[khaos974]

A fairly easy way to check a cable, or an amp/preamp is simply null testing.
You A/D both input and output signals, time and volume align both signals, and subtract them. And see the difference in dB.
 
Check Audio DiffMaker, someone even made a software for that.
More explanation:
 

 
If you want to, you can even do the null testing with an oscilloscope, it's more difficult to implement though.

 

[eucariote]

Quote:

 
 
 
Yes, I see audio science as applied science---making use of ideas from science and engineering, and balancing that will subjective evaluations of equipment. The important thing is not to think the measurements are the ultimate arbiter.
 
 
You keep bringing everything back to cables. I'm not talking about cables exclusively.
 
Let's talk about speakers. No one disagrees they have audible differences. So we can get a bunch of measurements, but we still have the problem of correlating those to perception.
 
There are more opinions than that. I believe that perception is unreliable, but I don't make the mistake of retreating to a "safe" empirical position because of that fact.

^ I don't see what this has to do with DBT.  DBT is simply a methodological control whereby possibly confounding sources of information that could influence a test outcome (ideas, biases and expectations that accompany knowledge of a treatment/drug/sound source) are eliminated by keeping a person unaware of what treatment they are receiving.  That's it.  
 
For example, double blind tests established that selective serotonin reuptake inhibitors improved patients' reports of mood to a statistically significant degree over placebo pills based only on the subjective results of the patients (the realm which you seem to favor).  And that is the same technique that could answer the question of whether differences heard between cables would still be heard if potentially confounding influence of knowledge of which cable is being listening to is eliminated.
 
Findings from DBT tests are accepted and valid, independently of knowledge of underlying processes.  For example, the paper I cite using DBT achieved positive results, without knowledge of mechanisms of how the actions of the drug actually alleviate depression.  Mind you, there are massive research projects that are detailing the drug actions on neuron receptors, relation of serotonin neurons to systems mediating affect and reward, etc etc.  And so there are theories that are outlining the possible reasons for these discovered effects on mood- but the mechanisms have not been completely worked out, and could all be wrong and could be revolutionized in 10 years.  And when they are discovered, that knowledge will have no bearing on the outcome of the DBT test result.  
 
Another example, lithium, reduced bipolar episodes to a statistically significant degree over placebo using DBT tests.  And the initial effects of lithium on mood were a purely accidental finding, with a total lack of awareness of mechanisms.  Even with this lack of mechanistic knowledge, the DBT test results (and seeming lack of averse health effects) were reason enough for successful medical use and treatment of bipolar disorder *for decades* before theories of the mechanism recently began to emerge.

 

[mike1127]

Quote:

^ I don't see what this has to do with DBT.  DBT is simply a methodological control whereby possibly confounding sources of information that could influence a test outcome (ideas, biases and expectations that accompany knowledge of a treatment/drug/sound source) are eliminated by keeping a person unaware of what treatment they are receiving.  That's it.  
.

You're right, my particular post was not about DBT. Perhaps it is off-topic in this thread, but this thread seems to be ranging everywhere.
 
I don't have any problem with the principle of audio DBT, but I do think the execution of it is generally flawed.


 

 

[maverickronin]

Quote:

I don't have any problem with the principle of audio DBT, but I do think the execution of it is generally flawed.
 

I can agree that some types of audio DBT are flawed and not always the best for picking up very subtle differences.  What sort of tests would you consider flawed and why?  Also the converse.  What sort of tests would you consider correct and why?

 

[Krav]

I think the reason 1)  you came up with is the biggest one. The side effect of some people sometimes learning something of value on this site notwithstanding, its primary purpose at the current phase of its evolution is to make money for its owners.
 
Money are made by collecting advertising fees, referral commissions, identification payments, and potentially in many other ways. The first two are obvious, but the identification payments are somewhat disturbing - as I type, an ad is shown that could only be targeted to me by sharing information with other sites - in my particular case, with Expedia - most likely through third-party aggregators.
 
This level of sophistication shows that this site is not a cozy small-time hobbyists hangout anymore but a serious commercial enterprise, and its Alexa Traffic Rank corroborates that: http://www.alexa.com/siteinfo/head-fi.org. As any normal enterprise, e.g. a printed hobbyist magazine, Head-Fi needs to make enough money to stay in business and to grow the business. This is perfectly normal and legal, just needs to be understood as a fact of life.
 
Correspondingly, the tactical objectives of site owners are to have as many users as possible, to make them go back to the site as frequently as possible, to stay there for as long as possible, and to make them buy as often as possible at as high price as they can bear. I think the reasons 2) and 3) you are referring to are marginal. You can find plenty of very flame-y threads with severe personal insults that are not being deleted. They are not considered a threat to the site business model I guess.
 
Quote:

Hey,
 
I got the impression that blind testing discussions on Head-Fi are explicitly _censored_, _marginalized_ and moved to the _outskirts_ of the forum, to one geeky sub-forum, mysteriously called 'Sound Science' (who cares about science anyway?).
 
Here is my own experience: in one ultra-popular thread, which discussed some flavor of the month gear, one reviewer made quite comprehensive review of some stuff, including USB DATA CABLES, and concluded that this but not that USB DATA CABLE makes a BIG DIFFERENCE SQ-WISE (I am not overemphasizing - this was the wording of the conclusion of the review). Now, I _instinctively_ posted a reply, something like 'Dude, did you perform a blind test, cause your sound perception might have been blinded by your shiny bourgeois cable'). Whoo, that was my mistake. I received a very strong reply from the reviewer, after which I felt like I have engaged in some illegal activity or something. It began like 'Did you compare? No? Then ****', which was kinda OK, but that was not it - the guy continued, relied on AUTHORITAHH (apparently there was a 'sticky' post somewhere prohibiting DBT suggestions!) and asked me not to do that again! Now I see the same guy posting comprehensive detailed reviews of different stuff, which are followed by numerous 'ooh ahh how nice', and then - procession to group buys, of course. To my humble mind, hearing BIG differences in USB DATA CABLES completely invalidates all other observations of that person with regard to SQ, and extensive reviews written by such a person in confident fashion may be very misleading. Therefore, when such a person writes hearing big differences in USB cables, people have the right (I would say, an obligation) to debunk them INSTANTLY, here and now. After all, our aim here is to be as objective as possible, as otherwise no single review would make any sense (except as a piece of arts, if it is written poetically).
 
Anyway, I was shown my place, but I thought for myself - what is the rationale behind such a prohibition? As you can't even mention DBT to some lunatic who claims that expensive usb data cables make big difference in sound. Here are some possible answers that I came up with:
 
1) Materialistic interests. Millions of bucks made from cable believers are the ultimate force, and openly debunking this stuff would interfere with this kind of business.
 
2) Feelings. Perhaps this is one of American 'exceptionalisms' - you can claim respect for your fragile _beliefs_, even when those are like the earth is 6000 years old, super sounding usb cables, or people riding dinosaurs and so on. Debunking stuff can make some people feel bad, how dare you to make fun out of them?
 
3) Flame prevention. Idea that blind testing suggestions may generate pages of unnecessary flame topics, departing from the original topic. But heck, are 'ooh ahh how nice' replies any better than a gentle debunking statement?
 
Other than that, I could not think of anything else. Did I miss something? I am really curious as to what other rationales might be behind this rude censorship.
 
It is a scientific FACT that human hearing (and not only that) is extremely prone to illusions facilitated by the intelligent brain (esp. when something is expensive and shiny); and the only way to get out of this situation is to use blind controlled tests. This is how scientific method works, this is how brilliant headphones and all music reproduction equipment were invented in the first place - through highly critical, impersonal, evidence based research.
 
Any thoughts?
 

 

[Prog Rock Man]

 
Quote:

A fairly easy way to check a cable, or an amp/preamp is simply null testing.
You A/D both input and output signals, time and volume align both signals, and subtract them. And see the difference in dB.
 
Check Audio DiffMaker, someone even made a software for that.
More explanation:
 

 
If you want to, you can even do the null testing with an oscilloscope, it's more difficult to implement though.

The above is the first real alternative to blind testing I have seen.

 

[maverickronin]

Quote:

Money are made by collecting advertising fees, referral commissions, identification payments, and potentially in many other ways. The first two are obvious, but the identification payments are somewhat disturbing - as I type, an ad is shown that could only be targeted to me by sharing information with other sites - in my particular case, with Expedia - most likely through third-party aggregators.
 

Ha!  I was not aware of those at all!   The only ads I see are the plain text sponsors list.
 
Another reason the talk of DBTs and testing cables is marginalized is because such tests show that those fancy cables are completely unnecessary.  A bad review of any decent product in another category such as an amp or a pair of headphones can easily be written off as a difference in preferences or tastes.  Maybe you prefer the HD800 to the T1.  Maybe vice-versa.  Pretty much no one who has heard one and who has any credibility around here doubts that you do get something for your money though.  There is quibbling over individual tastes or the question of relative value, but not a blanket dismissal of the entire market as useless and not good for anything but pose value.  That's the only reasonable conclusion you can come to after investigating fancy cables.  Therefore reason and logic are frowned upon in this arena.  They don't need to suppress it in other areas since those products do have objective value.  They don't need protection.  They can stand on their own.  No guards are needed to keep pesky skeptics like me from blowing over the metaphorical house of cards that serves as the foundation of the audiophile cable industry.
 
It seems to be creeping into other areas though.  The way the hifiman 801 RMAA thread got shoved into the sound science forum seems like a crude attempt at censorship to me.  Is objective measurement to be banned everywhere else but the sound science forum?  That's not a good long term strategy for running this site as a business.  If such a thing is enforced it will likely cause a divide within the community and split the user base.  The people who are driven out by such a split are unlikely to be a large enough percentage to cut into the ad revenue, but that's not the real danger.  The danger is that those likely to leave will be among the most influential.  If a large number of the respected trendsetters, opinion makers, and reviewers leave, then the rest of the user base will slowly follow.
 
I'm attempting to give some useful advice to the admins.  I haven't been a member here very long, but I've learned quite a bit from it and would like to see it continue indefinitely.  I can't say I know the true motives behind all their decisions.  I can say that scrupulous honesty is the only viable long term strategy.  Selling out to advertisers is at best a strategy destined to cause vicious boom and bust cycles.  At worst, it will bring a swift end to the enterprise.  If these actions are instead a misguided attempt to 'clean up' the forums, then I have another solution.  Censorship always fails.  The only solution to the misuse of freedom of speech is more freedom of speech.  If someone says something that is false or offensive then the solution is not to delete the post, but to argue against it.  People should be allowed to argue anything that is on-topic to the thread and sub-forum.  The admins should only be there to enforce civility.  Attacking an idea, any idea, should be fair game.  Attacking a person should not be.  Of course this server does belong to the admins and they have the right to regulate it as they see fit.  I make these recommendations not because they are, or should be, legally obliged to abide by them, but because I think they are in the best interest of the long term continuation of the site.

 

[sohels]

 
mike1127, please observe the following image:



 

Suppose you're convinced that the lines are not parallel (where as in fact they are), but you are not allowed to use a ruler to verify this. You are now asked to re-create this image and you get most of the details right except your lines are not parallel. You are then shown two images - the one that you re-created, and the original itself. You are then asked as to which of the two more accurately resembles the image that you were first shown - but this time you are allowed to use a ruler. Since you believe that the original did not have parallel lines, you claim that the image you came up with is more accurate, ie, closer to the image you were first shown. Do you now see the problem that relying on perception alone can create?

 

[Prog Rock Man]

The forum member Downsize who is a regular in the DIY section and at present has a thread on his latest DIY cables, has stated he has easily passed various blind test. He also stated that was true of his staff (I think he runs, or ran an audio shop). I asked him, he replied (stating a 90% pass rate as with his staff), someone else asked about blind testing, he answered and reminded everyone about the rules and that was the end of it.
 
http://www.head-fi.org/forum/thread/508252/the-best-interconnects-i-have-ever-heard-photos/30
 
 
1 - if that was the extent of blind testing comments then I am sure all would be fine, but obviously that does not happen here. At least small exchanges like that get through.
 
2 - I am sure most cable makers have tried blind testing. If it was in their favour, it would be used in all of their adverts and claims. But it is not.
 
3 - I wonder if the above exchange got through the moderators because it was in favour of cables?

 

[Catharsis]

 
Quote:

Aha! So you had to refer to a model of perception in order to answer. Which demonstrates my point: accuracy outside the context of perception is useless.

 
Ahem....mike, I'm going to agree with you (partially).  You don't ALWAYS need a model of perception to make measurements meaningful, BUT understanding the limits of human perception is why we only measure FQ response between 20-20,000hz, why mp3 files have the compression algorithms they have (based on psychoacoustics), why redbook audio CD maxes out at 44.1khz etc etc. 
 
Where I don't agree with you is that accuracy outside of perception is useless.  A measurable waveform input goes in, and comes out with either no degree of change, or some degree of change.  These changes can be measured (all types of measurements) and can exist independently from any model of perception.  The lesser the change in the output signal relative to the input signal, the greater the accuracy.
 
In your example, I chose device A because although I lack the knowledge to quantify which one is more/less accurate, a knowledge of perception would indicate that humans are more sensitive to changes of 1dB vs. 0.05% THD.  Still, I would think that on a relative scale, a change in frequency response 2-3dB is a greater deviation from the original signal than a tiny 0.05% THD.  You would have to put weight on each of the measurements based on which measurement is a greater indicator of signal change to determine this, and that could be accomplished quite easily.

 

[aimlink]

Quote:

mike1127, please observe the following image:

 
Suppose you're convinced that the lines are not parallel (where as in fact they are), but you are not allowed to use a ruler to verify this. You are now asked to re-create this image and you get most of the details right except your lines are not parallel. You are then shown two images - the one that you re-created, and the original itself. You are then asked as to which of the two more accurately resembles the image that you were first shown - but this time you are allowed to use a ruler. Since you believe that the original did not have parallel lines, you claim that the image you came up with is more accurate, ie, closer to the image you were first shown. Do you now see the problem that relying on perception alone can create?

Cute pic. 

 
Is there any real point in arguing from either extreme?  Should you rely on measurements alone? I'd say no since it implies that the measurements used are measuring everything that you'll hear.  Should you rely on perception alone?  I'd say no again since you not only gave a good example of perceptual issues common to all of us, but there are perceptual issues unique to each individual.
 
However, at the end of the day, I'd evaluate gear from two standpoints.  One standpoint would be for personal use, and from that point, it's my own perception that I'll more heavily rely on.  A blind test is a test using my own perception so even if I choose to go down that road, it still falls within that realm.  However, I'll not look at a measurement or chart and use that to appease my senses that aren't pleased with what I'm hearing.
 
OTOH, if I'm selling equipment or recommending equipment, the measurements, which are more objective, will have a more feature heavy role to play for the objectivists that walk in through the door.  Be that as it may, what will give special character to a product line is the human factor that goes in evaluating the external design, as well as the sound/performance of that line of gear.
 
It'll be a sad day when the sound of equipment is defended solely based on graphs and measurements.  If I had to live with one extreme, it would be the one without the machines taking the measurements.

 

[maverickronin]

Quote:

.
 
It'll be a sad day when the sound of equipment is defended solely based on graphs and measurements.  If I had to live with one extreme, it would be the one without the machines taking the measurements.

Assuming that suitable tests are devised, I'd go the other way and generally prefer measurements over subjective impressions.  That wouldn't be perfect, but I think it would be better than a sea of purely subjective impressions with little grounding in reality.  I believe we need both objective measurements and subjective impressions.  The important thing to understand is what those measurements are for.  They don't tell you everything about a piece of gear and they aren't supposed to.  They serve to keep the discussion grounded in reality.  They can keep you from fooling yourself and help you spend you money more wisely.  Most importantly they give you a way compare things without having having to hear it yourself.  It would certainly be preferable to listen to something before you buy it, but that is not always possible.  Good return policies help a little, but not a much as a showroom or a meet in you town, and those aren't exactly common.

 

[Prog Rock Man]

I am sure that if you have any two hifi products which have as flat a frequency response as possible, no matter the cost and construction, they are hard to differentiate between and are generally preferred as sounding good. Musical Fidelity and AKG are my preferred brands and both tend to have flat frequency responses, so that is the only measurement I look for.
 
I will go with subjective reviews, both here, other forums and in the audio magazines where the reviews tend to agree. Otherwise they are not very useful at all.

 

[aimlink]

Quote:

Assuming that suitable tests are devised, I'd go the other way and generally prefer measurements over subjective impressions.  That wouldn't be perfect, but I think it would be better than a sea of purely subjective impressions with little grounding in reality.  I believe we need both objective measurements and subjective impressions.  The important thing to understand is what those measurements are for.  They don't tell you everything about a piece of gear and they aren't supposed to.  They serve to keep the discussion grounded in reality.  They can keep you from fooling yourself and help you spend you money more wisely.  Most importantly they give you a way compare things without having having to hear it yourself.  It would certainly be preferable to listen to something before you buy it, but that is not always possible.  Good return policies help a little, but not a much as a showroom or a meet in you town, and those aren't exactly common.

You can never be grounded in the reality of this 'playground' of audiophilia that we discuss since there's no genuine reality to it.   I'm not being metaphysical either since careful reasoning will bring anyone to this conclusion.  It's all based on opinion and perception.  The moment we bring a piece of equipment into the picture, we suddenly feel that it's objective?  I don't get it.  For me, this isn't valid reasoning.
 
Aren't we the same ones who build these instruments and subjectively decide on their importance?  We subjectively decide that studies make for valid ways of proving things.  We subjectively decide to make claims on reality since we supposedly have 'enough' pieces of reproducible evidence/experience that supports our claim/s.   Who makes that decision of 'enough'?  IMO, you give yourself way too little credit.
 
What I'm saying is that we can't divorce ourselves from evaluations the way that you wish to or think that you are able to.  It's one thing to be grounded in reality.  It's another thing to not be grounded in reality, realise it, and try to find it.  It's yet another to think you're grounded in reality or are making headway in it under false pretense, and to be ignorant about this.  We can so easily fall into this latter category.
 
We are the masterminds behind our measurements, their interpretation and the validity/importance we put to them.  We will always remain at the center of it.
 
One such measurement and one which we base our satisfaction on is simply listening.  Afterall, what else would be the point?

 

[sohels]

The point I'm trying to make is that if the goal is to most accurately reproduce a piece of music, one must not be dismissive of objective measurements. Pursuing 'perceptual accuracy' is absolutely fine - but to claim that this pursuit is in fact identical to, or indeed is even more valid/effective than relying on scientific measurements in achieving the aforementioned goal, is simply not right.