Serious Question: How can DAC's have a SOUND SIGNATURE if they measure as transparent? Are that many confused?

[ahofer]

That’s fine. The other factors are in human’s brain, definitely there are no other factors in the electric signal.

Or in the very limited physical ability of human ears to discern a small difference in acoustic signal.

 

[solderdude]

As such, the ABX has generally a "human" as the detecting element. So bottom line, we need to calibrate our ABX test too.

To me ABX is to have the ability to compare X to A and B (and when one made up their mind chooses A or B as being X.
I don't care about "Audio ABX" or whatever company.
There is no need to calibrate anything. You either compare an unknown (X) to A and B and decide. This can be preference as well when you have a preference for B or A.

Maybe we're talking about a different ABX. For me it is the method.

I do not claim that the general ABX methodology is inherently flawed or problematic, only this specific subset, which is however what is normally meant in conversation here..

It can be used for preference as well. When you switch between A and B and have a preference for A or B you just enter the preference. Statistics do the rest.
Might require some hardware (relays switching outputs).

Did I claim I was being "scientific"? I used certain tools provided by science to provide a foundation for an engineering decision. The test was formulated hence to provide this kind of result. Controls were applied to make sure that the test results were reliable, or as Richard Feynman says, we do not fool ourselves, which is the easienst thing to do.

You made no such claim. You just 'went' with a found preference. The way you describe it has an 'air' of being scientific.

I defend my observation, the right to make my observation and to write: "I made this observation in such and such a manner and this what I observed."

I have no problems with that. I object to the conclusion that 140 SINAD and 120 SINAD would be the reason for the 140 to be 'boring' and by extension that technical excellence (signal fidelity) and preference do not do hand in hand. Without giving any reason other than... we found. This creates the idea for readers that, because you stated this, only the gear in question that does not have the highest signal fidelity is preferred, specifically your design.

 

[Thorsten Loesch]

Friendly advice: if you want to tell fairy stories, your life here will be Hobbesian: nasty, brutish, and short. Fairies may fly, but fairy tales don't.

I would consider this rather threatening, if addressed to me. I do not think this sort of language is appropriate.

I suggest you tone it down a notch or two.

Thor

 

[Thorsten Loesch]

If you're not interested in evidence and getting to the reality of issues, why in the world would you participate in a science-based forum?

From what I can tell, he is not participating in a "science based" forum.

I am still waiting for scientific evidence regarding audibility limits of the items measured in reviews here.

I am waiting for someone to demand scientific evidence for Bruno Putzeys "adding 60dB of NFB made a Tube sound immaculate" claim.

I am waiting for so much positive science, all that comes up is hot air and "Dog ate that homework but we have this here ABX test Chair and it's all we need."

I am waiting for someone to actually show what specific differences have been reliably detected with ABX tests, as in most or a majority or a preponderance even.

An AP2 dashboard is NOT science. Defending a discredited test as non plus ultra of testing for perception is NOT science.

So where is the science please? Anyone got any?

Thor

 

[Purité Audio]

Is it pantomime season?
Keith

 

[Thorsten Loesch]

To me ABX is to have the ability to compare X to A and B (and when one made up their mind chooses A or B as being X.
I don't care about "Audio ABX" or whatever company.
There is no need to calibrate anything. You either compare an unknown (X) to A and B and decide. This can be preference as well when you have a preference for B or A.

Maybe we're talking about a different ABX. For me it is the method.

Here the wikipedia page:

https://en.wikipedia.org/wiki/ABX_test

I would expand the potential flaws and alternatives section. But it's generally a fair outline.

More here:

http://djcarlst.provide.net/abx.htm

You will also find a results database. Rather old and outdated.

I suggest to perform a meta analysis of ALL "same" tests and see if taken as an overall grouping. Easily done in Excel or Google sheets which have the relevant statistical tools available.

Does a statistical analysis of the larger set of results still support the "same" conclusion, which it should if all results were really random? If on the other hand the meta analysis fails to support the null hypothesis, what could be the root cause for the difference?

Thor

 

[Thorsten Loesch]

I object to the conclusion that 140 SINAD and 120 SINAD would be the reason for the 140 to be 'boring'

So would I. I would describe this a case of parallel metrics, one metric was subjects preference of devices which VERY reliable and the other the observable in an AP2 dashboard similar to what does Amir does (I generally use Stereophile's format for obvious reasons).

My conclusion was the specific measures taken in the circuit design and execution of similar but subtly different circuits manifested as a reliable listening preference in one metric and differences in (T)HD in another metric.

and by extension that technical excellence (signal fidelity) and preference do not do hand in hand.

I think we already had a number of anecdotes stated that certainly support, independently from my own experiment, the hypothesis that there is no reliable link between objective signal fidelity (greater measurable absence of specific fidelity impairments) and listening preference. This hypothesis far predates my experiments and I was somewhat unaware that it had been reliably disproven and was under the impression it was still solidly in the running.

There is a subversion of this hypothesis I dislike but will NOT exclude, namely that preference for an "impaired fidelity" version of the signal is always a result of the fidelity impairment and not exists in spite of it.

I mean I prefer my Broccoli or Cauliflower boiled soft in saltwater and covered with breadcrumbs and having superhot clarified butter poured over all of it. That's a lot of "fidelity impairment" compared to the raw vegetable (which I also like in a salad with a nice dressing after being soaked a day or two in saltwater) but I like it anyway and until my doctor forbids butter I will eat it like that and when the doctor forbids butter I will change the doctor.

Without giving any reason other than... we found. This creates the idea for readers that, because you stated this, only the gear in question that does not have the highest signal fidelity is preferred, specifically your design.

Well, that is NOT what I wrote.

I am sorry, but I'm not taking responsibility for a lack of basic reading comprehension among self proclaimed "scientists".

As you know I am originally German, as Dutch (IIRC?) you should know that when I write something I write it as exactly and precise as possible and it means exactly what is written and ONLY what is written.

Thor

 

[Reynaldo]

Friendly advice: if you want to tell fairy stories, your life here will be Hobbesian: nasty, brutish, and short. Fairies may fly, but fairy tales don't.

Wonderful advice.
Thank you for his polite words.

 

[solderdude]

I am sorry, but I'm not taking responsibility for a lack of basic reading comprehension among self proclaimed "scientists".

Nowhere did I ever claim I was a 'scientist' nor a scientist.
I am a 'lowly' engineer/designer/repair guy. Not a self-proclaimed 'scientist'.
I am just challenging some of the things you write using common sense and practical experience.

A bit disappointed (t.b.h.) in your assertion which is not based on any facts other than how you prefer to see things, seeing as it must be addressed to me as it was a reply.

Here the wikipedia page:

it says: An ABX test is a method of comparing two choices of sensory stimuli to identify detectable differences between them.

A preference means a difference is identifiable is it not, otherwise how can there be a preference?
If one uses ABX test for preference and takes 10 mins to listen to A and B then why would that be invalid.

You may object to a quick ABX between 2 identical signals or nearly identical.
When you say there is a clear preference then it is audible, but may not be so using quick switching for certain reasons.
That may well be so but then use the ABX protocol based on preference and not on difference as a task. No need to set time limits to force a decision. That would not invalidate ABX test.

I am not a scientist, nor will I do meta analysis of the theory behind it with false negatives, forced decisions etc.

Well, that is NOT what I wrote.

I didn't quote you so it is my 'personal interpretation' of many of the replies to many people combine here and just as (in)valid as any 'interpretations' of yourself and other members.

You have already stated that your findings were very conclusive and the decision to build something was made on it. That is fine to do so. I have no objections.
What I find 'unpleasant' is the whole attitude and conclusions surrounding this considered 'fact'.

 

[ahofer]

Wonderful advice.
Thank you for his polite words.

So he's speaking for a lot of people. I more or less agree, although eventually I just put the "magical claims" people on ignore. But they catch a lot of flack, it is true.

Ultimately it is up to @amirm and his moderators whether untested subjective claims are encouraged or allowed here, it is his site. However, it is pretty clear that the majority of the frequent commenters prefer to have at least one place on the Web where the discussion is less cluttered with that stuff.

 

[Reynaldo]

So he's speaking for a lot of people. I more or less agree, although eventually I just put the "magical claims" people on ignore. But they catch a lot of flack, it is true.

Ultimately it is up to @amirm and his moderators whether untested subjective claims are encouraged or allowed here, it is his site. However, it is pretty clear that the majority of the frequent commenters prefer to have at least one place on the Web where the discussion is less cluttered with that stuff.

Here one should be writing about iFi.

I just wrote that I bought the DAC and I like it.
That I don't need a scientific explanation to like it.

Most of the people who frequent this forum want to know about the quality of the devices, their problems, etc.

I have nothing against discussing scientific data, I just wrote that I liked the quality of the device.

 

[ahofer]

I just wrote that I bought the DAC and I like it.
That I don't need a scientific explanation to like it.

You said you "liked the sound quality", not just features and build quality. That implies, directly, that you could tell it apart from other DACs by sound quality.

 

[solderdude]

It could simply be that he just liked how it sounded and wanted to just give his opinion ?
That's how I read it.
He prefers it over other DACs. There could be a myriad of reasons for that.

 

[Reynaldo]

You said you "liked the sound quality", not just features and build quality. That implies, directly, that you could tell it apart from other DACs by sound quality.

Now do I need to scientifically prove that I like the sound quality?
One question, have you heard this DAC?
Which DAC do you use?
Do you like or dislike your DAC?

 

[Reynaldo]

It could simply be that he just liked how it sounded and wanted to just give his opinion ?
That's how I read it.
He prefers it over other DACs. There could be a myriad of reasons for that.

Exactly, here is to write about this DAC.
I bought it and I liked it, simple as that.

 

[lateralous]

Exactly, here is to write about this DAC.
I bought it and I liked it, simple as that.

Forget this question of SINAD numbers.
Buy an iFi and be happy.
I have the ZEN One Signature with an extremely nice sound.

It seems to me your very first comment on this DAC has gone well beyond a simple statement of "I like this thing".

 

[Reynaldo]

It seems to me your very first comment on this DAC has gone well beyond a simple statement of "I like this thing".

There was a person asking about this unit and another DAC.
So as I liked it I gave the suggestion.
Even because I think this device has not been evaluated.

For you, does having a high number on SINAD mean indisputable quality?

 

[Thorsten Loesch]

Nowhere did I ever claim I was a 'scientist' nor a scientist.

Correct. And your reading comprehension seems ok as well. You state how things COULD BE interpreted. And you are quite conservative and reasonable.

A bit disappointed (t.b.h.) in your assertion which is not based on any facts other than how you prefer to see things, seeing as it must be addressed to me as it was a reply.

You very tamely echoed what has been thrown around a lot. My comment was more general and not personal. I also didn't read what you wrote as a literal expression of a view you actually literally hold as expressed.

It says: An ABX test is a method of comparing two choices of sensory stimuli to identify detectable differences between them.

Did you read the whole article?

I am not a scientist, nor will I do meta analysis of the theory behind it with false negatives, forced decisions etc.

Well, if as engineer you want to know (I did) the meta analysis is interesting.

What I find 'unpleasant' is the whole attitude and conclusions surrounding this considered 'fact'.

What attitude? What conclusions?

I posted conclusions about other things and subjects, like Audio ABX testing, but not on my observations.

I have zero interest to debate these things. But instead of simply ignoring my observations I get called out and even attacked personally (not by you) and it is tried to simply drown out voices found disagreeable not with facts or evidence, but just noise.

I could claim that as unpleasant, I fail to see however how I am being unpleasant in any way. Direct and forward yes.

Thor

 

[lateralous]

There was a person asking about this unit and another DAC.
So as I liked it I gave the suggestion.
Even because I think this device has not been evaluated.

For you, does having a high number on SINAD mean indisputable quality?

Of course not, but it's also not a parameter I would 'forget entirely' when making a purchasing decision on a DAC.

 

[HarmonicTHD]

There was a person asking about this unit and another DAC.
So as I liked it I gave the suggestion.
Even because I think this device has not been evaluated.

For you, does having a high number on SINAD mean indisputable quality?

You generalized your individual personal preference - that is the key point of this recent discussion - and for making such a generalizing remark, I would want to see evidence (not so about a statement where you would have said - I like it others might not).