[bigshot]

  I might suggest you find a DAC that uses the Wolfson 8741 chip.

 
I'm pretty sure that one of the iPods I tested with had a Wolfson chip in it. The filters weren't adjustable though. I wouldn't be surprised if optional filters did actually make a measurable and audible difference in the range of human hearing though. That's kind of the purpose of giving options, isn't it? I know some people prefer a smoother wider rolloff at the top end to give a "warm analogue sound" as opposed to keeping it flat up to beyond the range of human hearing and then filtering up there were it can't be heard.

 

[Steve Eddy]

I do have an interesting question for you... since you are the one making the "positive assertion" here (that "all good DACs sound the same").

That's funny. I can't find that phrase in his post. Yet you use quotation marks.

I think you're being rather disingenuous here.

se

 

[cel4145]

I might suggest you find a DAC that uses the Wolfson 8741 chip. That chip includes several different internal digital filters, and offers the ability to switch between them using a simple hardware switch - so many DACs that use it give you a front panel switch to switch between filters (and, by flipping the switch, the only thing you're changing is that internal filter). Regardless of whether you suspect Wolfson of deliberately "fiddling" with things to artificially produce audible differences, the fact remains that different filters, all with THD low enough that it should be inaudible, and frequency responses flat enough that any variations there should also be inaudible, nevertheless do manage to somehow sound different. This clearly proves that there's something else that constitutes an audible difference between them - which disproved your claim that no such difference can exist.

I didn't follow everything you wrote earlier. But did you establish that this claim is based on ABX testing? Or just what you perceive as different in sighted listening tests? And if so, with the ODAC?

 

[KeithEmo]

   
Level matched, direct A/B switchable, single blind on three different systems- headphones, my main speaker rig and the rig of a sound mixer friend of mine. We did a test between SACD and CD too the same way. Same results. That's good enough for my use. Everyone should do it for themselves. A lot of people would be very surprised about what they can't hear.

 
Cool, and that would certainly be sufficient for an individual to determine if the differences were audible in their system (which should be plenty if you're deciding what to buy).
 
However, I was looking for more specific information. For example, I find the differences between my various DACs to be clearly audible with my Koss ESP950's (electrostatic headphones - which include their own headphone amplifier and accept line level inputs), and on my Airmotiv 5 powered monitors, yet I am pretty much unable to hear any differences between them when I use my AKG K271 mkII or AKG K240 mkII headphones (using the headphone amplifiers in the DACs).  [In all instances using a computer to play FLAC losslessly compressed files - a mixture of files ripped from regular Red Book CDs and assorted 24/96 and 24/192 audio files, mostly from HDTracks. These being played through an Audiophilleo 2 USB-to-S/PDIF converter, which has low enough jitter to ensure that jitter isn't a factor, and then fed into the Coax S/PDIF input on the DAC.]
 
Differences clearly heard between: AudioQuest DragonFly (not through the converter since it is USB only); Emotiva XDA-2 (v1), Emotiva DC-1, Wyred4Sound DAC2, AudioLAB MDAC, Schiit Bifrost (original).
 
This clearly suggests that some components mask the differences I'm hearing (or some components exaggerate them).
 
Therefore, the fact that no differences are audible on specific equipment is useful information, but only in the context of knowing what equipment was used.

 

[bigshot]

I'll say it. A good DAC should be designed to be audibly transparent. And an awful lot of DACs are. I haven't found one that isn't myself yet, but I don't buy DACs with options to change the way the DAC sounds. That makes no sense to me.
 
My Oppo HA-1 and BDP103D sound exactly the same as my iPod and my iMac and my $120 Sony blu-ray player. Those are the things I use right now.
 
I'm not sure what specific equipment my sound mixer friend has. But I have a VERY good 5.1 speaker system and Oppo PM-1s if that helps you put it in perspective. I check all my electronic equipment to make sure it's audibly transparent when I buy each piece. I have carefully set the room acoustics and EQ and I don't want anything that is colored in the signal chain messing that up. So far, everything is plug and play... just as transparent as the piece it's replacing.
 
If I do happen across equipment with coloration, I'll pack it up and send it back for a return. I have no use colored equipment.

 

[KeithEmo]

   
I'm pretty sure that one of the iPods I tested with had a Wolfson chip in it. The filters weren't adjustable though. I wouldn't be surprised if optional filters did actually make a measurable and audible difference in the range of human hearing though. That's kind of the purpose of giving options, isn't it? I know some people prefer a smoother wider rolloff at the top end to give a "warm analogue sound" as opposed to keeping it flat up to beyond the range of human hearing and then filtering up there were it can't be heard.

 
I agree entirely.... however, my point was that most of their filters (including ones that clearly do sound different) are flat enough between 20 Hz and 20 kHz, and have low enough noise and THD, that, according to what some people seem to keep insisting, they should sound audibly the same. And, so, the fact that they don't proves that those parameters are insufficient to "define how a DAC sounds".
 
In other words, all by itself, that chip is proof that DACs with audibly flat frequency response, and inaudible levels of THD and noise, can indeed still sound different. (And, since virtually all manufacturers of DAC chips seem quite convinced that this is due to the differences in digital filters they use, I'm inclined to consider that to be the best theory to date as to why.)

 

[uchihaitachi]

   
I agree entirely.... however, my point was that most of their filters (including ones that clearly do sound different) are flat enough between 20 Hz and 20 kHz, and have low enough noise and THD, that, according to what some people seem to keep insisting, they should sound audibly the same. And, so, the fact that they don't proves that those parameters are insufficient to "define how a DAC sounds".
 
In other words, all by itself, that chip is proof that DACs with audibly flat frequency response, and inaudible levels of THD and noise, can indeed still sound different. (And, since virtually all manufacturers of DAC chips seem quite convinced that this is due to the differences in digital filters they use, I'm inclined to consider that to be the best theory to date as to why.)

http://www.tomshardware.com/reviews/high-end-pc-audio,3733.html 
 
Awesome link, you should have a read!

 

[KeithEmo]

I didn't follow everything you wrote earlier. But did you establish that this claim is based on ABX testing? Or just what you perceive as different in sighted listening tests? And if so, with the ODAC?

 
Nope.... based on individual (and perhaps some "group") listening... which is why I've been pretty careful to avoid making sweeping generalizations. In order to make valid generalizations, you need to have a lot of data, and you need to be very careful about how you interpret it.
 
For example, what if a dozen people perform a double-blind ABX test, yet they EXPECT to not hear any difference between the various units under test? Doesn't that constitute an expectation bias that biases them in the direction of hearing no difference? The only way I can think of to avoid that would be to be sure to include units that we DO expect to sound different - so we can have a control showing that the participants actually do hear differences that we know do exist. So, if you want me to accept a study showing that THD of 0.1% is "the threshold of audibility", then you need to present data showing both that THD < 0.1% was NOT detected by the participants, and that THD > 0.1% WAS detected. (I would compare the results of a whole bunch of samples, at all different levels of THD, and want to see a line - where everything below the line was inaudible and everything above it was.)
 
Otherwise it's possible that the test conditions simply weren't appropriate to determine the question. Like I said before, the fact that I heard no difference with AKG 240 headphones could show that either a) there is no difference or b) the difference, if present, cannot be heard with those headphones.

 

[icebear]

I got a Marantz SA11-S3 a couple of months back which comes with some options for filter rolling

 
http://us.marantz.com/us/Products/Pages/ProductDetails.aspx?CatId=ReferenceSeries&ProductId=SA11S3
quote :
[Multiple Filters Offer Versatile Listening Options
Two digital filter options are provided for both PCM and DSD bitstreams. With PCM content Filter 1 has a symmetrical impulse response with extremely short pre- and post-echo characteristics and an absolutely linear phase response for a sound that is the most faithful to the original digital audio source, while the Filter 2 option provides a more analog-like signal that strikes the best balance of fine detail coupled with rich harmonics. With DSD content from SACD, Filter 1 provides a direct decoding function with no additional filtering, while the Filter 2 option attenuates any residuals above 100 kHz.
As well, there are two additional digital filtering functions, one that includes noise shaping for improved in-band low level linearity as well as another one that provides effective filtering of extremely low frequency infrasonic signals (stop band is DC~1.7 Hz). All three digital filtering functions are controlled by dedicated buttons on the remote, with last memory function that retains the settings during disc changes and even when the player is powered down.] unquote
 
Acc. to the marketing speak some options do influence freq. above 100kHz and below 1.7Hz. Switching back and forth between these filters slightly alters the soundstage width. There is no night and day difference in sound and I most likely would have a hard time telling which filter is used when I left the room and came back and someone switched the setting. Never the less there is a difference in sound even if the shaping is supposedly influencing sub or far above the audible frequency spectrum. I always end up with filter one, no noise shaping and no DC filter in case I feel to try any of the filters. It sounds more open/airy is the best I could try to put it into words. And although it's not actually blind testing, the player is so far away and the display so tiny, that I can't see a thing, unless I stand up and walk over

.
 
btw: the player uses a Burr-Brown DSD1792A, about $20 a piece at Mouser (cheaper in bulk quantities)

.
And more importantly I do like the sound in general and particularly with SACD's - if that is caused by better mastering only and not the DSD format - which is highly likely - I do not really care, I just enjoy listening to the music

.

 

[Steve Eddy]

For example, what if a dozen people perform a double-blind ABX test, yet they EXPECT to not hear any difference between the various units under test? Doesn't that constitute an expectation bias that biases them in the direction of hearing no difference?

Don't know much about ABX testing, do you? If you did, you wouldn't even ask such a question because if the listener isn't confident that they're hearing a difference between A and B (which are always known), then the test can't even proceed, and there is no result. Period. Not a null result. Not a positive result. But no result at all.

The only way I can think of to avoid that would be to be sure to include units that we DO expect to sound different - so we can have a control showing that the participants actually do hear differences that we know do exist.

That's what every ABX test requires. You really need to brush up on the subject.

And what's this "...we know do exist" nonsense? When was this established except by way of ego, vanity and pure faith?

All we've had from your side for decades has been a lot of claims supported by nothing but a bunch of hand-waving. It grew old years ago and is tiresome. I've been listening to it for over 30 years.

se

 

[Roly1650]

I do have an interesting question for you... since you are the one making the "positive assertion" here (that "all good DACs sound the same"). Have you actually compared the 0Dac to other DACs, double-blind or otherwise, or are you just accepting other people's claims about what specifications are and are not audible?

I might suggest you find a DAC that uses the Wolfson 8741 chip. That chip includes several different internal digital filters, and offers the ability to switch between them using a simple hardware switch - so many DACs that use it give you a front panel switch to switch between filters (and, by flipping the switch, the only thing you're changing is that internal filter). Regardless of whether you suspect Wolfson of deliberately "fiddling" with things to artificially produce audible differences, the fact remains that different filters, all with THD low enough that it should be inaudible, and frequency responses flat enough that any variations there should also be inaudible, nevertheless do manage to somehow sound different. This clearly proves that there's something else that constitutes an audible difference between them - which disproved your claim that no such difference can exist.

(I don't know who's producing DAcs using the Wolfson lately - but AudioG*D used to make several reasonably priced ones.....

Certainly, Audio GD don't claim that the frequency response graphs are the same on their website. See the link for the various filter switch positions on the NFB12 with dual Wolfson 8741 dac chips.
Are the differences enough to be audible? Well not being identical and varying with different sample rate/bit depths would indicate there's a fair chance, but as you say maybe the quality of the equipment determines that.

http://www.audio-gd.com/Pro/Headphoneamp/NFB12/NFB12EN_Specs.htm

 

[bigshot]

 
I agree entirely.... however, my point was that most of their filters (including ones that clearly do sound different) are flat enough between 20 Hz and 20 kHz, and have low enough noise and THD, that, according to what some people seem to keep insisting, they should sound audibly the same. And, so, the fact that they don't proves that those parameters are insufficient to "define how a DAC sounds".

 
I would recommend going back and measuring again, and performing an ABX test to make sure a demonstrable difference exists. I bet if you did those two things, you would find out there is a mistake somewhere. Or you just might establish an entirely new theory of sound reproduction. Go for it! Maybe you'll make history!
 
 
(but I've got two bits bet on the mistake)

 

[KeithEmo]

That's funny. I can't find that phrase in his post. Yet you use quotation marks.

I think you're being rather disingenuous here.

se

 
I put quotation marks around the phrase because, while the concept I was getting at is well known in philosophy, I'm not sure of the proper name for it.
 
Basically, if I say I cannot hear any difference, then that statement makes no claims that impinge on you (I'm not saying that you can or can't hear anything).
Likewise, if I say that I can hear a difference, but with specific test equipment, and under specific conditions, I am still only making a claim about my personal experience.
 
However, if I state the "no difference exists" then, by inference, I am making a claim about YOUR experience.
(I'm stating that I don't hear a difference, but I'm also stating that, since no difference exists, either you won't hear one either or, if you say you do, then you must be wrong.)
 
It's like my saying that a certain address on a certain street doesn't exist - if you claim that it does, then either you're wrong or you're lying.
(And stating it as a fact is not the same as saying "I think you're wrong" or "I'd like to see your map".)
 
By stating that "no difference exists" as if it were a proven fact, he is essentially saying that I am wrong.
He could have said something like "all of the evidence and results of studies I've seen seem to suggest that the difference shouldn't be audible" if he preferred to be less contentious.
 
The simple fact is that, if he were actually able to demonstrate that there were no differences between various DACs, then I would be on his side.
However, the differences are very well documented (just look at impulse response test graphs of ay DAC).
The sole "item under discussion" is whether the differences are audible or not......
 
And, while it's true that some studies have indeed failed to prove that differences in digital filters are actually audible, the few actual test results people keep trotting out are rather thin. Face it, the fact that two dozen volunteers were unable to hear something, using their own equipment, and their own sample material, really doesn't constitute "ironclad proof that no audible difference exists" - at least not on the planet where I live.
 
I'm not asking anyone to take my word for it that a difference exists, but I also haven't seen any "proof" that such a difference doesn't exist.
(I will concede that the differences I'm hearing might be due to something else, or even that I might be imagining them, but I will expect actual proof of that before I consider it to be true.)
The "standards" about what is and isn't audible, like much of what we "know" of science, are subject to periodic revision...

 

[KeithEmo]

Certainly, Audio GD don't claim that the frequency response graphs are the same on their website. See the link for the various filter switch positions on the NFB12 with dual Wolfson 8741 dac chips.
Are the differences enough to be audible? Well not being identical and varying with different sample rate/bit depths would indicate there's a fair chance, but as you say maybe the quality of the equipment determines that.

http://www.audio-gd.com/Pro/Headphoneamp/NFB12/NFB12EN_Specs.htm

 
If you look at the graphs, you will see that the graphs for several of the filters are all within the arbitrary limits that seem to be accepted as "inaudible variation".
(Let's ignore the apodizing filters with the high-frequency roll off.)
For example, if you use 24/96 content, you will find that many of them remain "flat" to within a very small fraction of a dB from 20 to 20 kHz.
Therefore, those filters should sound exactly the same if only frequency response matters.
I believe you will find that the THD figures for all of them also fall safely under the 0.01% which most people seem to consider as "obviously inaudible".
 
And, unless EVERY audible difference can be credited to a significant difference in frequency response or THD, then my point is made....

 

[sonitus mirus]

It is not logical to expect someone to be able to test and determine that no difference exists.  We can determine if a difference exists through testing.  Until someone does this, I think it is fair to claim that no difference exists.