[pinnahertz]

I understand your concerns regarding the methodology used. Even though I doubt an ABX switch that would mess with the inputs and outputs of the C370 amplifier would be better than directly and manually switching from one channel to the other without any noise or click whatsoever: that's the closest thing to transparency that occurs to me.

Something important to understand here. An ABX switch should not "mess" with anything. However, since even the effects of an ordinary switch, like the input switch on a C370, can be measured, it's effects should be as minimal as possible, but more importantly, identical to sources being compared, thus being a controlled variable.

I also doubt that increasing the number of subjects from 5 to 50 would yield a different outcome. But hey, that's science, right?

Come on guys, it's not that difficult to understand that when 5 different people —2 of them without a "proper" listening ear, but with personal preferences and 3 of them with "trained" ears, 2 of these without any bias at all for I didn't ask them anything other than to express their opinion to me on what, if any, difference they perceived— manifest that they find no difference whatsoever between the two DACs maybe it is because there is none?

In other words: I would expect a $220 device to sound so much better than a $5 one, such as to make a noticeable difference within the very first seconds of listening to it, right? Well, here it's not the case.

If this were an actual controlled DBT, the above would become part of the hypothesis which would have to be proven, disproved, or modified by the results of experiment, which then circles back around to further experiment. Scientific method at work.

This is not aimed at anybody in particular but I perceive a lot of https://en.wikipedia.org/wiki/Obsessive-compulsive_disorder in the adiophile scene, mixed with phat wallets. A combination waiting for financial disaster if you ask me...

Anyway, if you advice me which DAC to get to test it against this Soundblaster, I'll happyly try to get it in the coming months. I was thinking of getting either the Chord Mojo or the Audiolab M-DAC. Both have "impressive" reviews within the scene.

I have learned here that money should be put on a good pair of speakers or headphones instead, probably the only? variable in the equation that makes real sense to throw money at.

The way I see it, once you pass a certain quality threshold, enjoy the music instead of OCD'ing over the last .9999% of quality to reach that impossible 100% of audio nirvana, it's that simple!

I'd like to dedicate this theme to all of you. It sounds absolutely wonderfull on my limited budget $5 soundcard + second hand $250 C370 amp + second hand $300 B&W 602 speakers + $350 Etymotic ER4XR IEMs and I'm sure? it will sound much better on way more expensive sound systems...

...or maybe not.

Couldn't agree more. The gap between choosing your system components for enjoyment and scientific research is rather huge, and really doesn't intersect at all. The problem we run into here is that people arrive at a "definitive" conclusion using completely unscientific methods, then argue their point to death in the face of reality. If someone says, "I picked the new Glorator 6000 Super Analog Humungo-Amp because I thing it sounds fantastic!" I wouldn't say a thing, perfectly fine. But when they say, "I picked the Glorator 6000 super Analog Humungo-Amp because I and 4 friends compared it to the CheezeyTech Honkhorn 20 and it was like lifting the veil!" I'd probably question the test methods.

In AB testing it's possible to show not only a statistically confirmed perceived difference but a preference for one of two identical choices simply by applying bias towards that choice. And by identical, I mean they're the same source connected to the switching device with a "Y" cord. Then tiny things like a fraction of a dB of level differences biases results. It's one of the oldest HiFi demo tricks in the world, play the one you want to sell just slightly louder, they pick it every time.

As to the cost vs quality thing, I will personally agree that there is no direct link between the two until you get to the extreme low end of cost. An example for me is my "reference" IEMs retail for $50, and blow away anything up to 10X their cost. I've auditioned headphones in the $6K price range, and found them "good". Not Nirvana. And out of scale with their price. I think the concept here is right on, but we should be careful about the difference between opinion and fact.

Absolutely NOT true!

 

[bigshot]

When someone shows me their sound system or home theater, I'm always more impressed if they've gotten great sound out of midrange equipment. With enough money, a chimpanzee can put together a system with lots of beautiful lights and tubes and shine black cases. But someone who knows how to actually *use* their tools to get the most out of them is more interesting to me than someone who takes the "money is no object" approach. That's why I've never listed my gear in my sig. The equipment doesn't sound good. It's the way I've put it all together. And a laundry list of brands and model numbers isn't going to tell you anything about how it sounds.

 

[pinnahertz]

When someone shows me their sound system or home theater, I'm always more impressed if they've gotten great sound out of midrange equipment. With enough money, a chimpanzee can put together a system with lots of beautiful lights and tubes and shine black cases. But someone who knows how to actually *use* their tools to get the most out of them is more interesting to me than someone who takes the "money is no object" approach. That's why I've never listed my gear in my sig. The equipment doesn't sound good. It's the way I've put it all together. And a laundry list of brands and model numbers isn't going to tell you anything about how it sounds.

I know someone else who doesn't list their gear! And I love to make it sound great without spending the equivalent of a nice summer home.

After designing and calibrating many more rooms than I want to count I've found even the "good" stuff benefits from EQ, calibration, and acoustic treatment, rarely less so than the modest stuff. Focus on the big issues.

Back in the day there were a few companies that centered on great sound for affordable prices, the likes of Henry Kloss at KLH and Advent, Apt, and NAD. And they accomplished that goal. Sure, the stuff didn't look terrific, but it was well made and sounded like it cost 10X the price.

I've seen some of those chimp designed systems. Impressive that they did it. Of course, one also typed out the Gettysburg Address in a room with an infinite number of chimps and typewriters.

 

[mBTX]

I don't think you have very good ears. I have both a soundblaster audigy fx, an an STX II with burson v5's, and an Asus xonar essence one......

With the SB both the jitter and noise floor are borderline unlistenable.

Looking through this thread, I saw nowhere where you express being able to hear minute differences. Just as a point of curiosity, I wonder if you'd pass a color test on a high RGB gamut monitor? (are the 2 even linked???)

Anywho, within any hobby you always run into diminishing returns. Of corse sound devices are %90 the same. It's that %5,%6,%7,%8,%8.1,%8.2, and so on we're after.

 

[deafmutelame]

I don't think you have very good ears. I have both a soundblaster audigy fx, an an STX II with burson v5's, and an Asus xonar essence one......

With the SB both the jitter and noise floor are borderline unlistenable.

Looking through this thread, I saw nowhere where you express being able to hear minute differences. Just as a point of curiosity, I wonder if you'd pass a color test on a high RGB gamut monitor? (are the 2 even linked???)

Anywho, within any hobby you always run into diminishing returns. Of corse sound devices are %90 the same. It's that %5,%6,%7,%8,%8.1,%8.2, and so on we're after.

Hi mBTX, thanks for your reply.

Fortunately my ears are fine, thanks. But even if my ears weren't fine, the ears of 5 more people, the youngest one a 19 yo and the oldest one a 42 yo, two of them "trained" to listen to audiophile hardware haven't been able to hear ANY difference whatsoever when switching between the two. What does that tell you?

When I "crank" the SB to its loudest level in the amp (we are talking a 120W x 2ch. kind of amp, the NAD C370), other than the typical "hiss" (or noise floor, as you call it) that you can hear with ANY device we don't hear ANY jitter or any kind of noise at all, even when the CPU workload is working close to its highest load.

Maybe I am lucky to have a completely rubbish power supply, for that is what is preventing me from listening to any kind of "internal noise"...

Let me ask you: have you really been able to do an ABX test with both soundcards playing at the same time? In other words: a test where you could manually and transparently switch between one and the other without any sound gap between them. Or have you listened to one first and some minutes later to the STX II "with Burson v5's"? (I was even considering to purchase the MUSES01 for the buffer but... no way now!).

Also, I'm writing this looking at a high RGB gamut monitor for I work with photography, my eyeseight is very well, thanks again and no, I don't think the two are linked anyway...

It's ok that you are after that last 5%. In my opinion we are not talking about that last 5% here but about a 100% similarity between both: which tells me that the STX II is a scam, pure and simple.

It should be obvious that you would expect an INSTANT jaw dropping at the very first seconds of listening to it and after investing $220 on it, right? Well, that difference simply is not there, confirmed by 6 people.

Or maybe the Soundblaster Audigy FX is real rubbish compared to the X-Fi Xtreme, which I very much doubt...

Finally, I'd suggest you not to overstress about it. But anyway, who am I to prevent you from going for that last 5%?

After all it's just a harmless hobby... (other than to your own wallet, of course).

 

[deafmutelame]

Couldn't agree more. The gap between choosing your system components for enjoyment and scientific research is rather huge, and really doesn't intersect at all. The problem we run into here is that people arrive at a "definitive" conclusion using completely unscientific methods, then argue their point to death in the face of reality. If someone says, "I picked the new Glorator 6000 Super Analog Humungo-Amp because I thing it sounds fantastic!" I wouldn't say a thing, perfectly fine. But when they say, "I picked the Glorator 6000 super Analog Humungo-Amp because I and 4 friends compared it to the CheezeyTech Honkhorn 20 and it was like lifting the veil!" I'd probably question the test methods.

We agree that there is absolutely NO PERFECT way to prove *ANY* audiophile claims even after following the best "scientific method" at all, right?

Absolutely NOT true!

 

[pinnahertz]

We agree that there is absolutely NO PERFECT way to prove *ANY* audiophile claims even after following the best "scientific method" at all, right?

The statement is "spun" toward a particular direction that I have issues with. There are several audiophile claims that can be proven to be within an acceptable statistical range of audibility (or inaudibility).

With statistical analysis you don't actually get absolutes very often, which is why analysis often includes arbitrary thresholds. However, if you use a control test (A=B=X) with a similar number of data points the differential becomes pretty sensitive.

 

[deafmutelame]

statistical

That's precisely the word that I was expecting to read from someone intellectually honest and knowledgeable about it: there are no proven facts but statistical "truths". Thanks.

In other words: even though I have tried to carefully set this up, with reasonable gear and non-biased subjects (2 of them at least), the procedure I have followed here "might" not be within the .99999% percentile of the statistical "truth" for it has not followed a much more strict "scientific" procedure (which would not prove anything neither), yet it is still relevant for the average ear (=non dog ears ), right?

 

[bigshot]

I don't think you have very good ears. I have both a soundblaster audigy fx, an an STX II with burson v5's, and an Asus xonar essence one......

With the SB both the jitter and noise floor are borderline unlistenable.

What does jitter sound like? I'm betting you've never done a line level matched blind test.

 

[castleofargh]

That's precisely the word that I was expecting to read from someone intellectually honest and knowledgeable about it: there are no proven facts but statistical "truths". Thanks.

In other words: even though I have tried to carefully set this up, with reasonable gear and non-biased subjects (2 of them at least), the procedure I have followed here "might" not be within the .99999% percentile of the statistical "truth" for it has not followed a much more strict "scientific" procedure (which would not prove anything neither), yet it is still relevant for the average ear (=non dog ears ), right?

your test is very valid for your testing conditions. you demonstrate that under those circumstances you don't seem to notice a difference. but such a test might not necessarily be valid to extrapolate about the gears being identical and stuff like that. it's all just a game of conditional truth and not stepping outside the conditions.

about your previous post, there are many claims which can absolutely be proved or disproved with more or less ease. but the most important IMO is that if there is no way to confirm or disprove a claim, then you shouldn't consider it in the first place. I personally reject empty claims for being empty claims. I don't care if the claim ends up being true. just like I don't care for someone who reaches the right result after 2 consecutive mistakes using math formulas. the wrong way to do things is IMO enough to reject any subsequent result no matter if they agree with common knowledge or not.

 

[pinnahertz]

That's precisely the word that I was expecting to read from someone intellectually honest and knowledgeable about it: there are no proven facts but statistical "truths". Thanks.

In other words: even though I have tried to carefully set this up, with reasonable gear and non-biased subjects (2 of them at least), the procedure I have followed here "might" not be within the .99999% percentile of the statistical "truth" for it has not followed a much more strict "scientific" procedure (which would not prove anything neither), yet it is still relevant for the average ear (=non dog ears ), right?

But what I think you're trying to say is that your sighted, non-controlled, biased test is just as good as a controlled ABX/DBT with many times your data set because when the results are reduced to threshold based binary they would be the same, or that both are equally good because no absolutes can be proven.

Well, good thing you aren't doing drug testing.

The concept of S/N ratio exists in statistics as well as audio. Bias in data is noise. When a data set is small it's difficult to separate signal and noise. And when it's small enough, S/N discrimination becomes impossible. Two trials, for example, even if the result is identical, is still nearly all noise.

You simply cannot look at test results like these as 100%, that's not even the goal. But to think an uncontrolled, sighted test is at all comparable with a blind, controlled one is just ridiculous.

 

[bigshot]

I think his test works fine for his purposes. More audiophiles should do casual comparison tests. They would have a better perspective on things than just taking sales literature's word for everything. Demanding clinical testing standards to be able to come to a conclusion about whether a DAC sounds different or not only discourages people from doing any testing at all. If the difference is so small it doesn't show up in his test, it doesn't matter.

 

[pinnahertz]

What does jitter sound like?

Noise or one or more discrete tones modulated by the desired signal. Depending on frequencies involved it generally falls below masking profiles. Means it's not audible.

 

[pinnahertz]

I think his test works fine for his purposes. More audiophiles should do casual comparison tests. They would have a better perspective on things than just taking sales literature's word for everything. Demanding clinical testing standards to be able to come to a conclusion about whether a DAC sounds different or not only discourages people from doing any testing at all. If the difference is so small it doesn't show up in his test, it doesn't matter.

But what would you say about the same test, but with results that shows audible differences that reinforce audiophile mythology? With that methodology it could easily go either way.

 

[bigshot]

If he could clearly hear a difference, I would say it matters. If he has to struggle to hear a difference, it probably doesn't.