or Connect
Head-Fi.org › Forums › Equipment Forums › Sound Science › Audiophile cables, an interesting question.
New Posts  All Forums:Forum Nav:

Audiophile cables, an interesting question. - Page 54

post #796 of 1186
Quote:
Originally Posted by x838nwy View Post

This was in response to a claim that there was no difference between (hifi) dacs while there is - you yourself mentioned on manufacturer using a different filter.
 


Different filter, as in:
–"Holy smokes, dude, output's -3dB at 15kHz, and look at those side bands. Something must be wrong here!"
–"Naw, man, it's audiophile, innit. 'tis s'posed to be like tha' – pass me the filters, and the papers, will ya?"

We've all seen the the horror stories, Wadia, Macintosh, HiFi-man etc. Of course a DAC can sound different if it's been sufficiently botched up.

post #797 of 1186

Look, as I said earlier. Find the most expensive DAC you can (that you think earns that price point) and I'll put up my $60 DAC in a DBT anytime, any day.

 

We run enough trials and you will not be able to consistently pick the more expensive DAC. I'd put money on it.

post #798 of 1186
Quote:
Originally Posted by limpidglitch View Post
 


Different filter, as in:
–"Holy smokes, dude, output's -3dB at 15kHz, and look at those side bands. Something must be wrong here!"
–"Naw, man, it's audiophile, innit. 'tis s'posed to be like tha' – pass me the filters, and the papers, will ya?"

We've all seen the the horror stories, Wadia, Macintosh, HiFi-man etc. Of course a DAC can sound different if it's been sufficiently botched up.

Can you tell me where I can read up these stories you are referring to?  

post #799 of 1186
Quote:
Originally Posted by SilverEars View Post
 

Can you tell me where I can read up these stories you are referring to?  


I can't remember the specific model of Wadia, but they seem to prefer slow roll-off for many of their DACses.
Macintosh had that music server a while back, and HiFi-man had the HM-801.
The NuForce uDac and CDP were pretty bad too.

Stereophile has measurements for most of these, you-know-who measured the uDAC, and various RMAA runs exist of the HM-801.

post #800 of 1186
Quote:
Originally Posted by x838nwy View Post

Esldude - i actually think we're on the same page on most things. I never said that more expensive dacs are better neither did i say that cheap dacs are no good. If you go back to my original post on the subject, i said the differences are possible and design play a large role. This was in response to a claim that there was no difference between (hifi) dacs while there is - you yourself mentioned on manufacturer using a different filter.

The purrin thing i referred to simply as they used some of the dacs i have or have heard so that it's not all in my head or all that the dealer's facilities.

Data - i cannot find any that leads to a conclusion that shows there is no difference between dacs. None that is more stringent of conditions than a bunch of guys listening to stuff. The only addition was that it was done blind.

In my little tests, the differences were tiny if they were even there at all. But the differences between dacs in my (probably invalid again) experience can be very apparent that in my opinion it is not an absolute requirement to do so blind. I'm not saying sighted tests are as valid, but for cases between some dacs, there's no need specially if it's to see if they are different.


I know sighted listening of DACs sounds convincing.  But like other things it just doesn't pan out.

 

You say you cannot find data leading to the conclusion there isn't a difference between DACs that is more stringent than guys listening to stuff.   There is plenty.

 

Frequency response is 20hz-20khz for humans.  Keep that flat and there is no room for that to matter.   Generally distortion is inaudible below .1% and usually much higher with music.  Signal to noise, instantaneous SNR for ears is maybe 60 db but just to be safe less ask for 80 db.  But with the fact few rooms are quieter than 30 db noise levels adding 96 db on top of that puts you over the threshold of pain.  None of this is unsupported data.  It is all well established material.  I am not going to cite each and every source for that, but if you doubt it such is easy to find.  Or we can point to textbooks with the info for you to read.  So, did I mention DACs in this paragraph yet?  Nope I didn't.  However, any musical delivery system, any blackbox that can take in a music signal digital or analog and put it out with accuracy equal to or better than the above envelope will sound audibly transparent.  .01% THD will not sound audibly different than .00001%.  So on and so forth. 

 

So  looking at the above guidelines how many DAC's fail to be flat in response, less than .1% in THD and IMD, with less than 80 db of SNR?  I think you will find the answer is very few indeed.  So for a DAC to sound different you need a signal different.  When the differences lie at extremely low levels (and they do I have measured a few, so have many people) what is present in the signal to hear as better or worse?  So no we don't go around performing an ABX test on every pair of DACs in existence.  We don't need to do so.  We have well supported data to the abilities of human hearing, and when something fits well within it the result won't be discernibly different.  Those times when more rigorous testing has been done give results that line up with that knowledge.  No need to re-invent the wheel in each case so to speak. 

 

Now the other part of this is how we get fooled in hearing differences we can perceive as so real, but aren't there.  In this there is much data, much research, not just in hearing, but in all of our senses showing how easily our brain generates non existent differences.  Our brains could be described as pattern matching difference engines.  Predisposed in a myriad of ways for finding a difference.  There was great evolutionary and survival value in that built in bias.  But at the margins of perception it generates differences which aren't real.  Usually it is much safer to operate on possible differences that aren't real than it is to perhaps miss a difference which is real.  Real missed differences could cost you your life.  Just as obviously is the emotional, gut level, strength at which those perceptions will function.  It had survival value for our kind.  But in a modern context for which we have not biologically adapted it can lead you astray.  Buying overly complex or expensive audio equipment isn't deadly for the most part. 

post #801 of 1186
Quote:
Originally Posted by limpidglitch View Post


Different filter, as in:

–"Holy smokes, dude, output's -3dB at 15kHz, and look at those side bands. Something must be wrong here!"

–"Naw, man, it's audiophile, innit. 'tis s'posed to be like tha' – pass me the filters, and the papers, will ya?"


We've all seen the the horror stories, Wadia, Macintosh, HiFi-man etc. Of course a DAC can sound different if it's been sufficiently botched up.

Umm no, not that one. Although i don't know which esldude was talking about, there are many dacs that have switchable filters (apodising, pre/post ringing, that sort of thing) in the digital domain. I think Meridian use their own one of these which is supposedly different to others. What esldude mentioned sounds like an analogue filter, so best pm him to find out.
post #802 of 1186
Quote:
Originally Posted by limpidglitch View Post
 


I can't remember the specific model of Wadia, but they seem to prefer slow roll-off for many of their DACses.
Macintosh had that music server a while back, and HiFi-man had the HM-801.
The NuForce uDac and CDP were pretty bad too.

Stereophile has measurements for most of these, you-know-who measured the uDAC, and various RMAA runs exist of the HM-801.

Looks like Wadia and HM-801 use PCM1704K.  This seems like poor choice of DAC chip used or I don't know why they chose that one.  Any ideas??  I know about u know who.  These were unacceptable. I really don't know what was going through their minds.  How can they design a circuit with measurements like this?  It's so rediculous.  

 

 

 


Edited by SilverEars - 5/31/14 at 10:53pm
post #803 of 1186
Quote:
Originally Posted by x838nwy View Post


Umm no, not that one. Although i don't know which esldude was talking about, there are many dacs that have switchable filters (apodising, pre/post ringing, that sort of thing) in the digital domain. I think Meridian use their own one of these which is supposedly different to others. What esldude mentioned sounds like an analogue filter, so best pm him to find out.

 

Mostly gimmick, I'm afraid. Different reconstruction filters have different draw backs, but applied correctly any of them will operate transparently.

 

 

Quote:

Originally Posted by SilverEars View Post
 

Looks like Wadia and HM-801 use PCM1704K.  This seems like poor choice of DAC chip used or I don't know why they chose that one.  Any ideas??  I know about u know who.  These were unacceptable. I really don't know what was going through their minds.  How can they design a circuit with measurements like this?  It's so rediculous.  

 


Design by ear? Sighted listening tests?
To be fair, the HM-801 is orders of magnitude worse than the Wadia. I doubt I could detect -.5dB at 14kHz, or whatever it is.


Edited by limpidglitch - 5/31/14 at 11:41pm
post #804 of 1186
Quote:
Originally Posted by SilverEars View Post
 

Can you tell me where I can read up these stories you are referring to?  


I can testify to Wadia.  I owned a couple of them.  They used different filters which did result in something like -3 db at 15 khz which in those days I could hear.   Now I may or may not notice it.  They also had kick ass output stages with the ability to put out several volts and very low output impedance.  Fittingly enough one of those is why I learned about level matching somewhat.  I loaned it out to several audiophile friends for weeks at a time.  All thought it great.  Now all had to turn it down as it put out something like 4.5 volts vs usually somewhere around 2-2.5 volts for others.  I got to noticing everyone turned it down, but not quite enough.  Eventually I would go to retrieve it and insist on level matching for a final audition before I took it home.  Several people who had raved about it when comparing it sighted to level matched DACs they owned found,   "oops, not as different as I thought".  None could shake the idea the Wadia was better, but they did hear it was not the huge difference they thought it was. 

 

Now the latter Wadia also convinced me that digital was not a problem. It was a Wadia 25.  I paired it with an MSB AD converter.  People thought I was crazy to digitize my phono.  But it simply was so transparent as to be a big step forward.  Phono, reel-to-reel tape from my Revox, even FM radio seemed simply transparent.  The MSB seemed to do nothing more than a set of analog interconnects.  Feed the digital result to the 25 which had a good digital volume control and I had very nice sound.   The great thing is you can get that for a fraction of the price with current equipment.  So really the MSB Audio Director AD is really what convinced me.  I digitized all analog fed it to the Wadia and controlled volume digitally.  The Wadia did sound different, different by design.  Not fully transparent in fact. 

post #805 of 1186
Quote:
Originally Posted by esldude View Post
 


I know sighted listening of DACs sounds convincing.  But like other things it just doesn't pan out.

 

You say you cannot find data leading to the conclusion there isn't a difference between DACs that is more stringent than guys listening to stuff.   There is plenty.

 

Frequency response is 20hz-20khz for humans.  Keep that flat and there is no room for that to matter.   Generally distortion is inaudible below .1% and usually much higher with music.  Signal to noise, instantaneous SNR for ears is maybe 60 db but just to be safe less ask for 80 db.  But with the fact few rooms are quieter than 30 db noise levels adding 96 db on top of that puts you over the threshold of pain.  None of this is unsupported data.  It is all well established material.  I am not going to cite each and every source for that, but if you doubt it such is easy to find.  Or we can point to textbooks with the info for you to read.  So, did I mention DACs in this paragraph yet?  Nope I didn't.  However, any musical delivery system, any blackbox that can take in a music signal digital or analog and put it out with accuracy equal to or better than the above envelope will sound audibly transparent.  .01% THD will not sound audibly different than .00001%.  So on and so forth. 

 

So  looking at the above guidelines how many DAC's fail to be flat in response, less than .1% in THD and IMD, with less than 80 db of SNR?  I think you will find the answer is very few indeed.  So for a DAC to sound different you need a signal different.  When the differences lie at extremely low levels (and they do I have measured a few, so have many people) what is present in the signal to hear as better or worse?  So no we don't go around performing an ABX test on every pair of DACs in existence.  We don't need to do so.  We have well supported data to the abilities of human hearing, and when something fits well within it the result won't be discernibly different.  Those times when more rigorous testing has been done give results that line up with that knowledge.  No need to re-invent the wheel in each case so to speak. 

 

Now the other part of this is how we get fooled in hearing differences we can perceive as so real, but aren't there.  In this there is much data, much research, not just in hearing, but in all of our senses showing how easily our brain generates non existent differences.  Our brains could be described as pattern matching difference engines.  Predisposed in a myriad of ways for finding a difference.  There was great evolutionary and survival value in that built in bias.  But at the margins of perception it generates differences which aren't real.  Usually it is much safer to operate on possible differences that aren't real than it is to perhaps miss a difference which is real.  Real missed differences could cost you your life.  Just as obviously is the emotional, gut level, strength at which those perceptions will function.  It had survival value for our kind.  But in a modern context for which we have not biologically adapted it can lead you astray.  Buying overly complex or expensive audio equipment isn't deadly for the most part. 

 

Stuff in bold: Yup, that's pretty much normal.

 

Sutff underlined: That's very few. Different to none. Even "hifi" DACS. SilverEars just managed to post an FR that isn't flat to 20kHz.

 

Stuff in bold + italics: The magnitude of the difference really depends on the DAC. I'm sure if you listen to the Wadia or the 801 posted earlier, you'd notice a difference between them and other DACs. I'm pretty sure the difference between the HM-801 and the Clip+ will be audible.

 

I really think we're arguing over different things here. What you seem to be saying is that all modern Hi-fi DACs are the same. I am saying NOT ALL of them are the same. Most are very similar, but not ALL. I'm not talking about only obscure DACS from flea bay, but DACs you find on most audio forums and sites. That is all. I am not stating that there is, say a difference between a Mytek and a Benchmark or whatever. Simply that not ALL sound the same.

 

The information I mentioned I cannot find is that there is no information that states that ALL DACS sound the same.

 

Again, I refer you to the ODAC's development blog. One part of the very long blog mentions:

 

Quote:
The ODAC has been through four lengthy revisions—two of the earlier boards are shown to the right. Despite the fact we started with essentially the reference design from the datasheets, the devil was in the details. The  first version played music and sounded OK. Many companies and DIYers that “design by ear” would have stopped there. But that first version didn’t come close to delivering what the DAC chip is capable of. Each revision cycle took at least several weeks, cost hundreds of dollars, and involved countless hours of work. But, in the end, it resulted in muchbetter performance compared to where we started..

 

I'd like to focus on the "much better performance" part. The designer was someone who's never far away from the "what's audible" since the start and began with what was close to what you mentioned in terms of human hearing ability. So I am assuming that he's saying version 4 is audibly better than version 3, 2 or 1.

 

At the end he said that in a blind test he and one other could not tell the difference between ODAC rev.4 and the Benchmark 1. Which is great and follows what you're saying. But had they stuck to Rev.1, 2 or 3 for production it would not have been. Which is sort of my point - DACs can be designed poorly and the difference between a poorly designed DAC and a well designed one is audible. I don't know how good DAC's are in things like cheap CD players and so on, but the ODAC was designed using many very sophisticated equipment and the designer himself insisted that he would not be able to achieve this result without said equipment. Some people design "by ear" and some simply just don't have the test equipment or simply do not/cannot use them who knows. But it is possible to design a poor dec and even more possible to sell them.

post #806 of 1186
Quote:

Originally Posted by esldude View Post
 

<Snip>

 

As for the statement analog design doesn't matter?  I never made that statement.  I did say if the analog circuits were of a enough quality to be audibly transparent the circuit didn't matter.  A subtle, but important difference.


<Snip>

 

 

Quote:
Originally Posted by esldude View Post
 

<Snip>

 

Now if a device like a DAC gets the same data stream, puts out an analog result that is close enough to be audibly transparent it will sound the same.  Chips to make it happen don't matter, analog design don't matter, switched or linear supplies don't matter, just the resulting signal.  The resulting signal is what you actually listen to in the end. 

 

<Snip>

 

Just for your information that I didn't claim you said what you didn't. I think your point is that you're looking at the DAC as a box and its output as the RCA or XLR output.

 

What I was talking about was that while DAC chips do the same thing for a given input, the analogue design stage will influence the final output. As such the analogue stage is one more variable for DACs to sound different.

post #807 of 1186

I own several DACs, and continue to look for better DACs because I hear differences.  Dragonfly which was popularized for it's asynchronous ability sounds very smooth, lacking details.  One of my headphones can pick up certain details because of it's sound signature the other cannot.  The one that cannot pick up the subtle details from the dragonfly, puts out the details when sourced from ODAC.  I also have gamma 2.  I have fed in both(Gamma 2 and ODAC) to my Objective 2 amp, and can decern difference between the two as ODAC sound more transparent.  Matter a fact, ODAC is probably my favorite DAC as it sounds transparent compared to my other sources.  I have also tried the Fiio E17, which does have a more colder sound signature, and not as transparent as the ODAC.  I hear the differences, and if you do not, that's fine.  No need to waste you money on looking for best DAC.  All of these have measurements public and all are spec'd less than what are audible distortions.  I haven't found measurements for Dragonfly though. 

 

I'm not looking for the most expensive DAC or have the desire to spend a lot, but I need to know if high end DACs make sonic difference as I'm searching for greater sounding setup.  Some maybe hogwash, but there are many out there, and I do research to make as much educated guess as possible for my purchases.  I have not delved into expensive DACs yet, but I'm itching to because I'm curious.  The way I think about it is, I cannot know if there is not difference unless I try listening to DACs and compare them. I long figured out specs cannot help me know how they would sound.

 

I have also listened to the overpriced AK240, and was surprised.  I unfortunately compared it to my galaxy s4, and it was very obvious coming out of Ultimate Ears reference monitors IEM.  My opinion is based on my own experiences, you experiences may differ.  I like to try everything before concluding.  I like to make the most educated purchases, but people that has tried lot more gear blindly know much more than I do on what sounds the best since they have tried many gear.


Edited by SilverEars - 6/1/14 at 12:02am
post #808 of 1186

Should we start a new thread to discuss issues with DAC design? Because it's a little off topic in the cables thread.

 

Anyone know the science behind "diode effects" in "directional, burned-in" cables?

 

How about dielectric-bias systems?

 

 

Cheers

post #809 of 1186
Quote:
Originally Posted by ab initio View Post

Should we start a new thread to discuss issues with DAC design? Because it's a little off topic in the cables thread.

Anyone know the science behind "diode effects" in "directional, burned-in" cables?

How about dielectric-bias systems?


Cheers

Yeah, it was going off a bit wasn't it smily_headphones1.gif

And yes, the cable directionality thing actually exist i found. In XLR cables. Plugging them in the wrong way round really messes up the sq. biggrin.gif

Sorry, that was a really bad joke.
post #810 of 1186
Quote:
Originally Posted by x838nwy View Post

Yeah, it was going off a bit wasn't it smily_headphones1.gif

And yes, the cable directionality thing actually exist i found. In XLR cables. Plugging them in the wrong way round really messes up the sq. biggrin.gif

Sorry, that was a really bad joke.

I'm all for bad jokes!

Cheers
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Audiophile cables, an interesting question.