[art]

Ethan,

I can measure changes in jitter with different cables. Period.

I have shown the levels that can be expected with SPDIF fall into "your" 2 nSec criteria.

I am trying to be civil, but I feel as if I am talking to myself. And maybe one or two other folks.

So, unless someone has a direct question to ask me, I see no point in going on. I have spent far too much time as it is. I have a product to finish developing. I feel that my time is best spent there.

Pat

[Geardaddy]

Pat, thanks for your time and energy in debating this matter....

[opaqueice]

I know it angers - or at least irritates - some people to be told they can't possibly be hearing a change when they swap digital cables etc.

As it should.

Really?  I think it's a bit of a shame you feel that way. 

If you can actually hear differences it's pretty easy to prove with a blind test, and then you could shut up the skeptics.  If you can't, and you just imagined then, so what?  What's the shame in that?  As long as you enjoy your system and your favorite digital cables, what difference does it make?

[Double Ugly]

I know it angers - or at least irritates - some people to be told they can't possibly be hearing a change when they swap digital cables etc.

As it should.

Really?  I think it's a bit of a shame you feel that way. 

If you can actually hear differences it's pretty easy to prove with a blind test, and then you could shut up the skeptics.  If you can't, and you just imagined then, so what?  What's the shame in that?  As long as you enjoy your system and your favorite digital cables, what difference does it make?

You assume too much.

The primary participants have determined this thread has run its course, and rightly so IMHO.  If you wish to begin a new thread in which you divulge your methods of determining my skepticism or lack thereof, please do so, but I'd prefer the remainder of this thread not morph into a new topic.

Thank you.

The Management

[Geardaddy]

I know it angers - or at least irritates - some people to be told they can't possibly be hearing a change when they swap digital cables etc.

As it should.

Really?  I think it's a bit of a shame you feel that way. 

If you can actually hear differences it's pretty easy to prove with a blind test, and then you could shut up the skeptics.  If you can't, and you just imagined then, so what?  What's the shame in that?  As long as you enjoy your system and your favorite digital cables, what difference does it make?

Agreed.  I am not offended by Ethan's or others sentiments in regards to this subject, and some form of experiment would be fun.  I may be "smoking crack" as one of my teenaged brother-in-laws likes to say, but I have no shame in being proven wrong.   Anyway, I am looking forward to a blind testing of myself, my wife and her punk brothers with Ethan's jitter files....   As far this thread is concerned, DO is right.  It feels as if it has run its course.  We need a Part II focused on tweaks and technology.  

[opaqueice]

You assume too much.

Sorry - I didn't realize I was assuming anything.

The primary participants have determined this thread has run its course, and rightly so IMHO.  If you wish to begin a new thread in which you divulge your methods of determining my skepticism or lack thereof, please do so, but I'd prefer the remainder of this thread not morph into a new topic.

You're the moderator - do as you think best.  I haven't participated in this thread until now, so I certainly don't have much perspective on it.

I'm looking forward to the results of Ethan's tests, even though I doubt they'll differ much from others done in the past by professional audio researchers.  In any case I for one learned something here:  I now know how to calculate the effects of jitter with a given spectrum on an audio signal.  It's much simpler than I had thought.

[Double Ugly]

I may be "smoking crack" as one of my teenaged brother-in-laws likes to say, but I have no shame in being proven wrong.  

Nor do I.  Happens at least a few times every day, and to my chagrin more often than not.

You assume too much.

Sorry - I didn't realize I was assuming anything.

Yes, I know.

[Geardaddy]

Okay, I know this thread was supposed to be dead, but I did find a reference regarding audibility of jitter:

Eric Benjamin and Benjamin Gannon, "Theoretical and Audible Effects of Jitter on Digital Audio Quality", Pre-print 4826 of the 105th AES Convention, San Francisco, September 1998.

This paper concluded that the threshold of audibility of jitter on normal music signals is around 20nS (that's 20,000pS - rather more than 71pS

For you geeks out there, this may be worth looking up....or not

[Geardaddy]

Here are Ethan's jitter files:  http://www.audiocircle.com/index.php?topic=52363.new#new

 

[Ethan Winer]

Eric Benjamin and Benjamin Gannon, "Theoretical and Audible Effects of Jitter on Digital Audio Quality", Pre-print 4826 of the 105th AES Convention, San Francisco, September 1998. This paper concluded that the threshold of audibility of jitter on normal music signals is around 20nS (that's 20,000pS - rather more than 71pS

Thanks for posting that link. I bought the paper yesterday and you are correct - it concluded that even with the cheapest DVD players they tested - and this was ten years ago in 1998 - the jitter was well below the threshold anyone was able to hear. I also found a similar 2005 paper from another technical journal that concluded the same based on extensive testing.

This is not to say that people don't hear differences when swapping digital cables or A/D/A converters and word clocks. But if they do hear a difference, it's not due to reduced jitter.

Again, I agree with Pat and Bruno Putzeys and all the other competent EE's out there that aim to make their specs as good as possible. To do otherwise is just lazy, unless it can substantially reduce cost to the consumer while still preserving "good enough not to matter" fidelity.

--Ethan

[miklorsmith]

What's the point of good specs if you can't hear the benefit?  It surely is cheaper to "not sweat it".

[Geardaddy]

Eric Benjamin and Benjamin Gannon, "Theoretical and Audible Effects of Jitter on Digital Audio Quality", Pre-print 4826 of the 105th AES Convention, San Francisco, September 1998. This paper concluded that the threshold of audibility of jitter on normal music signals is around 20nS (that's 20,000pS - rather more than 71pS

Thanks for posting that link. I bought the paper yesterday and you are correct - it concluded that even with the cheapest DVD players they tested - and this was ten years ago in 1998 - the jitter was well below the threshold anyone was able to hear. I also found a similar 2005 paper from another technical journal that concluded the same based on extensive testing.

This is not to say that people don't hear differences when swapping digital cables or A/D/A converters and word clocks. But if they do hear a difference, it's not due to reduced jitter.

Again, I agree with Pat and Bruno Putzeys and all the other competent EE's out there that aim to make their specs as good as possible. To do otherwise is just lazy, unless it can substantially reduce cost to the consumer while still preserving "good enough not to matter" fidelity.

--Ethan

It's interesting isn't it.  Jitter may well be a total boogeyman.  This is especially intriguing to me since I was feeling angst about computer audio besting my faithful CDP due in part to lower "jitter" levels. 

[Geardaddy]

What's the point of good specs if you can't hear the benefit?  It surely is cheaper to "not sweat it".

Amen...

[Daygloworange]

What's the point of good specs if you can't hear the benefit?  It surely is cheaper to "not sweat it".

You're preachin' to the choir, dude...

Cheers

[darrenyeats]

I can't believe this. I'm siding with the audiophile believers in a debate about audibility! (Normally, I am a skeptic and disbelieve 90% of the differences people speak about. I believe it's the placebo effect in many many cases.)

As mentioned I did a blind test of my old CDT vs SB3 both feeding via S/PDIF into my old Sony DAC. Very old equipment (apart from the SB3) but my listening impressions blind and the listening results pointed to a real audible difference. I'm pretty sure it wasn't just in my head since I heard pretty consistent differences blind, and I chose one transport consistently. Given that both transports were just passing 1s and 0s to the DAC the differences must have been read errors (by the CDT) or jitter.

So stating "it doesn't make a difference" is a bit of a blanket statement. Maybe, with modern equipment jitter isn't audible. But that is not the same as saying jitter is never audible in domestic replay. I can't prove it but I feel I have good justification for this view.
Darren

[miklorsmith]

Different digital gear sounds different to me.  If jitter was positively proven to be irrelevant (which it hasn't been as far as I can tell), it wouldn't change that.  Jitter is one unknown that always surfaces as to why that is.  Audioheads, as people in general, try to simplify things into terms and concepts they can understand.  It's a lot easier to discuss "jitter" than all the variables and decisions that went into making a particular piece.  It may fully be that jitter is not the central problem but that like preamplifiers or amplifiers or anything else, the design choices made by the designer ultimately define its mettle.

So, to me the discussion of jitter is a bit of a non-starter.  I personally believe it's an important design consideration but I equally believe a low-jitter setup could sound like crap.  Trying to get our heads around what makes good digital sound is a natural inquiry but focusing on a single element is like trying to pull grapes out of wine.

[mfsoa]

Darren,
Although I sometimes disagree with that other 90% of you that remains skeptical (while totally admitting that the placebo effect must always be considered), I gotta give you snaps for using your ears and being honest about it.

I was posting a lot the other day, home from work with a nasty head cold so I don't know how my posts really came off.

I think where I stand between the two camps is that I do have a problem with "My textbook and this published, peer reviewed paper says so, and therefore it is so" but I have no problem with "I tried it and couldn't hear a difference"

So hears (pun intended) to those who are skeptics but still actually listen!

-Mike

[Geardaddy]

I agree with you guys.  I wonder if part of the reason HD-driven systems sound better than CDPs is the "read until right" issue and other more theoretical phenonema such as the Reed Solomon error correction programming that Nova Physics mentions rather than or above and beyond jitter.  As I may have mentioned, I plan on doing some blinded listening with friends (modded SB3 vs Sonos + Exemplar USB DAC vs Granite Audio CDP).  Should be interesting and potentially informative...as long as we do not introduce too many confounding variables such as beer...

[mfsoa]

Ethan,

This is not to say that people don't hear differences when swapping digital cables or A/D/A converters and word clocks. But if they do hear a difference, it's not due to reduced jitter.

I'd be curious to know what you think might be the reason that different CD-R media sound different to each other and to the original CD they were copied from?  I assume the bits are the same, so the digital part is not the issue. What's left is the analog part (the timing) which is what I consider to be jitter-relevent. And since it's been shown that Gold CD-R media can have lower jitter, why are you so sure that if bit-perfect copies (if indeed the copies are) sound different when made on media with lower jitter, that the jitter is not what's causing the difference?

Thanks!

-Mike

[Russell Dawkins]

Re gold media, FWIW.

I have an acquaintance who has an interest in testing digital media for error rate.

He says that, while most archival CDRs use gold as the reflective surface (or in the case of the ones I use, silver over gold for better reflectivity) the discs that show the fewest errors are never gold.

The value of gold is that it has good archival characteristics, not that it has the lowest error rate.

300 years is claimed for the best, although the jury is still out on that one.