Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)
post-11512472
Post #4,111 of 6,423

Maxx134

Headphoneus Supremus
Joined
Oct 15, 2010
Messages
4,755
Reaction score
1,829
Location
NYC
Joined
Oct 15, 2010
Location
NYC
Posts
4,755
Likes
1,829
Sorry that I am so late to this party, but with 250+ pages in this thread to peruse I'm just not willing to plow through all of them. However, I'm more than willing to ask the question as to how the comparisons (in this thread) were made. Were they double-blind, or is all of this opinion? If someone can provide some post links in which these matters are addressed, I would be grateful.

What I am most interested in is whether these comparisons were doing using typical double-blind controls when paired comparisons are made (i.e. ABX, Bradley-Terry et al) or if all of this thread is based on uncontrolled tests.

Lastly, does anyone know of any recently-published papers either by IEEE or AES that looks into such matters? That is, a published paper that can be found in a peer-reviewed journal?

Thanks,

Mark (immersifi)
The problem with double blind test is that it removes all strengths clues that you would use to differentiate anything by not using your own equipment and fav familiar test songs as well as placing you mentally under stress with unfamiliar environment so as to distract you .
The whole while you are already in assumption of doing a test that promotes not hearing anything and thus you mind is already bias and its active filter is mentally closed...

Anf for those that have sensitive eyes,
I don't use spellcheck so please forgive any typos from my smartphone..
:p

The bottom line is you gotta believe somebody and what I do is accept an opinion once a general consensus made over time with many users...
that and whatever Purrin says, lol
:p
 
     Share This Post       
  • Like
Reactions: Argo Duck
post-11512771
Post #4,112 of 6,423

Articnoise

1000+ Head-Fier
Joined
Sep 4, 2013
Messages
1,480
Reaction score
490
Location
Sweden
Joined
Sep 4, 2013
Location
Sweden
Posts
1,480
Likes
490

If conducting a study, a blind or A-B test, it is important to reduce all the factors that can influence the participants in one way or the other. A more experienced participant is normally less infected to some of the factors, but never immune.

1 If only listening for 1-3 track in an unfamiliar system and then changing gear it is less likely that one will find it very fatigue or spot other minor flaws that is more obvious in longer listening sessions.

2 If only listening for 1-3 track in an unfamiliar system and then changing gear back and forth, it is less likely that one will really feel and connect to the music (this is SQ aspect) a as much as in a more relaxed situation.

3 If only listening for 1-3 track in an unfamiliar system and then changing gear back and forth, one is more likely to prefer a more crispy sound that stands out and that has the wow factor than in a longer listening sessions.

4 If only listening for 1-3 track in an unfamiliar system it is hard to conclude if the tonality of the specific tested gear is natural, neutral, lean etc. only if the whole synergy of the system sound that way.

5 If only listening for 1-3 track and then changing gear one is more likely to prefer a sharper and more present sound, especially after a couple of hours.  

6 The synergy between different gears and their tolerance to pure/dirty power etc. are important aspects that can infect some gear more than others.

7 The music chosen is important as some gears do some categories (bass, vocal, symphony etc.) of music better than others.  

I agree with Maxx. To conducting the same short blind test in a familiar system and with well-known music is so much easier and one can normally detect even minor difference between the tested gears and also find the distinguishing difference in both tonality and technicality.

IME.

 
     Share This Post       
  • Like
Reactions: Maxx134
post-11513092
Post #4,113 of 6,423

Ableza

Headphoneus Supremus
Joined
Jan 18, 2014
Messages
3,976
Reaction score
12,318
Location
Valley of the Sun
Joined
Jan 18, 2014
Location
Valley of the Sun
Posts
3,976
Likes
12,318
A blind test is never necessary to determine a preference.  It is only necessary when trying to determine a reason for the preference, or when trying to ascribe causation (causation that is not found through electrical parameter testing, that is.)
 
For example, me saying that I like the new Whammo Super Bajinga amplifier more then my old Gizmodo Flipperoo amplifier is a perfectly valid statement and the reason why doesn't matter at all.  I like it more, that's all that matters.  It's a preference based on my total personal perception of the Whammo.  But if I want to determine which one sounds better based solely on the signal output, then I really must remove all other variables - and get a consensus opinion from more people than just myself - for that judgement to be valid.
 
     Share This Post       
post-11513195
Post #4,114 of 6,423

Articnoise

1000+ Head-Fier
Joined
Sep 4, 2013
Messages
1,480
Reaction score
490
Location
Sweden
Joined
Sep 4, 2013
Location
Sweden
Posts
1,480
Likes
490
  A blind test is never necessary to determine a preference.  It is only necessary when trying to determine a reason for the preference, or when trying to ascribe causation (causation that is not found through electrical parameter testing, that is.)
 
For example, me saying that I like the new Whammo Super Bajinga amplifier more then my old Gizmodo Flipperoo amplifier is a perfectly valid statement and the reason why doesn't matter at all.  I like it more, that's all that matters.  It's a preference based on my total personal perception of the Whammo.  But if I want to determine which one sounds better based solely on the signal output, then I really must remove all other variables - and get a consensus opinion from more people than just myself - for that judgement to be valid.

 

I don’t think that the reason for doing a blind test is to determine a reason for the preference. It is to prevent unwanting bias to influence the result. It is common in many diffrent studies.

 
     Share This Post       
post-11513295
Post #4,115 of 6,423

kugino

Headphoneus Supremus
Joined
Nov 15, 2003
Messages
5,742
Reaction score
91
Joined
Nov 15, 2003
Posts
5,742
Likes
91
The balanced digital signal should be one twisted pair.  Since CAT-5 is unshielded you will have to use one or more of the other wires/pairs in place of the shield.  If the connection is short enough the shield connection may not be vital.
Note: I have not tried this in practice, as I don't have anything that outputs AES/EBU balanced digital to my Adcom GDA-700.  By the way, the match is close, but not exact: CAT-5 is 100 ohm and AES/EBU is 110 ohm. Should be close enough.

One thing to note about the sound of the Adcom GDA-700 DAC is that the sound will probably be dominated by the PMD-100 digital filter and the Adcom op-amps (6AA = AD711) which have class-A biasing but may "smooth" the sound a bit.
interesting tidbit about the adcom. I have the 700 coming in later this week. isnt it true of most DACs that the sound is determined a lot by the filter? isn't one of the main reasons the theta sounds so smooth is due to the super burrito filter that moffatt designed?
 
     Share This Post       
post-11513340
Post #4,116 of 6,423

haywood

New Head-Fier
Joined
Nov 24, 2010
Messages
45
Reaction score
26
Joined
Nov 24, 2010
Posts
45
Likes
26
 



I don’t think that the reason for doing a blind test is to determine a reason for the preference. It is to prevent unwanting bias to influence the result. It is common in many diffrent studies.
However valid the reason for double-blind testing is it has nothing to do with this thread which is titled "thoughts on a bunch of DACs" and not "a definitive study on listener preference as it relates to music and DACs" or whatever.

Now... back to DACs. Please?
 
     Share This Post       
post-11513399
Post #4,117 of 6,423

Articnoise

1000+ Head-Fier
Joined
Sep 4, 2013
Messages
1,480
Reaction score
490
Location
Sweden
Joined
Sep 4, 2013
Location
Sweden
Posts
1,480
Likes
490
However valid the reason for double-blind testing is it has nothing to do with this thread which is titled "thoughts on a bunch of DACs" and not "a definitive study on listener preference as it relates to music and DACs" or whatever.

Now... back to DACs. Please?

 

I will very soon have a new tube dac to play with. Just wonder how I should conduct the test against the Master 7; blind, double blind, triple blind or atx LOL!

 
     Share This Post       
post-11513454
Post #4,118 of 6,423

icebear

Headphoneus Supremus
Joined
Jan 9, 2014
Messages
1,640
Reaction score
359
Joined
Jan 9, 2014
Posts
1,640
Likes
359
I take a wild guess here :
Only a very small minority of customers buys a hifi product blind (any product for that matter).
 
So what's the purpose of double blind or whatever study in the context of discussing hifi gear?
There is the sound science forum for the inclined members

 
Of course having validation of some sort, preferably with numbers black on white makes every argument much easier. So comparisions often try to settle on a number of good practises. One of them SPL matching. Hoenstly if the louder one sounds better, why just not kick it up a notch and be happy? I never use any SPL at home, I just adjust the volume to match the music to my liking and mood.
 
Ultimately if you buy a product e.g. a DAC and you like it, just be happy and enjoy listening to music. There no absolute best that is valid for everyone. Usually everbody has his/her own set of ears and music preference. It's no surprise that there are different opinions about gear and there will be no study that will reveal the ultimate truth in DAC's, although it can only be DAC42

 
     Share This Post       
post-11513466
Post #4,119 of 6,423

Argo Duck

Formerly known as "AiDee"
Joined
Nov 6, 2007
Messages
2,245
Reaction score
416
Joined
Nov 6, 2007
Posts
2,245
Likes
416
Mark, you didn't offend at all. Coming out of my field my points may have been a little obscure. I'm not putting it very well here but...the question is whether 'subjective judgment' or opinion has any validity or truth-power. If it does, even marginally, then forums that share opinions are useful in steering us toward better choices, even if only slightly., i.e. we have better utility. The 'control' case is to choose one's equipment blind - through random selection - with perhaps a sideways glance at the pitifully few objective measures published by equipment manufacturers. Let's say we set a lower limit before gear qualifies for our 'selection pool'.

Where this comes right into my area of interest is to what extent a community sharing opinions (messages we tend to call them) enhances its functioning or survival. Specialist communities like head-fi are not different in this respect. Increasingly unusual in this internet age though is the merging of multiple and disparate communities and groups with all kinds of other vocabularies and 'understandings'. Critical to a group like head-fi is to what extent is our 'shared language' - with its specialist terms such as 'dry', 'wet', 'warm' etc - is rooted in a shared experience. Probably not much - though informal groups like purrin's and mini-meets as well are strong steps to something better - but is it enough anyway to achieve greater validity than would be the case from an entirely solo audiophile journey?

I don't pretend to have any answers. A couple of years in the sound 'science' forum convinced me there's nothing interesting going on there. Had some research ideas of my own but I my expertise is too peripheral, the task too large and I don't have the time or inclincation anyway!

In the end I pursue this hobby to relax :wink:

And btw welcome.

Social dynamics. Very interesting stuff (I say that in earnest). Good point about overlap - I can give you one such example, namely, when we do the listening trials, we try to visually isolate (if possible) the jurors' points of view. That is, just like back in grade school, we don't want people looking at their neighbor's keypad to see which key is being pressed. What I don't know about social dynamics though is a great deal. It's interesting how often such simple tests - and in paired comparison stuff we're asking jurors which they prefer, nothing more - how the juried study can impart anxiety in some subjects. We try never to call them "juried tests" because the second word can impart anxiety (we simply refer to them as "listening sessions", and anxiety (as I'm sure you know) can cloud one's ability to judge.

I'm not sure specifically which of your points that I missed, but I'll go back and re-read your posts; I hope that I didn't offend. My reply (ies) was (were) more than likely due to my focus and background, and I hope that I didn't come across as heavy handed or rude.

Mark
 
     Share This Post       
post-11513469
Post #4,120 of 6,423

jacal01

You've just been blocked by Mr. Conviviality!
Joined
Dec 1, 2011
Messages
1,459
Reaction score
540
Location
Slightly left of bubble
Joined
Dec 1, 2011
Location
Slightly left of bubble
Posts
1,459
Likes
540
I'm getting the impression that the eye candy fanboy crowd is a larger percentage of this forum than I'd given them credit for.  Gotta look good in your living room or desktop.  SQ evaluations?  Pffft.
 
     Share This Post       
post-11513522
Post #4,121 of 6,423

Stillhart

CanJam@RMAF 2015 Karting Champion
Joined
Feb 14, 2014
Messages
6,638
Reaction score
3,193
Location
Las Vegas
Joined
Feb 14, 2014
Location
Las Vegas
Posts
6,638
Likes
3,193
   
The balanced digital signal should be one twisted pair.  Since CAT-5 is unshielded you will have to use one or more of the other wires/pairs in place of the shield.  If the connection is short enough the shield connection may not be vital.
Note: I have not tried this in practice, as I don't have anything that outputs AES/EBU balanced digital to my Adcom GDA-700.  By the way, the match is close, but not exact: CAT-5 is 100 ohm and AES/EBU is 110 ohm. Should be close enough.
 
One thing to note about the sound of the Adcom GDA-700 DAC is that the sound will probably be dominated by the PMD-100 digital filter and the Adcom op-amps (6AA = AD711) which have class-A biasing but may "smooth" the sound a bit.
 
 
 
Thanks, this is super handy.  I'll give it a shot and see what happens.
 
I also found out that my Theta (which gets here tomorrow!!!) doesn't have AES3, just coax and some old style optical/BNC connector.  Oh well, at least it has balanced out.
 
Regarding the sound of my 600, the biggest part of the mod that I'll be doing is replacing the opamps with OPA637.  Apparently that makes the biggest difference and all the cap upgrades are minor in comparison.
 
     Share This Post       
post-11513595
Post #4,122 of 6,423

Tuco1965

Headphoneus Supremus
Joined
Dec 28, 2012
Messages
4,545
Reaction score
184
Joined
Dec 28, 2012
Posts
4,545
Likes
184
  It's taken me a long time to come to this realization--because I was once of the belief there is a "best" component that existed somewhere in the universe--that there's nothing wrong with placebo effect. We do this hobby for enjoyment. Music moves us. So, if component A is bright and shiny and costs more than dingy component B, there's nothing wrong with believing A sounds better than B. Do what brings happiness and enjoyment to the hobby. We do this with other hobbies in our life--clothes, cars, food, wine, vacations, etc. Audio should be no different.
 
The problems arise when someone pontificates that A is better than B, and you're not a skilled listener if you can't tell them apart and come to the same conclusion. Purrin was pretty clear he said "his preference" in "his system" to avoid pontification.
 
That's the problem that bothers me.  My ears know what I like, just as your ears know what you like.  We all have preferences.
 
     Share This Post       
  • Like
Reactions: hodgjy
post-11513700
Post #4,123 of 6,423

conquerator2

Headphoneus Supremus
Joined
Jul 19, 2012
Messages
5,638
Reaction score
1,280
Location
Czechia
Joined
Jul 19, 2012
Location
Czechia
Posts
5,638
Likes
1,280
Website
www.youtube.com
   
Thanks, this is super handy.  I'll give it a shot and see what happens.
 
I also found out that my Theta (which gets here tomorrow!!!) doesn't have AES3, just coax and some old style optical/BNC connector.  Oh well, at least it has balanced out.
 
Regarding the sound of my 600, the biggest part of the mod that I'll be doing is replacing the opamps with OPA637.  Apparently that makes the biggest difference and all the cap upgrades are minor in comparison.
#ThetaHype 

 
     Share This Post       
post-11513709
Post #4,124 of 6,423

murrays

100+ Head-Fier
Joined
Jan 10, 2009
Messages
302
Reaction score
42
Joined
Jan 10, 2009
Posts
302
Likes
42
B
Thanks, this is super handy.  I'll give it a shot and see what happens.

I also found out that my Theta (which gets here tomorrow!!!) doesn't have AES3, just coax and some old style optical/BNC connector.  Oh well, at least it has balanced out.

Regarding the sound of my 600, the biggest part of the mod that I'll be doing is replacing the opamps with OPA637.  Apparently that makes the biggest difference and all the cap upgrades are minor in comparison.

Be very careful using the OPA637 if you don't have an oscilliscope to confirm their stability. The OPA637 is not unity gain stable and may be prone to HF oscillations. It would be safer to use the related unity gain stable part, which is the OPA627.
Do you have the GDA-600 or GDA-700?
 
     Share This Post       
post-11513726
Post #4,125 of 6,423

Stillhart

CanJam@RMAF 2015 Karting Champion
Joined
Feb 14, 2014
Messages
6,638
Reaction score
3,193
Location
Las Vegas
Joined
Feb 14, 2014
Location
Las Vegas
Posts
6,638
Likes
3,193
B
Be very careful using the OPA637 if you don't have an oscilliscope to confirm their stability. The OPA637 is not unity gain stable and may be prone to HF oscillations. It would be safer to use the related unity gain stable part, which is the OPA627.
Do you have the GDA-600 or GDA-700?
 
Actually, you're right.  I just double checked and I misspoke, it was the OPA627.  I was typing from memory.  

 
I'm modding the 600.  Here's the thread I'm using for reference on the mod.  I've read through the whole thing twice, checked all the photos, and now I'm trying to put together all the parts...which I'll probably have to source from 3-4 different places.  DIY is a crazy journey...
 
     Share This Post       

Users Who Are Viewing This Thread (Users: 1, Guests: 3)

Top