I will fire up my scope and measure it and report back.

Finally, let me say that is very counterintuitive to read that the shortest cable is not right.
 

mep

Member Sponsor & WBF Founding Member
Apr 20, 2010
9,481
17
0
$500 beans for a 1 meter digital cable is "not very expensive at all"? Perspective Steve, perspective.
 

DonH50

Member Sponsor & WBF Technical Expert
Jun 22, 2010
3,946
305
1,670
Monument, CO
Note that if the receiver was perfectly, matched cable length would not matter (within reason). If the source was perfectly matched, or a perfect voltage source, cable length would not matter, either. Barring a perfect world, anything's possible. For most systems significant mismatch is tolerated, and in fact in this case that's true for the data stream itself. The problem is jitter on the clock recovered from the data stream that ends up driving the DAC.

This is another example where you only want just enough bandwidth and no more...
 

DonH50

Member Sponsor & WBF Technical Expert
Jun 22, 2010
3,946
305
1,670
Monument, CO
As an aside, one of the reasons I chose my AVR (Pioneer SC-27) was its very low jitter on coax and HDMI (almost the same and both very low). Some of the other AVRs had high jitter on coax, though lower than on HDMI, and very high (up to 10's of ns) on HDMI. As a matter of fact, my surprise at the high numbers, and number-crunching to prove its relevance on another forum, is how Amir and I "met"! :)
 

Vincent Kars

WBF Technical Expert: Computer Audio
Jul 1, 2010
860
1
0
Very interesting cable, Vincent. Do you use it and how did you find it different from other cables? It is well made and not very expensive at all

As AmirM was referring to an article by Steve Nugent, I pointed to a cable manufactured by him.
As I do think the article in Positive Feedback is a nice advertorial I expected the joke to be obvious. Obvious not.
As far as cable length and impedance matching is concerned, I love the quarter wave length approach: http://www.thewelltemperedcomputer.com/KB/DigitalCable.htm
 

RBFC

WBF Founding Member
Apr 20, 2010
5,158
45
1,225
Albuquerque, NM
www.fightingconcepts.com
Vincent,

Many of us are not Electrical or Audio Engineering types. The transition from all-analog music systems to digital systems, especially computer-based, has not always been clearly explained. Please keep this in mind when you post "answers" to questions on the forum. This modern age can be confusing with all the conflicting information offered around the web.

Lee
 

Orb

New Member
Sep 8, 2010
3,010
2
0
As an aside, one of the reasons I chose my AVR (Pioneer SC-27) was its very low jitter on coax and HDMI (almost the same and both very low). Some of the other AVRs had high jitter on coax, though lower than on HDMI, and very high (up to 10's of ns) on HDMI. As a matter of fact, my surprise at the high numbers, and number-crunching to prove its relevance on another forum, is how Amir and I "met"! :)

Shame the latest Pioneers have gone backwards with their implementation of 1.4
Paul Miller has some shocking measurements for the latest Pioneer AVR, even with their PQLS (or whatever it is called) engaged, it is far worse than the older model with 1.3 support not using PQLS.

Don, to help some appreciate timing-singal-jitter (where some may feel timing is just timing); are you aware of any online papers that show the framing format-specification and how it operates for Redbook CD?

I do not think me sharing G.703 stuff,etc I have is going to help hehe.

Thanks
Orb
 

RBFC

WBF Founding Member
Apr 20, 2010
5,158
45
1,225
Albuquerque, NM
www.fightingconcepts.com

Much of it does make sense to me now, but mostly because I've started to spend some time wrapping my head around the issues of setting up a computer-based music player. In some of the sentences there, you throw out a term that most folks will not understand (i.e. "it's an ARM processor", etc.) because it's not second-nature to us.

Now, I'm not a complete caveman, as I can easily download music files or rip CDs and play back redbook or mp3 material out of my computer speakers. I'm trying to wade through the configuration options to play standard AND high-resolution files from a new computer in my music room into my Krell preamp.

The frustrating comments I've gotten from other places on the web usually sound like Just get "X" and you'll be good to go. These comments literally make me want to jump through the computer screen and show the author what I do as my second job.

I've witnessed first-hand the intricacies and problems associated with getting RELIABLE playback of media with different native resolutions.

I have not yet seen anything resembling a step-by-step guide that:

1.walks users through the process of purchasing the correct components (based on their choice of PC vs. Mac) and discusses options based on future expansion, etc.

2. connecting it all successfully: what is needed and various options to get the job done

3. clearly understanding how to do the setup in MIDIs, player program preferences, etc. (typically we see one static screenshot of a preference panel)

4. choosing and operating music player software, while understanding how to negotiate different sampling rates on playback

5/ understanding how to store and catalog your collection, and how different storage options must be configured

6. integrating the internet for metadata retrieval and access to music download stores, etc.

7. steps to take when certain, archetypal glitches and problems surface

If someone would clearly go through all this information, that would be a huge success on the internet. Folks (yourself included, no offense intended) start out great and soon begin to take for granted that folks understand something that is simple TO YOU. Then, certain steps, obvious to experienced users, are omitted. That ends the chance of successful setup for many of us, where we are brickwalled by a missing step, etc.

I hope this explanation helps you to understand how those who are just getting into this area are struggling with the vast amount of conflicting information and incomplete advice. If I told you how to change your oil in your car, but neglected to tell you how to use an oil filter wrench or even what one was...., you can see the situation more clearly.

Thanks,

Lee
 

DonH50

Member Sponsor & WBF Technical Expert
Jun 22, 2010
3,946
305
1,670
Monument, CO
@Vincent -- The 1/4-wave approach is the one I usually use as well, with the caveats that (a) it relates to the highest frequency in the signal so edge speed matters more than the clock/bit/whatever frequency, and (b) in a cable the velocity of propagation is only about 70% of c. This does remind me that I was going to try to put something about this in a thread here, with some pretty pictures... Not enough time! Maybe we should start a Techie Speak Forum. ;)

@Orb -- No, I have not really looked. While sampling has been my life, I have not actually done a lot of research into audio systems -- my work has been mostly up in the GHz region. However, since the problem is the jitter on the recovered clock, I am not sure anything else really matters.

I think my earlier threads on terms, jitter and such might help. I'll take a step at helping out; apologies if I am too high/low a level, and bear in mind I am not trying to be scientifically rigorous.

Signals have frequency (how many times a second they go through a complete cycle), velocity (how fast they travel -- near light speed in vacuum, somewhat less in air, and only perhaps 7/10 that speed in a wire), and wavelength (the distance it takes for a complete cycle whilst traveling through a medium, be it vacuum or wood). The wavelength is the velocity divided by the frequency, w = v / f.

The problem is that if the source (DAC output), load (preamp input), and cable are not all the same impedance ("resistance"), part of the signal will "bounce" when it hits. Think of trying to connect three slightly different sized water hoses, offset just a little, so whenever the water passes from one to the next some splashes out due to the mismatches. (OK, not a great a analogy, but it's early, and 22 below out!) In a cable, instead of splashing out, the mismatched part of the signal reflects and heads back to the load. A key thing to know is that, when it reflects, it inverts ("turns around" so it is 180 degrees, or out of phase, with the original). So, when the reflected part gets back to the source, it cancels some of the out-coming signal as it hits. That "moves" (varies the amplitude of) the outgoing signal a bit and causes jitter.

When the cable is 1/4 wavelength and the mismatch is complete (e.g. load is very badly matched), the end-to-end (round-trip) travel time matches 1/2 wavelength. For a steady-state wave at just the right frequency, or travel time short enough, that means the signal peak hits the load, inverts, and comes back just in time to completely cancel the outgoing wave. This is actually used to make filters in the microwave world. In the audio world, it messes with the signal by changing the amplitude, and (as seen in my jitter threads) that causes jitter on the recovered clock signal.

Probably too deep for some and way too loose for others, but such is trying to reach a wide audience.

HTH - Don
 

RBFC

WBF Founding Member
Apr 20, 2010
5,158
45
1,225
Albuquerque, NM
www.fightingconcepts.com
Interesting and valid points.
I do see a task of Herculean proportions ahead.
What about (as we are horribly of topic) to move this to http://www.whatsbestforum.com/showth...-out-of-the-PC?

Thanks, Vincent. Yes, it would be a tall order to fill. Of course, I understand that the landscape of computer music and the attendant interfaces, etc. is changing day-to-day. I'll address any further thoughts in the other thread or a new one.

Lee
 

Orb

New Member
Sep 8, 2010
3,010
2
0
Don,
yeah.
I am trying to find information that makes it pretty clear for some to appreciate how complex the subject is when one considers that the 1s and 0s must co-incide with signal/coding format that also includes timing information or structure, which is impacted exactly by what you and Amir are saying.
The simplest I can find is what I am going to link; but I get the feeling it will only make sense to those of us that are telecom engineer or other engineers used to data/audio transmissions.
Still it provides a wealth of info that others can use to do further searches if they want to try to get to grips with understanding how 1s and 0s are part of a very complex structure/framing for transmitting digital audio.
The key part to look at is the Bitr rate and then Coding Format section further down the page:
http://www.hardwarebook.info/S/PDIF

And for those who may not know just bear in mind that the above is not the same as what is done for the audio standards applied to reading CDs by a mech unit or transport and defining 16bit PCM coding at sampling rate 44.1khz.

Thanks
Orb
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing