[bigshot]

Measure those differences and let me know what you find out. I have 8 or 9 iPods and they all sound exactly the same from line out... flat response and clean.

 

[customNuts]

Measure those differences and let me know what you find out. I have 8 or 9 iPods and they all sound exactly the same from line out... flat response and clean.

 
I have measured them, with my ears. Since thats how we actually listen to music, I find it a far more accurate way to determine differences then by comparing numbers - especially when it comes to sound signatures. If you can't hear any difference then fair enough, but that doesn't mean the majority of audiophiles can't. Its not surprising when the models have different DAC's, different amp modules, different OS's, different firmware etc, etc. In fact with that many variables, I would expect it impossible for them all to sound exactly then same. 
I do agree that the differences in general are small and they are flat & clean although quite musical (as a portable player should be imo).

 

[higbvuyb]

   but that doesn't mean the majority of audiophiles can't. 

Nobody is saying that you can't hear the difference; it's just that there's no evidence that the difference you are 'hearing' actually exists.

 

[xnor]

  Nobody is saying that you can't hear the difference;

 
Indeed.
I guess everyone can hear a difference depending on personal mood, how much the person has slept, when the person has a cold, etc. But the physical sound waves that are being output are virtually identical so there is no objective difference.
 
That's the main problem when people compare audio components - bias, changing interpretations/feelings/imaginings. Well, the real problem is attributing those subjective impressions solely to the components' performance.
 
Correlation does not imply causation. When a component A sounds better than component B in an uncontrolled comparison it does not mean B has higher sound quality. Bias can easily influence what you hear, probably preferring A because it's more expensive, looks better, is newer, has some rave reviews (FOTM trap) ... or it's due to trivial things like improper level matching.

 

[customNuts]

   
Indeed.
I guess everyone can hear a difference depending on personal mood, how much the person has slept, when the person has a cold, etc. But the physical sound waves that are being output are virtually identical so there is no objective difference.
 
That's the main problem when people compare audio components - bias, changing interpretations/feelings/imaginings. Well, the real problem is attributing those subjective impressions solely to the components' performance.
 
Correlation does not imply causation. When a component A sounds better than component B in an uncontrolled comparison it does not mean B has higher sound quality. Bias can easily influence what you hear, probably preferring A because it's more expensive, looks better, is newer, has some rave reviews (FOTM trap) ... or it's due to trivial things like improper level matching.

Evaluating in an uncontrolled environment would throw endless variables into the test making it completely invalid. I thought that was a given.  Clearly the testing needs to be as controlled as possible. If the evaluator is unable to do that then then the review looses all credibility. 

 

[customNuts]

  Nobody is saying that you can't hear the difference; it's just that there's no evidence that the difference you are 'hearing' actually exists.

 
I get you but my point is the evidence isn't just numbers. Its also what people actually hear which is not entirely measurable - hence the many discrepancies between numbers/charts and actual sound quality & sound signatures. They just don't match up a lot of the time. Thats the mystery of the human ear & of course the endless variables that effect opinions as stated by xnor. Since hearing is the true medium for listening to music as opposed to evaluating numbers,I choose to say that there is substantial 'evidence' through that medium also.

 

[xnor]

  I thought that was a given.

It really should be and I wish it was, but last time I checked some amp/DAC comparison reviews there was either no mention of level-matching at all, or if there was, it was "done by ear".
 
 

  I get you but my point is the evidence isn't just numbers. Its also what people actually hear which is not entirely measurable - hence the many discrepancies between numbers/charts and actual sound quality & sound signatures. They just don't match up a lot of the time. Thats the mystery of the human ear & of course the endless variables that effect opinions as stated by xnor. Since hearing is the true medium for listening to music as opposed to evaluating numbers,I choose to say that there is substantial 'evidence' through that medium also.

Here is where it gets hairy.
 
We can for example measure differences in frequency response that nobody can distinguish by ear, or tiny distortion levels etc. The problem is again not in that those measurements are wrong, but that people expect worse measurements to sound worse completely ignoring the limits of our hearing. So this is one reason why they ignore measurements.
It gets worse when people try to forcibly match measurements with their biased, subjective impressions. Most of the time that doesn't work (due to many uncontrolled variables, see above), so that's a reason why people may reject measurements.
 
If a couple of people tell you that amp A sounds "warmer" than B, guess what, it's gonna sound at least a bit more warm than B even if the frequency response is completely flat for both. Only when you do something like an ABX test you will notice that both sound equally "warm".
If they don't then the FR measurement was done unloaded or the effect of the output impedance was ignored. Does that make the unloaded FR measurement invalid per se? Nope.. just not very useful.
 
I don't think there is a mystery to measurements at all, it's just a lot of bias and bad tests that skews expectations. The mystery to me is why DBTs are not the standard in audio.
 
Sure, I agree that what you hear is more important than numbers, BUT only if you've accounted for bias in your listening tests. Otherwise it's just anecdotal evidence which is usually dismissed in science (for a good reason)!

 

[customNuts]

  It really should be and I wish it was, but last time I checked some amp/DAC comparison reviews there was either no mention of level-matching at all, or if there was, it was "done by ear".
 
 
Here is where it gets hairy.
 
We can for example measure differences in frequency response that nobody can distinguish by ear, or tiny distortion levels etc. The problem is again not in that those measurements are wrong, but that people expect worse measurements to sound worse completely ignoring the limits of our hearing. So this is one reason why they ignore measurements.
It gets worse when people try to forcibly match measurements with their biased, subjective impressions. Most of the time that doesn't work (due to many uncontrolled variables, see above), so that's a reason why people may reject measurements.
 
If a couple of people tell you that amp A sounds "warmer" than B, guess what, it's gonna sound at least a bit more warm than B even if the frequency response is completely flat for both. Only when you do something like an ABX test you will notice that both sound equally "warm".
If they don't then the FR measurement was done unloaded or the effect of the output impedance was ignored. Does that make the unloaded FR measurement invalid per se? Nope.. just not very useful.
 
I don't think there is a mystery to measurements at all, it's just a lot of bias and bad tests that skews expectations. The mystery to me is why DBTs are not the standard in audio.
 
Sure, I agree that what you hear is more important than numbers, BUT only if you've accounted for bias in your listening tests. Otherwise it's just anecdotal evidence which is usually dismissed in science (for a good reason)!

 
I actually agree with what your saying. Especially:
 
"I don't think there is a mystery to measurements at all, it's just a lot of bias and bad tests that skews expectations."
 
My point is - which you've covered - That accurate measurements and accurate hearing tests are both viable mediums for evaluation. Ultimately, in an ultra controlled environment - both would be used in conjunction. 

 

[bigshot]

Evaluating in an uncontrolled environment would throw endless variables into the test making it completely invalid. I thought that was a given.  Clearly the testing needs to be as controlled as possible. If the evaluator is unable to do that then then the review looses all credibility. 

When I compared my iPods to see if there were differences, I used the same AIFF file in each one, set up a line level matched switcher, and did direct A/B comparisons using several pairs of headphones and my main speaker rig. I couldn't hear any differences between them at all.

How did you control the variables when you determined that there was a difference?

 

[MagnumOpus]

I just replaced my HTC One S with the iPhone 5 and the difference is monumental. The iPhone sounds thin, sharp, cold, dynamics are gone, bass is gone, highs are unpleasant and mids are not worth mentioning. Like a cat fighting its way out of a tin can.

I dont want this phone at all.

 

[higbvuyb]

I just replaced my HTC One S with the iPhone 5 and the difference is monumental. The iPhone sounds thin, sharp, cold, dynamics are gone, bass is gone, highs are unpleasant and mids are not worth mentioning. Like a cat fighting its way out of a tin can.

I dont want this phone at all.

The HTC One S has 'beats audio' which is apparently an EQ setting that makes it more 'fun' but actually decreases fidelity.

 

[SilverEars]

I just replaced my HTC One S with the iPhone 5 and the difference is monumental. The iPhone sounds thin, sharp, cold, dynamics are gone, bass is gone, highs are unpleasant and mids are not worth mentioning. Like a cat fighting its way out of a tin can.

I dont want this phone at all.

This is quite true when I heard the iphone 5.  Wasn't pleased with the sound There is no warmth.  I like a bit of warmth to the sound, it add perceived sound stage.  I have the Galaxy S4 with Neurtron Player App, and it sounds very clear with perceived sound stage and nice vocals.  The app uses the processor more intensively for DSP purposes to output better processed sound via software.

 

[MagnumOpus]

I

The HTC One S has 'beats audio' which is apparently an EQ setting that makes it more 'fun' but actually decreases fidelity.

never use the beats EQ..

 

[MagnumOpus]

Ca

This is quite true when I heard the iphone 5.  Wasn't pleased with the sound There is no warmth.  I like a bit of warmth to the sound, it add perceived sound stage.  I have the Galaxy S4 with Neurtron Player App, and it sounds very clear with perceived sound stage and nice vocals.  The app uses the processor more intensively for DSP purposes to output better processed sound via software.

Can You recommend a dsp for iphone?

 

[higbvuyb]

I
never use the beats EQ..

The iPhone 5 measures pretty much perfect; it isn't the one at fault.