Kinda beating a dead horse but... (non blind level-matched DAC comparison)

[rkbates]

I didn't want to spell it out, but for me I strongly avoid products made in China (and certain other labor markets) largely based on human rights issues... I accept that some things I cannot avoid but when I can, I always do.

...but then in some countries working in a factory for $1 a day is still better than starving. No easy answer.

 

[abdo123]

I didn't want to spell it out, but for me I strongly avoid products made in China (and certain other labor markets) largely based on human rights issues... I accept that some things I cannot avoid but when I can, I always do.

Mad respect tbh, most people here don’t give a damn about their actions because they think they’re past the age of giving damns.

 

[Vasr]

No, not any and all types of conclusions, but just some conclusions concerning the audibility or otherwise of measured differences between well-performing DACs.

But that is where the problem is. There is a lot of body of work but to make a claim, one has to point to the exact study and the logic that makes the inference for the particular claim. Otherwise, it is as meaningless as saying "there is Physics, go look it up, my claim is proved there".

Layout the exact argument with the research that supports it and then that argument can be analyzed for correct assumptions and logic. If it passes then the claim is valid. I have not seen such an argument for this specific case yet.

The arguments here are very much like the Boglehead discussions. They will point at some research and say all that is already proved and make all kinds of assertions about active management but when you look at it in detail, not all of those claims are supported by the studies.

For example, a typical Boglehead argument...
Active management is worse than indexing and no active managers can beat an indexed funds.
Why do you say that? There are many active managers who outperform index funds.
There are studies that show active managers cannot consistently beat index funds.
These studies show that an active manager cannot get higher performance every year but here is the manager that outperforms much higher when he beats than underperforms when he lags in some year. So, although this would fit the conclusions of the study that he is not beating the index every year, over a market cycle or a long enough period he has built up a sufficient lead over the index fund.
But that is because he must be investing outside his category into higher beta stocks.
But isn't that why you pay in active manager to select funds that pay-off.
No, but then it cannot be compared to this index fund.
So what you are saying is that if the fund manager invests exactly in the same fund as the index fund, he would under-perform because of higher fees?
Exactly.
But that is stupid. Then the study does not say anything of practical value.
You are stupid to invest in active funds. Go away...

The above is the problem in these kinds of ideological and pseudo-scientific arguments

If I tell you, "I can jump 100m into the air (unassisted)", should you accept my claim?

Why would I even if there are other peole who could. I have no idea if you could.

Do you insist that the counterclaim "That is impossible" is merely conjectural? Do you insist that my claim is "perfectly fine" and that any counterclaim "suffers from a lack of conclusive evidence"?

The counterclaim is conjectural unless backed by sufficient proof that it is impossible for anybody to jump that high. In the same way your claim would remain of very little validity unless accompanies by proof that you could.

I think this where people misunderstand science and confuse it with a court of law (far too many TV shows about law and order and not enough about science). The latter is adversarial and has a default assumption with the concept of a burden of proof on one or the other relative to the default.

Science does not work that way and this really needs to be understood. Science works by consensus. The "burden of proof" is on anyone who makes a claim and that proof is accepted by "consensus" based on an examination of the assumptions/axioms and the logic used in it.

In science, it is perfectly fine to have no established conjecture/theory either way. So, it doesn't become something is true if the opposite cannot be proven. This misconception is at the bottom of the statements you are making.

It's not the "same thing here" at all. Psychoacoustics is not some fledgling field that just began yesterday with a spot poll of ten people on the street.

Counting is part of an a large and ancient system of Mathematics.

But seriously, just saying Psychoacoustics like brandishing a bible is not a scientific argument. You need to provide a logical proof/line of thought of which parts of those studies apply here so it can be judged on its own. Otherwise, it could be as silly as the Boglehead argument above.

You may have a conjecture that nobody can hear the difference between two DACs that measure similarly or pass some threshold with a specific and finite set of measurements. You likely believe that this is because of a study that says human hearing cannot differentiate between two things that differ in only those. But here is the problem that I have already written earlier which you do not seem to have read/understood. For that claim to be valid, you would need to establish the proof of necessary and sufficiency of that set of measurements to affect what people hear or differentiate on , establish the limits of measurement and difference in measurement outside of which it is possible to differentiate and the applicability and statistical validity of the studies that you use in support to support this exact claim. Just thumping the bible is not going to cut it, at least if one is interested in science. Because all kinds of claims can be made which may or may not follow logically from the studies.

The burden of proof in science (as opposed to court of law) is very clear. Anyone making a claim, provides the proof for it. If there is a consensus after clearly defining the proof (no, number of "likes" in a forum to bible-thumping Psychoacoustics is not consensus, that is an echo chamber) that can be studied.

I'm very happy to argue for the statistical validity of psychoacoustics research.

What is more important is relating the actual study and its conclusions to support a logical line of thought as explained above.

What is your novel audibility claim, specifically? (I'll need a specific claim if I'm going to try to use research to refute it.)

This is where the binary thinking impedes. You believe that because I have not accepted a particular claim, I must have an alternate competing claim.

As I have stated before, there are two possible conjectures - one that nobody can hear the difference between competent/similar measuring DACs and the other some people may be able to. I am perfectly fine with neither of them having been proven.

If you want to prove either of them purely empirically, then it is difficult to prove the first one because it is not feasible to do so over the entire universe (unless supported by other arguments and you are free to make them). The second one can be proven by a single incident of a person being able to so distinguish in a controlled and repeatable experiment. I have not seen anyone do that either. So neither are proven in my opinion and I am OK with that. And hence my first comment on beating dead horses.

Problem is, people are uncomfortable with that uncertainty and non-binary situation and confusing their conviction with proof.

Science.does.not.work.that.way.

 

[gn77b]

To see the distinctions more clearly, let us say we ask one person on the street if they saw a unicorn that day. The person answers no. You ask him on 10 or 100 consecutive days. The answer is still no.

Tu put your analogy in context, in the audiophile press, in the online discussion everyone sees a few unicorns each day. Me and me friends thought we saw unicorns too, but when really looking (not for months or years, admittedly) we realized we couldn't see them.

Now, when we'll meet people claiming they saw unicorns just the other minute we'll be skeptical.

One thing that I noticed is that the manufacturers seem to all move now to streamers and maybe it's because they realized that people are beginning to understand that nowadays the technology can get you 99% of perfect sound on the cheap and they can get more money from features compared to claimed sound improvement. I speculate it's the same with SACD, DVD-A, DSD, MQA. They invented something new (not worse in any way, of course) because they couldn't get money anymore from "improved" sound out of old Redbook.

 

[andreasmaaan]

But that is where the problem is. There is a lot of body of work but to make a claim, one has to point to the exact study and the logic that makes the inference for the particular claim.

Yes, it is a lot of work. That's precisely why I'm asking you to put forward a claim. I'm willing to do that work for you.

The arguments here are very much like the Boglehead discussions. They will point at some research and say all that is already proved and make all kinds of assertions about active management but when you look at it in detail, not all of those claims are supported by the studies.

I'm sorry but this is completely lost on me, as I have no background in business management (???) or finance (???) and have never heard of "Boglehead".

The counterclaim is conjectural unless backed by sufficient proof that it is impossible for anybody to jump that high. In the same way your claim would remain of very little validity unless accompanies by proof that you could.

I think this where people misunderstand science and confuse it with a court of law (far too many TV shows about law and order and not enough about science). The latter is adversarial and has a default assumption with the concept of a burden of proof on one or the other relative to the default.

Perhaps "people" do confuse it in this way. I don't know.

Why don't you make a novel claim that I can then attempt to refute using models and evidence? This discussion is not only highly abstract, but in addition to that it seems to me that you're arguing with someone other than me.

But seriously, just saying Psychoacoustics like brandishing a bible is not a scientific argument. You need to provide a logical proof/line of thought of which parts of those studies apply here so it can be judged on its own. Otherwise, it could be as silly as the Boglehead argument above.

Give me a claim and I will do exactly this. I'm only citing psychoacoustics in general because no claim has been made in the specific.

As I have stated before, there are two possible conjectures - one that nobody can hear the difference between competent/similar measuring DACs and the other some people may be able to. I am perfectly fine with neither of them having been proven.

And I am not fine with the latter. Because the science tells us it is extremely improbable.

Let me ask you this: How well-versed are you in the field of psychoacoustics?

If you want to prove either of them purely empirically, then it is difficult to prove the first one because it is not feasible to do so over the entire universe (unless supported by other arguments and you are free to make them). The second one can be proven by a single incident of a person being able to so distinguish in a controlled and repeatable experiment. I have not seen anyone do that either. So neither are proven in my opinion and I am OK with that. And hence my first comment on beating dead horses.

No, it is a starting limitation of science that it will never prove that anything doesn't exist. That question is not even interesting.

So let's not pretend this is about "proof" now. This is about statistical probability, just like the claim that no human can jump 100m into the air unassisted.

Problem is, people are uncomfortable with that uncertainty and non-binary situation and confusing their conviction with proof.

People this.... people that....

Let me put it this way: If your sound system's maximum output is 100dB SPL peak, the ambient noise level in your living room is 30dB, and your normal listening level is 70dB average, it is extremely unlikely (for everyday purposes "impossible", in the same way jumping 100m or seeing a unicorn is "impossible" even though it cannot be universally disproven by science) that you can hear differences between two DACs that measure flat through the audio band and have SINADs of 100+dB.

 

[M00ndancer]

Let me put it this way: If your sound system's maximum output is 100dB SPL peak, the ambient noise level in your living room is 30dB, and your normal listening level is 70dB average, it is extremely unlikely (for everyday purposes "impossible", in the same way jumping 100m or seeing a unicorn is "impossible" even though it cannot be universally disproven by science) that you can hear differences between two DACs that measure flat through the audio band and have SINADs of 100+dB.

I have un-curable cancer. Can anyone predict with any certainty that I will have a normal life span and not die from something else long before the cancer gets me? Absolutely not. The only thing I know with some degree of confidence is that I will die from my cancer first, but not when. If I look at statistics about my condition, it only states probabilities.

The thing that I think that we as humans have problems with is that there are infinite number of cases where it's NEVER black or white. Making hard to make a perceived, informed decision that we are content with.

Same goes for what @andreasmaaan states above. If we set the threshold to 90 dB? 80? 60? 50? The odds are that there is a point where it is audible for a majority of people. But @90 dB? If i remember correcly from the research I've read and a lot of what really smart people here have done and can PROVE is that 90 seem to be a good value. Unless something is broken and I have exceptional hearing and trained in analyzing sounds, it's totally transparent. For a trained person, the value is 91? 92? But at 100? Thats audiofoolery and as good as the rock you put on amps for clarity you need.


(Sorry for the wall of ramblings. Just wanted put it my words for easier understanding it, english not my first language, so please correct me if didn't understood @andreasmaaan

 

[Vasr]

Yes, it is a lot of work. That's precisely why I'm asking you to put forward a claim. I'm willing to do that work for you.

You still have not understood the last part of my last message to be asking this and making no attempt to do so to be asking this. You are stuck in a binary world. So I am going to end it here with final thoughts.

There is a claim (by you also presumably) that people cannot differentiate two DACs which measure over certain threshold or do not differ in some metrics. I am claiming that this proof has not been established. I have been asking you repeatedly for this but you still keep thumping Psychoacoustics as a bible.

If you care to read through my posts than knee-jerk line by line without context, you will see that I have out together the requirements for such a proof. Of what precisely needs to be established to provide such a proof. You are not doing this but doing the equivalents of Bogleheads do, just hand-waiving. This can be done without any tangential request for a counter claim which has nothing to do with offering the logical line of thought and evidence for this claim.

I have explained that empirical evidence from a population sample isn't sufficient to establish a universal statement (nobody can..). I have told you that it isn't sufficient to say studies exist but provide the logic line of thought that relates a particular study to the claim in hand. Since you seem to be unfamiliar with forms of a rigorous argument, it would be something like these lines

1. People perceive a difference between two systems (all else remaining the same) by these factors and no more no less
- stage balance, tonality, noise/distortion, ... Here is the study that establishes that these are the only things (sufficiency) and nothing else that makes people realize two things are different.

2. Each of the above is captured in this particular metric/measurement. Here is the study that establishes this by showing that when these measurements are equal people do not perceive a difference in that particular dimension.

3. Here are the bounds of EACH of these metrics/measurements beyond which they are differentiable but not within it. These studies show it and hence why we can assert that it is not differentiable within these bounds by anybody. We do this for each such metric.

Thumping Psychoacoustics as a bible isn't the above nor is some partial quoting of data.

There is possibly another opposing claim that may be not all but some may be able to hear the difference. This has not been proven either (and it is not my claim). But this is easier to prove by a simple empirical test where someone would pass the test in controlled conditions. I have not see this either.

So neither opposing claims have been proven and that is OK with me. This is what you seem to be having problems with.

Your insistence for some "novel claim" to disprove whatever that means is missing this whole understanding.

Let me put it this way: If your sound system's maximum output is 100dB SPL peak, the ambient noise level in your living room is 30dB, and your normal listening level is 70dB average, it is extremely unlikely (for everyday purposes "impossible", in the same way jumping 100m or seeing a unicorn is "impossible" even though it cannot be universally disproven by science) that you can hear differences between two DACs that measure flat through the audio band and have SINADs of 100+dB.

Unfortunately, this does not constitute a proof that two DACs are not differentiable. I have already mentioned the list of things that need to be established to offer this as a proof like for example the sufficiency of measurements to capture what the perception is, etc. If you do not understand what the standard for a proof is, that is ok. We are not working in the same framework. Before one offers proof, we do have to agree on what constitutes proof. I do not believe you have understood what a rigorous proof means (in a Science and Logic framework) despite repeated attempts to tell you. So we go back to beating the dead horse.

Basically, it comes down to not understanding the method of science and understanding what constitutes proof.

I do find a lot of similarities between ASR and bogleheads.org.

 

[andreasmaaan]

Basically, it comes down to not understanding the method of science and understanding what constitutes proof.

@Vasr it is apparently you who are misunderstanding science, since you continue to demand that it must offer proof.

If proof of a universal (as opposed to, say, very persuasive evidence) is your hurdle, then no argument based on science will overcome it - not in respect of audio, and not in respect of anything (and you’ll note I never claimed otherwise, as such a claim would be philosophically untenable).

IMO, if you can’t offer any good reason to reject my claim that I can jump 100m high that doesn’t also apply to someone else’s claim that they can hear differences between well-performing DACs, insisting instead on thrashing around in the epistemological weeds, then I think we’ve exhausted this as a topic of discussion

 

[M00ndancer]

MO, if you can’t offer any good reason to reject my claim that I can jump 100m high that doesn’t also apply to someone else’s claim that they can hear differences between well-performing DACs, insisting instead on thrashing around in the epistemological weeds, then I think we’ve exhausted this as a topic of discussion

Burn! That was one of the most politest answers today. And I totally agree.

 

[Vasr]

I think we’ve exhausted this as a topic of discussion

That is the only thing I agree with given we are speaking different languages.

Fundamental differences in what establishing validity of something in a scientific method means (hint: no it isn't just "persuasive evidence" still confusing lawyering with science). This isn't the first group to use Science as a malleable tool to fit one's conviction!

 

[andreasmaaan]

That is the only thing I agree with given we are speaking different languages.

Fundamental differences in what establishing validity of something in a scientific method means (hint: no it isn't just "persuasive evidence").

We certainly are speaking different languages atm, but I'd honestly like to better understand what you're trying to say.

For example, you said in your previous post:

Before one offers proof, we do have to agree on what constitutes proof. I do not believe you have understood what a rigorous proof means (in a Science and Logic framework) despite repeated attempts to tell you.

The language of "proof" is incompatible with science in the frameworks with which I'm familiar (my brief undergraduate background in this area is in post-war philosophy of science, FWIW).

I think the ideas you're trying to get across are not hitting the mark here because there is a language barrier (for example, the very notion of a "scientific proof" is a misconception in the frameworks I'm familiar with). I'd like to think this barrier is not insurmountable, though.

Could you please point me to some texts that lay out the particular "Science and Logic framework" that you're referring to?

 

[raistlin65]

We certainly are speaking different languages atm, but I'd honestly like to better understand what you're trying to say.

For example, you said in your previous post:


The language of "proof" is incompatible with science in the frameworks with which I'm familiar (my brief undergraduate background in this area is in post-war philosophy of science, FWIW).

I think the ideas you're trying to get across are not hitting the mark here because there is a language barrier (for example, the very notion of a "scientific proof" is a misconception in the frameworks I'm familiar with). I'd like to think this barrier is not insurmountable, though.

Could you please point me to some texts that lay out the particular "Science and Logic framework" that you're referring to?

One has to grok the notion that knowledge is socially constructed to understand where you're coming from. Unfortunately, the paradigm shift is still happening within the sciences. Not everyone is ready to accept that yet.

 

[Blumlein 88]

We certainly are speaking different languages atm, but I'd honestly like to better understand what you're trying to say.

For example, you said in your previous post:


The language of "proof" is incompatible with science in the frameworks with which I'm familiar (my brief undergraduate background in this area is in post-war philosophy of science, FWIW).

I think the ideas you're trying to get across are not hitting the mark here because there is a language barrier (for example, the very notion of a "scientific proof" is a misconception in the frameworks I'm familiar with). I'd like to think this barrier is not insurmountable, though.

Could you please point me to some texts that lay out the particular "Science and Logic framework" that you're referring to?

Yeah, I think he conflates logical proof and mathematical proofs with how science works in other areas. There actually isn't all that much science which can be constructed from the ground up like doing proofs in geometry. It might be an optimum ideal, but if we wait until that is possible much useful science wouldn't be made use of.

 

[brimble]

I've been dying to come in on this as a professional philosopher of science ... but also not dying to, because I think the issues are all very complicated. One simple thing I can say though is that pretty much all philosophers of science these days avoid the word "proof" outside logic, mathematics and mathematical physics, whether they're social constructionists or not. That doesn't really prove (!) that much though, because they might be avoiding it for different reasons. It's a quagmire. A fun quagmire, but still a quagmire.

As for all the other philosophy of science issues, professionals don't agree so I don't see why anyone else should either. (I'm being lazy here, and you can have fun finding things about which professionals do agree. But there aren't many.)

 

[gn77b]

Maybe I'm too pragmatic but for me it's not about the philosophy of science. If everyone hears differences even with cheap headphones it's not about that 1 in 1000 DAC that really sounds different for no obvious reason, isn't it? It's not like finding a person who really can levitate because even the craziest person would think they're harder to find than DACs that sound different but measure identical with today's measurement technology.
All speakers sound and measure different, some amps sound different and measure different but DACs... it's voodoo. As I said before, the Musical Fidelity V-DAC II which was reviewed and measured by Stereophile is admirable for its price. So we have a $200 DAC which was rated B by this renowned magazine. And we have people like MDAguy claiming that you need to pay 10x or 100x for that last 1% improvement in sound. So do we have 99% of perfect sound with $200? And what does 1% mean after all, it bugs me when people say stuff like food A is 30% tastier than food B. Why not 28%? Why not 34%? Where do those numbers come from? And why are DACs that are made to spec (and not tuned by ear) frowned upon? Why do they sound different/bad if they measure well? That $20k dCS must do something different. A lot of vague stuff, if you look close it doesn't hold water.

And talking about Stereophile's A, B, C quality rating. It's a mystery to me and hopefully to everyone with a brain. You see cheap stuff with class A rating and very expensive stuff with class B rating. And why are some products removed from the official recommended list? Why don't they just add to that list and maybe from time to time come back and say let's compare this class A DAC from 2010 with this class A DAC from 2020? See, they never gave even the vaguest definition of the rating system. To me it's politics, it's like saying 1+1=2 but we can't risk offending people claiming 1+1= 57.

Summing up, there are a lot of unanswered (unanswerable?) questions, in the audiophile world the logic system is not self-consistent.

 

[Vasr]

The language of "proof" is incompatible with science in the frameworks with which I'm familiar (my brief undergraduate background in this area is in post-war philosophy of science, FWIW).

Could you please point me to some texts that lay out the particular "Science and Logic framework" that you're referring to?

I think you are taking proof as a word too literally and conflating it with different notions of proof (formal and informal). You should really find a comprehensive book on the Methodology of Science itself if interested or there may even be online resources.

Many colleges will have courses and material. For example

http://logic.berkeley.edu/

is a fantastic and very useful program.

I have already mentioned earlier the various stages of how discovery progresses in science - from conjecture to hypothesis to theory to law of nature (in physical sciences).

The method of science requires certain rigor in establishing what people (to be exact peers) would accept to go from just a conjecture to a theory to be classified as a natural law. Typically, this uses one or more of empirical evidence, experimentation, observation, deduction or a logical inference, etc.

The default for science for some X is unknown/unproven, not some Y unless there is a counter-example. And there can be more than one conflicting theory or hypothesis at any time unless one of them can be rigorously invalidated.

If you propose a Y, then it would be contingent on you to make the case for Y including one or more of the above. That includes, why another plausible Z isn't the case. The thing that can trip these logical line of thought to advance a theory is incomplete logic or some aspect of it can be shown to be demonstrably false or has a false premise. This is why detailing the whole argument is important and peer review of that whole argument is so important.

In lawyering as a contrast (which is what most Internet "science" brandishing is about), the goal is to prevail not find the truth (except as it may be coincidental) which can include incomplete or incorrect logic as long as those remain persuadable (or the opposing lawyer or jury didn't catch it). In science, that would be discouraged and a show-stopper.

Here is an example of incomplete argument:
Conjecture: A human being cannot distinguish between two DACs that have SINADs above 100db
Argument: We have studies that show with statistical significance that humans cannot hear noise below that level. Therefore, it is impossible for human beings to differentiate between those two DACs with measurements above 100db

A possible objection from a peer review:
The argument is true but incomplete for the thesis. The problem is people are listening to some content presumably in stereo and you are making a claim for this scenario. What if the channels of a DAC were not in balance and one of them showed a perceptible tilt to right or left. People would be able to differentiate between that one and another that didn't have the problem, right?

Reply to the objection:
But we fix all of those issues to ensure that both DACs have the same levels and that they have the same channel balance.

Peer objection: But you didn't explicitly specify channel balance as part of the argument. What other parameters are also fixed that would otherwise affect the perception. Do you have confidence that you have caught all possible parameters? If so, you would need to provide an argument to establish this. Otherwise, the argument is incomplete.

Etc. This is just a start.

So if you were to go through this rigor of scientific peer review, the list I posted earlier of what you would need to show to be is what would be acceptable for such a peer review. So that a theory can be advanced and accepted as a plausible theory. Nobody has really done this for this specific issue (this is different from there being multiple controlled studies on various things). Because, it is polarized into two groups - one who believe with no evidence that passes scientific rigor that they can hear a difference and another group that is fully convinced that is the case without actually constructing the entire argument and having it checked for validity (not agreement within an echo chamber or similarly convinced people). Until then, neither side has really "proved" their case.

Note that I am not saying that such a view is wrong, just that it has not been established to a rigor (evidence, studies, limits, assumptions, caveats, relation to alternate theories, etc., that directly deal with the hypothesis of interest) from which it is possible to say "science shows...".

There is no such thing as "persuasive evidence" in science.

The rigorous method of science leads to an effective vaccine. Persuasive evidence leads to accepting hydroxychloroquine as a solution. That is the difference.

We should be glad we have the rigor of science as a common basis for such research. Anyone who is in the scientific field and has gone through it to a higher level of studies gets exposed to and fully understands what that involves, a lot of it from simply the difficulty of getting papers peer-reviewed and accepted. So, the Internet argument nonsense typically does not happen there.

It is unfortunate that this rigorous methodology (including what constitutes a controlled experiment, what is valid logic, what constitutes a logical argument, etc) is not taught much earlier. This would avoid much of the misuse of "Science" whether it is in audio, or climate issues or any of the many problems facing the world today where people misuse "science" simply to advance their convictions (which as we know are subject to multiple biases).

 

[raif71]

I think you are taking proof as a word too literally and conflating it with different notions of proof (formal and informal). You should really find a comprehensive book on the Methodology of Science itself if interested or there may even be online resources.

Many colleges will have courses and material. For example

http://logic.berkeley.edu/

is a fantastic and very useful program.

I have already mentioned earlier the various stages of how discovery progresses in science - from conjecture to hypothesis to theory to law of nature (in physical sciences).

The method of science requires certain rigor in establishing what people (to be exact peers) would accept to go from just a conjecture to a theory to be classified as a natural law. Typically, this uses one or more of empirical evidence, experimentation, observation, deduction or a logical inference, etc.

The default for science for some X is unknown/unproven, not some Y unless there is a counter-example. And there can be more than one conflicting theory or hypothesis at any time unless one of them can be rigorously invalidated.

If you propose a Y, then it would be contingent on you to make the case for Y including one or more of the above. That includes, why another plausible Z isn't the case. The thing that can trip these logical line of thought to advance a theory is incomplete logic or some aspect of it can be shown to be demonstrably false or has a false premise. This is why detailing the whole argument is important and peer review of that whole argument is so important.

In lawyering as a contrast (which is what most Internet "science" brandishing is about), the goal is to prevail not find the truth (except as it may be coincidental) which can include incomplete or incorrect logic as long as those remain persuadable (or the opposing lawyer or jury didn't catch it). In science, that would be discouraged and a show-stopper.

Here is an example of incomplete argument:
Conjecture: A human being cannot distinguish between two DACs that have SINADs above 100db
Argument: We have studies that show with statistical significance that humans cannot hear noise below that level. Therefore, it is impossible for human beings to differentiate between those two DACs with measurements above 100db

A possible objection from a peer review:
The argument is true but incomplete for the thesis. The problem is people are listening to some content presumably in stereo and you are making a claim for this scenario. What if the channels of a DAC were not in balance and one of them showed a perceptible tilt to right or left. People would be able to differentiate between that one and another that didn't have the problem, right?

Reply to the objection:
But we fix all of those issues to ensure that both DACs have the same levels and that they have the same channel balance.

Peer objection: But you didn't explicitly specify channel balance as part of the argument. What other parameters are also fixed that would otherwise affect the perception. Do you have confidence that you have caught all possible parameters? If so, you would need to provide an argument to establish this. Otherwise, the argument is incomplete.

Etc. This is just a start.

So if you were to go through this rigor of scientific peer review, the list I posted earlier of what you would need to show to be is what would be acceptable for such a peer review. So that a theory can be advanced and accepted as a plausible theory. Nobody has really done this for this specific issue (this is different from there being multiple controlled studies on various things). Because, it is polarized into two groups - one who believe with no evidence that passes scientific rigor that they can hear a difference and another group that is fully convinced that is the case without actually constructing the entire argument and having it checked for validity (not agreement within an echo chamber or similarly convinced people). Until then, neither side has really "proved" their case.

Note that I am not saying that such a view is wrong, just that it has not been established to a rigor (evidence, studies, limits, assumptions, caveats, relation to alternate theories, etc., that directly deal with the hypothesis of interest) from which it is possible to say "science shows...".

There is no such thing as "persuasive evidence" in science.

The rigorous method of science leads to an effective vaccine. Persuasive evidence leads to accepting hydroxychloroquine as a solution. That is the difference.

We should be glad we have the rigor of science as a common basis for such research. Anyone who is in the scientific field and has gone through it to a higher level of studies gets exposed to and fully understands what that involves, a lot of it from simply the difficulty of getting papers peer-reviewed and accepted. So, the Internet argument nonsense typically does not happen there.

It is unfortunate that this rigorous methodology (including what constitutes a controlled experiment, what is valid logic, what constitutes a logical argument, etc) is not taught much earlier. This would avoid much of the misuse of "Science" whether it is in audio, or climate issues or any of the many problems facing the world today where people misuse "science" simply to advance their convictions (which as we know are subject to multiple biases).

Thank you @Vasr for your long posts explaining that all is not quite kosher with the statement "DACs that measure well will not be audibly indistinguishable from each other". I have long believe this to be the case but lack the language mastery and knowledge as you so eloquently presented. I will not say much but from my understanding the two sides are neither right nor wrong but like the title of this thread, the dead horse will still be beaten.

 

[JustJones]

Beating this poor dead horse on super humans hearing differences in DACs that measure beyond the ability of human hearing is starting to remind me of beating another dead horse on the theory of evolution.

 

[andreasmaaan]

You should really find a comprehensive book on the Methodology of Science itself if interested or there may even be online resources.

Many colleges will have courses and material. For example

http://logic.berkeley.edu/

is a fantastic and very useful program.

I have already mentioned earlier the various stages of how discovery progresses in science - from conjecture to hypothesis to theory to law of nature (in physical sciences).

@Vasr thanks for the links.

I presume you realise that this is not the only model of science that is widely accepted at an academic level? For example, the idea of a "law of nature" is heavily critiqued, and barely appears in any philosophy of science of the past sixty years.

I think you are taking proof as a word too literally and conflating it with different notions of proof (formal and informal).

Pardon me?

If you don't intend a word to mean what it says, then choose another word.

Meanwhile, your past two posts have seized on my choice of "persuasive evidence". Instead of asking for clarification of what this term means, you've assumed that it implies whatever you think the legal definition of "persuasive evidence" is.

So on one hand, your informal choice of language is something I am "taking too literally", while on the other hand, mine is assumed to have a specific meaning derived from a completely separate field (which I have never even mentioned, btw) .

Can you see how this is not an intellectually generous approach to this discussion?

The default for science for some X is unknown/unproven, not some Y unless there is a counter-example. And there can be more than one conflicting theory or hypothesis at any time unless one of them can be rigorously invalidated.

Yes, I absolutely agree with this, although I would tend to prefer "falsified" in this context.

If you propose a Y, then it would be contingent on you to make the case for Y including one or more of the above. That includes, why another plausible Z isn't the case. The thing that can trip these logical line of thought to advance a theory is incomplete logic or some aspect of it can be shown to be demonstrably false or has a false premise. This is why detailing the whole argument is important and peer review of that whole argument is so important.

I broadly agree, too (although I would suggest that peer review in practice tends to fall short of this requirement).

Here is an example of incomplete argument:
Conjecture: A human being cannot distinguish between two DACs that have SINADs above 100db
Argument: We have studies that show with statistical significance that humans cannot hear noise below that level. Therefore, it is impossible for human beings to differentiate between those two DACs with measurements above 100db

A possible objection from a peer review:
The argument is true but incomplete for the thesis. The problem is people are listening to some content presumably in stereo and you are making a claim for this scenario. What if the channels of a DAC were not in balance and one of them showed a perceptible tilt to right or left. People would be able to differentiate between that one and another that didn't have the problem, right?

Reply to the objection:
But we fix all of those issues to ensure that both DACs have the same levels and that they have the same channel balance.

Peer objection: But you didn't explicitly specify channel balance as part of the argument. What other parameters are also fixed that would otherwise affect the perception. Do you have confidence that you have caught all possible parameters? If so, you would need to provide an argument to establish this. Otherwise, the argument is incomplete.

Etc. This is just a start.

Interestingly, I don't think we disagree at all on this.

So if you were to go through this rigor of scientific peer review, the list I posted earlier of what you would need to show to be is what would be acceptable for such a peer review. So that a theory can be advanced and accepted as a plausible theory.

"Theory" shouldn't pop up here IMO. That a Person A can't distinguish between DAC X and DAC Y is a mere hypothesis.

Nobody has really done this for this specific issue (this is different from there being multiple controlled studies on various things). Because, it is polarized into two groups - one who believe with no evidence that passes scientific rigor that they can hear a difference and another group that is fully convinced that is the case without actually constructing the entire argument and having it checked for validity (not agreement within an echo chamber or similarly convinced people). Until then, neither side has really "proved" their case.

As I argued earlier, the side that says Person A can't distinguish between DAC X and DAC Y simply cannot "prove" its case (not even if we subject Person A to experimentation).

Would you like me to expand on the reasons why I say this, or is it clear enough already?

Note that I am not saying that such a view is wrong, just that it has not been established to a rigor (evidence, studies, limits, assumptions, caveats, relation to alternate theories, etc., that directly deal with the hypothesis of interest) from which it is possible to say "science shows...".

I essentially agree. In fact, I try to avoid ever making categorical statements that begin with "science shows". However, to argue that scientific evidence suggests that X or Y is extremely unlikely is not an abuse of science, nor should I need to conduct a study and produce a peer-reviewed paper every time I make such a claim.

There is no such thing as "persuasive evidence" in science.

The rigorous method of science leads to an effective vaccine. Persuasive evidence leads to accepting hydroxychloroquine as a solution. That is the difference.

How about this? If you agree to drop the patronising, intellectually empty rhetorical flourishes, I'll refrain from them, too - deal?

We should be glad we have the rigor of science as a common basis for such research. Anyone who is in the scientific field and has gone through it to a higher level of studies gets exposed to and fully understands what that involves, a lot of it from simply the difficulty of getting papers peer-reviewed and accepted. So, the Internet argument nonsense typically does not happen there.

It is unfortunate that this rigorous methodology (including what constitutes a controlled experiment, what is valid logic, what constitutes a logical argument, etc) is not taught much earlier. This would avoid much of the misuse of "Science" whether it is in audio, or climate issues or any of the many problems facing the world today where people misuse "science" simply to advance their convictions (which as we know are subject to multiple biases).

It seems to me that your entire argument amounts to this: No one may invoke scientific evidence in any discussion unless they have conducted a study on the specific question and produced a peer-reviewed paper first.

Is this a fair characterisation?

 

[Feanor]

Beating this poor dead horse on super humans hearing differences in DACs that measure beyond the ability of human hearing is starting to remind me of beating another dead horse on the theory of evolution.

The misunderstanding is that hearing differences between DACs and other audio components is a matter of "super humans hearing". It isn't about super hearing, rather it has to do with very attentive listening and some experience of what to listen for.