Benefits of using expensive DACs

[knownothing]

View attachment 410594

Good one!

 

[knownothing]

Since you've asked for constructive feedback,
Your description of the "blind test" appears anecdotal and lacks tangible evidence to support your claim. To ensure credibility, it would be helpful to provide detailed documentation or photos of the setup, as simply describing the test in a post isn’t enough to substantiate such a claim.

If you and other participants genuinely observed a difference, the burden of proof is on you to provide convincing evidence, especially when the claim seem unlikely.

Again, I freely admit this exercise could be greatly improved to achieve anything more than a slightly more controlled subjective evaluation to inform my own buying decisions. I put it forward here for critique, not to defend.

Here is a link to my current system set up.

knownothing's Music/Library/Theater System | Virtual Listening Room

knownothing's listening room featuring Leviton 20 Amp Isolated Ground Duplex Outlet, Nanotec Systems #308 "Wonderful" Power Cable, Shunyata Research Alpha HC ZTRON, Shunyata Research Venom PS8, Shunyata Research Venom Defender, FURUTECH FP TCS21 cable/Rhodium connector power cord, FURUTECH...

www.audiogon.com

The digital front end has changed since the test, with replacement of the DAC and an external power supply now supports both the Node and the DAC, but it’s physical layout is identical. All the major pieces in this system were bought used, on sale/closeout, or were factory seconds except for the subwoofer. I assembled several of the power cords I am using from bulk cable and purchased connectors.

The subject sat in the center of the couch directly between the speakers on the opposite long wall and did not move from that spot for the duration of the test. I sat near the electronics on the left front portion of the room and had the four different cables laid out behind the DAC and the Node. The power to the Node was left on for the whole test. The power to the DAC was disconnected immediately before the test and between each cable swap. The subject could not see the cables being swapped from where they were sitting. As noted previously, five different cuts were played in sequence for each cable, and then cables were swapped and the five cuts played again. It was up to the subject to determine how much of each cut they wanted to listen to. This entire sequence was repeated so there were two complete listening sessions for the four cables and five cuts. The tests were run at 60-70 decibels average volume as measured in the room, depending on the loudness of the cut played as the volume was held constant throughout.

These were the cuts used for the test:

1 - Sarasate: Carmen Fantasy, Op.25 Moderato, Anne Sophie Mutter, Weiner Philharmoniker, James Levine, “CARMEN-FANTASIE” - 24/88.2

2 - Lay Lady Lay, Cassandra Wilson, “Glamored” - 24/96

3 - Horizontal Blue, David Piltch, Kevin Breit, “Take One” - 24/192

4 - The Spot, Mickey Hart, “Planet Drum” - 24/96

5 - Siren Singing, Yello, “Point” - 24/48

I had said previously that some cuts at 16/44.1 were used in the test. On review, I was mistaken, all cuts used were higher resolution.

The front speakers are about 9ft apart on the long wall of the room and angled slightly towards the listening position. The listening position is about 11ft from the front wall. The subwoofer is positioned along the left wall about 4ft from the front wall. No DSP was engaged other than the settings in the subwoofer itself. The room is insulated on three sides and the ceiling with rock wool and the ceiling sheet rock is floated on aluminum spacers. This has more to do with keeping sound from affecting the rest of the house than acoustics in the listening room. There is a bookshelf covering the entire left wall. The floor is concrete with a large throw rug. The room is a little “dead” at lower volumes, eating some treble, but comes alive in terms of treble at about 65 to 70 decibels. I would say the bass without DSP in the room is “pretty good”. Turning the sub down one of two decibels on high bass content material like electronic or hip hop helps. The sub was not adjusted from its standard setting during the test.

kn

 

[Old_School_Brad]

Again, I freely admit this exercise could be greatly improved to achieve anything more than a slightly more controlled subjective evaluation to inform my own buying decisions.

I agree. At this point, it is, unfortunately, nothing more than a subjective assessment.

 

[knownothing]

I agree. At this point, it is, unfortunately, nothing more than a subjective assessment.

If you are interested and able, pointing out the specific areas of greatest failure to meet objective standards would be helpful.

kn

 

[SIY]

If yo

If you are interested and able, pointing out the specific areas of greatest failure to meet objective standards would be helpful.

kn

Double blind with good protocols.

Understand that you're making an extremely dubious claim. I mean the equivalent of claiming that you can see the back side of Pluto without a telescope. That means you have to do your controls properly for this to have any validity at all. And your blinding procedure is woefully inadequate.

 

[Old_School_Brad]

If yo

If you are interested and able, pointing out the specific areas of greatest failure to meet objective standards would be helpful.

kn

I believe explanations and links have already been shared, which should provide you with a clear understanding of areas for improvement.

 

[ahofer]

If you are interested and able, pointing out the specific areas of greatest failure to meet objective standards would be helpful.

here:

The subject could not see the cables being swapped from where they were sitting.

But he knew they were being switched and, if I'm not mistaken, which cable it was (by letter or other non-brand identifier)? As indicated in my prior post, that is the big one. The set up and everything else don't really matter when basic controls are lacking. It also helps to have the person switching the cables not know which is which (that's harder).

a few quick ways to add more control:

1) randomize the order of cables, including, from time to time, leaving the same cable in without telling the subject (e.g. in the first trial, cables A and B could be the same one)
2)rename the cables between trials, keeping track of which was which in each trial
3) collect impressions and preference order across trials
4) measure the cables for any damage or anomalies (LCR)

Due to the brevity of auditory memory, the listener will have a much better chance of telling the difference if the time between switching is minimized, so a switching appliance makes for a better test.

 

[knownothing]

Double blind with good protocols.

Understand that you're making an extremely dubious claim. I mean the equivalent of claiming that you can see the back side of Pluto without a telescope. That means you have to do your controls properly for this to have any validity at all. And your blinding procedure is woefully inadequate.

Thanks. The point of sharing was to examine my testing “protocol”. I appreciate and accept that concept “tested” in my example is controversial here to say the least. I have no intention of seeking any converts on cables as the original subject of this thread was “Benefits of using expensive DACs”. If I can get my protocols sharpened and can find the time and willing participants, I may try to test that question with human subjects.

Again, thanks,

kn

 

[knownothing]

here:


But he knew they were being switched and, if I'm not mistaken, which cable it was (by letter or other non-brand identifier)? As indicated in my prior post, that is the big one. The set up and everything else don't really matter when basic controls are lacking. It also helps to have the person switching the cables not know which is which (that's harder).

a few quick ways to add more control:

1) randomize the order of cables, including, from time to time, leaving the same cable in without telling the subject (e.g. in the first trial, cables A and B could be the same one)
2)rename the cables between trials, keeping track of which was which in each trial
3) collect impressions and preference order across trials
4) measure the cables for any damage or anomalies (LCR)

Due to the brevity of auditory memory, the listener will have a much better chance of telling the difference if the time between switching is minimized, so a switching appliance makes for a better test.

Very helpful

 

[olieb]

If you are interested and able, pointing out the specific areas of greatest failure to meet objective standards would be helpful.

I see the problem not so much in the test protocol, but in the basic design.
What exactly is the question that you want to answer with this test?
What do you even consider as the result?
As I see it, the result is that your test person "heard" differences between cables. Well, that is by no means a surprise, people "hear" differences all the time, when they are asked/told to do so (even if there are none).
You seem to consider it a result, that you came to similar conclusions as your test person, by sighted listening.
But this is not at all the question your test can give an answer to.

Summary: I would say you do not have a result, as there is no question.

To begin I would recommend a test design to answer the question
"Can the test person reliably reproduce identify (edited) differences in sound from different cables?"
For this more than two runs are necessary and specific variables ("brightness", "preference", "precision of leading edge", ...) have to be chosen and answered for each run.
Of course you need to randomly change the order of the cables from run to run.
Then check the correlations after a sufficient number of runs. And make VERY sure that the test is really blind. The human brain is extremely good at "feeling" clues to better jump at conclusions, hundreds of millions of years of evolution made it that way.
For this the gold standard is double blind tests.

 

[knownothing]

I see the problem not so much in the test protocol, but in the basic design.
What exactly is the question that you want to answer with this test?
What do you even consider as the result?
As I see it, the result is that your test person "heard" differences between cables. Well, that is by no means a surprise, people "hear" differences all the time, when they are asked/told to do so (even if there are none).
You seem to consider it a result, that you came to similar conclusions as your test person, by sighted listening.
But this is not at all the question your test can give an answer to.

Summary: I would say you do not have a result, as there is no question.

To begin I would recommend a test design to answer the question
"Can the test person reliably reproduce differences in sound from different cables?"
For this more than two runs are necessary and specific variables ("brightness", "preference", "precision of leading edge", ...) have to be chosen and answered for each run.
Of course you need to randomly change the order of the cables from run to run.
Then check the correlations after a sufficient number of runs. And make VERY sure that the test is really blind. The human brain is extremely good at "feeling" clues to better jump at conclusions, hundreds of millions of evolution made it that way.
For this the gold standard is double blind tests.

Also helpful

 

[Wolf11Man]

What I miss is a plain, basic quality cable with a reasonable price range, maybe even DIYS, so you know each component.
Ready made not more than 20 $.
A digital cable with 100% perfect technical function is about 2-5 $ for a 3' length of coaxial cable and 4-8 $ for two high quality plugs. Just as they are used in studios all over the world. Such a cable is not to beat in simplicity, function and cost.

Anyway, the "test" if it ever happened in reality, is as senseless as a test about HDMI or satellite TV cables that are promised to give a better picture, more vivid colors and better sound.
Products only based on brazen lies to sell overpriced wires to technical uneducated folks with too much money. If there is a fool with money in his pocket, there will always be a crook near him to take advantage of.
I can't tell which of the two types "knownothing" is, but his whole test is dripping of snake oil and seems to promote cables that are overpriced. Even his "basic version" is 400% more than a reasonable price would be.
Has nothing to do with DAC's anyway. Complete nonsense to spend more money on a coaxial wire than DAC and amp together.

I could thing about a poll: How many members here are so naive to think that some digital cable could give a "better sound" than a technically, measureable 100% perfect one.

1. Nonsense, digital is OK or broken, nothing to "improve" by elves and unicorns
2, Maybe, cousin Billy told me so when we shared a bottle of booze
2. Sure, there is no limit in sound improvement if they are expensive enough

 

[Mart68]

Again, I freely admit this exercise could be greatly improved to achieve anything more than a slightly more controlled subjective evaluation to inform my own buying decisions. I put it forward here for critique, not to defend.

Here is a link to my current system set up.

knownothing's Music/Library/Theater System | Virtual Listening Room

knownothing's listening room featuring Leviton 20 Amp Isolated Ground Duplex Outlet, Nanotec Systems #308 "Wonderful" Power Cable, Shunyata Research Alpha HC ZTRON, Shunyata Research Venom PS8, Shunyata Research Venom Defender, FURUTECH FP TCS21 cable/Rhodium connector power cord, FURUTECH...

www.audiogon.com

The digital front end has changed since the test, with replacement of the DAC and an external power supply now supports both the Node and the DAC, but it’s physical layout is identical. All the major pieces in this system were bought used, on sale/closeout, or were factory seconds except for the subwoofer. I assembled several of the power cords I am using from bulk cable and purchased connectors.

The subject sat in the center of the couch directly between the speakers on the opposite long wall and did not move from that spot for the duration of the test. I sat near the electronics on the left front portion of the room and had the four different cables laid out behind the DAC and the Node. The power to the Node was left on for the whole test. The power to the DAC was disconnected immediately before the test and between each cable swap. The subject could not see the cables being swapped from where they were sitting. As noted previously, five different cuts were played in sequence for each cable, and then cables were swapped and the five cuts played again. It was up to the subject to determine how much of each cut they wanted to listen to. This entire sequence was repeated so there were two complete listening sessions for the four cables and five cuts. The tests were run at 60-70 decibels average volume as measured in the room, depending on the loudness of the cut played as the volume was held constant throughout.

These were the cuts used for the test:

1 - Sarasate: Carmen Fantasy, Op.25 Moderato, Anne Sophie Mutter, Weiner Philharmoniker, James Levine, “CARMEN-FANTASIE” - 24/88.2

2 - Lay Lady Lay, Cassandra Wilson, “Glamored” - 24/96

3 - Horizontal Blue, David Piltch, Kevin Breit, “Take One” - 24/192

4 - The Spot, Mickey Hart, “Planet Drum” - 24/96

5 - Siren Singing, Yello, “Point” - 24/48

I had said previously that some cuts at 16/44.1 were used in the test. On review, I was mistaken, all cuts used were higher resolution.

The front speakers are about 9ft apart on the long wall of the room and angled slightly towards the listening position. The listening position is about 11ft from the front wall. The subwoofer is positioned along the left wall about 4ft from the front wall. No DSP was engaged other than the settings in the subwoofer itself. The room is insulated on three sides and the ceiling with rock wool and the ceiling sheet rock is floated on aluminum spacers. This has more to do with keeping sound from affecting the rest of the house than acoustics in the listening room. There is a bookshelf covering the entire left wall. The floor is concrete with a large throw rug. The room is a little “dead” at lower volumes, eating some treble, but comes alive in terms of treble at about 65 to 70 decibels. I would say the bass without DSP in the room is “pretty good”. Turning the sub down one of two decibels on high bass content material like electronic or hip hop helps. The sub was not adjusted from its standard setting during the test.

kn

Your room is beautiful, your blind test not so much

Can you test your subject again, this time just between two cables, the one he liked least and the one he liked the most? With proper controls, and all he has to do is reliably tell which one is in play (get it right 9 out of 10 times or better).

Have to say I already know from experience he has no chance, but worth doing for your own peace of mind (and bank balance). Divesting oneself of FOMO is wonderful.

 

[SIY]

Have to say I already know from experience he has no chance,

If the controls aren't done right (and I have serious doubts that they will), I'd say he has an excellent chance.

 

[A Surfer]

We also have to be careful as the host himself may accidentally bias the subjects. Body language, changes in tone of voice can tip a subject off that something is different.

I would literally make sure the subjects are blindfolded, have someone besides the host do the swapping and also make sure that when swapping happens, the subjects can't hear.

I would include some test runs without swapping and also once a good question for the test is decided on, included a confidence rating per trial. So if subject A says there was an improvement in sound between two cables, how confident are they in that judgement.

 

[ahofer]

In addition to fast switching, headphones increase the likelihood of hearing any actual difference. Especially if you think, as the OP’s system description suggests, there is an increase in detail when you change digital sources.

BTW, after looking at that, I hope @knownothing is aware that Chord DACs have much higher output, which is an easy hack for winning poorly level-matched comparisons (see the third quote in my signature).

 

[A Surfer]

Yes, level matching is absolutely crucial.

 

[knownothing]

Your room is beautiful, your blind test not so much

Can you test your subject again, this time just between two cables, the one he liked least and the one he liked the most? With proper controls, and all he has to do is reliably tell which one is in play (get it right 9 out of 10 times or better).

Have to say I already know from experience he has no chance, but worth doing for your own peace of mind (and bank balance). Divesting oneself of FOMO is wonderful.

Thanks for your kind remarks. This has been a fourteen year effort from designing and constructing the room in a new basement to the selection of major components to tweaking everything repeatedly to get the sound I wanted. If an added piece didn’t improve the perceived sound, I got rid of it. I am now very pleased with the results and am done changing anything unless something breaks or I decide I “need” a new feature like 4k capable receiver. And since the subwooofer, receiver, and speakers are all 13 years old, that is a distinct possibility. My prized Panasonic plasma TV quit last year and there were no parts available or service people who would touch it so I got a new well-reviewed 4k set to replace it that is not as pleasing to me. “Progress”.

I can possibly convince the original subject to do another digital cable test. I do not have access to those exact cables but have one similar to the higher end and could grab a $3 cable from the hardware store or Amazon to compare it with, using the improved protocols described by others or linked in this thread. But don’t hold your breath because we both travel a lot for work.

kn

 

[knownothing]

BTW, after looking at that, I hope @knownothing is aware that Chord DACs have much higher output, which is an easy hack for winning poorly level-matched comparisons (see the third quote in my signature).

Yes, I understand level matching will be critical, tested at the output of any DACs used in a comparison. The Chord Qutest has three nominal output settings, 1, 2, and 3 volts. Amir tested the Qutest DAC at the 2 volt setting and registered 2.017 volts on his equipment. FWIW, I run my Qutest at the 3 volt setting.

kn

 

[Imsjry]

Benefits of using expensive DACs are mainly:

• Placebo effect is real and works really well so you might perceive a better SQ where there is none knowing something is recommended in luxury (web)magazines/sites.
• Looks may be important for the owner to get enjoyment (feelgood effect)
• Knowing something expensive may be important for the owner to get enjoyment (pride of ownership,feelgood effect).
• Owning something 'special' may work wonders for someone (pride of ownership, feelgood effect).
• Most benefits go to the manufacturers, importers and retailers and are financial in nature.
• Very few manufacturers actually produce something that measures/performs objectively better. That does not mean it will objectively sound better too but maybe a reason to buy a product.

Feeling good about a product (knowing it is among 'the best' according to [insert reviewers/individuals one trusts] is an important part of enjoyment as is pride of ownership to some.
One gets some 'insurance' it does not get 'better' than that (within or even above their budget).
In a similar way as owning 'objectively good' and 'ASR recommended' gear kind of works (people buying on recommendations and SINAD for instance and buying something new the moment a device has a slightly higher SINAD or a good number one aspires but suddenly falls within budget.

In the end it is a 'dream' one buys. A dream our eyes/brains are perfectly willing to believe.

This place is wild and this post is for sure the silliest thing I'll read on the internet today. Oh wait, is there a blind test I should use to prove it? Of course different Dac's tested using the same source tracks/volume level/cables/streamer/speakers will have audible differences. I know because you can hear them. With your ears.