[gregorio]

[1] I am a proponent of how our perceptions are always changing and can effect how we can "hear" something from day to day.
[2] The main "controlled" anecdote I have is the computer vs iPhone source through my Benchmark.
[3] My bias was that they'd sound the same: so I was taken aback that they sounded different.

1. I'm a proponent of our perceptions do change, subtly over time and sometimes quickly and quite dramatically but regardless, our perception constantly affects how/what we "hear". Just the fact that you listen to "music" and can differentiate music from semi-random sound/noise demonstrates that fact.

2. I'm not sure how "controlled" your test was, or even if you were testing what you thought you were testing. For example, you might have just been testing an impedance difference in the supplied signal, noise/interference from the computer's USB output the DAC was not expecting/designed to cope with on it's coax input or various other possibilities.

3. A lot of audiophiles fall into the trap of thinking that "bias" is that thing which they consciously expect. This is an erroneous belief! We all have numerous biases, some/many of which are sub-conscious and can take precedence over conscious biases. Furthermore, expectation bias itself isn't just one thing, our "expectation" is usually a judgement comprised of various different expectations and again, some of our expectations are sub-conscious and therefore not taken into account. An example of this is the Mcgurk Effect, watch this 3 min video. but first I'll tell you (even though it states it in the video) there is no "faa", you only ever hear "baa", there's no difference. You should now have the conscious expectation of there being no difference and you shouldn't hear "faa".

- Blind listening comparisons with matching of volumes and music segments, and ability to switch rapidly: should be better than casual sighted listening comparisons, but could still be unreliable, especially for detection of small objective differences in the physical sound

Sure but why doesn't your summary include the other obvious missing points? Shouldn't your summary effectively be:

1. Casual sighted tests can be (typically are) unreliable.
2. Blind testing has it faults but is typically more reliable.
3. Double blind tests are typically more reliable still.
4. DBX test are typically the most reliable.

No test is absolutely reliable although with a sufficient sample size we can get close. The gap between the first point and the second is typically larger than between the second point and the rest because a casual sighted test is by definition "casual", while the other testing methods are usually less so. For example, as you mention, blind testing usually includes matching volumes and material, thereby eliminating potentially serious testing flaws. And, ABX testing typically allows instant switching and the comparison of extremely short extracts, thereby eliminating the limits of echoic memory which can affect the other testing methods.

G

 

[KeithEmo]

The general connotation of "resolving" is that "it allows you to hear differences in other components or the music itself more easily". So, for example, if you heard a distinct difference between two DACs when listening with one pair of headphones, but not with another, then the one that enabled you to hear the difference would be "more resolving". While many audiophiles seem to believe that this means that "the more resolving component is 'better'", that may not always be the case. It's important to note that this can mean different things, or different combinations of characteristics, depending on the person speaking, and the context. It's equally important to keep in mind that resolving is not necessarily synonymous with "good". For example, a speaker with a huge lump in the treble response will make it easier to resolve certain types of distortion - by exaggerating them. That obviously does NOT mean that it sounds good or is especially accurate.

It is worth noting, however, that any audio system is subject to having a weakest link. So, for example, if your speakers have tweeters whose response doesn't extend all the way up, then they might prevent you from hearing a difference between two other components that have differences in high frequency response. (This might limit the sound quality they can deliver with high quality program material; but it might also hide the flaws in low quality MP3 files, which might be considered to be a benefit.)

Also note that "the weakest link" is not always a simple distinction. For example, some tweeters have a lot of energy storage.... if you ask them to play a single sharp high frequency tone they continue to ring after the tone ends. However, they will still test as having a flat overall frequency response. This is easily seen on their waterfall plot. We might well consider that 50 milliseconds of ringing on a tweeter may be totally inaudible. However, without knowing whether that's always true or not, we can say that we are definitely NOT going to be able to assess whether a DAC that exhibits 5 msec of ringing sounds better than one that exhibits 10 msec of ringing if the speakers we've chosen for our test setup exhibit ten times that much... because the speakers will obviously mask the difference - ensuring that we won't be able to hear it.

So, WITHOUT getting into an argument about whether the difference between 5 msec of ringing on a DAC is audible or not, attempting to make that determination using a speaker that exhibits far more ringing would invalidate the test entirely. (We call that "a flawed test protocol".)

Whatever resolving means... whether it's a balanced response or low distortion levels, the Oppo PM-1s have it. And they work with an iPhone without an amp. I think people just like the complication of a convoluted system to haul around in a backpack.

 

[Phronesis]

Sure but why doesn't your summary include the other obvious missing points? Shouldn't your summary effectively be:

1. Casual sighted tests can be (typically are) unreliable.
2. Blind testing has it faults but is typically more reliable.
3. Double blind tests are typically more reliable still.
4. DBX test are typically the most reliable.

No test is absolutely reliable although with a sufficient sample size we can get close. The gap between the first point and the second is typically larger than between the second point and the rest because a casual sighted test is by definition "casual", while the other testing methods are usually less so. For example, as you mention, blind testing usually includes matching volumes and material, thereby eliminating potentially serious testing flaws. And, ABX testing typically allows instant switching and the comparison of extremely short extracts, thereby eliminating the limits of echoic memory which can affect the other testing methods.

G

I could go about midway between what I originally said and what you said. The only way we can "know" how perception and memory are affecting results of listening comparisons (regardless of the protocol) is comparison with objective information on the physical sound which tells us the true objective differences. We generally don't have that type of objective information available, and statistics have only limited benefit in helping us get to the truth if the subjective "measurements" of audio perception and memory are inherently variable and unreliable. It's like trying to figure out how different two signals may be, with each signal having a lot of random noise added to it; if there's a small difference in the signals, the stats may miss that because the difference in the signals is swamped by the random noise.

 

[Davesrose]

2. I'm not sure how "controlled" your test was, or even if you were testing what you thought you were testing. For example, you might have just been testing an impedance difference in the supplied signal, noise/interference from the computer's USB output the DAC was not expecting/designed to cope with on it's coax input or various other possibilities.

3. A lot of audiophiles fall into the trap of thinking that "bias" is that thing which they consciously expect. This is an erroneous belief! We all have numerous biases, some/many of which are sub-conscious and can take precedence over conscious biases. Furthermore, expectation bias itself isn't just one thing, our "expectation" is usually a judgement comprised of various different expectations and again, some of our expectations are sub-conscious and therefore not taken into account. An example of this is the Mcgurk Effect, watch this 3 min video. but first I'll tell you (even though it states it in the video) there is no "faa", you only ever hear "baa", there's no difference. You should now have the conscious expectation of there being no difference and you shouldn't hear "faa".

Maybe it was differences in impedance with iPhone vs computer output, or other factors. As I stated, I don't know the reason, but my expectations were not confirmed when these digital sources sounded different. Also, the Mcgurk effect is not relevant as I'm comparing sound during approximately the same time and with consistent stimuli. Your video is showing the change of audio perception from visual perception (and that if you close your eyes, you may hear something differently then when you're watching the video).

 

[Zapp_Fan]

If there really is an audible difference when you're just changing the digital source and the analog stage is the same, then there's something wrong. Especially if the difference is in tonality. More likely some weird impedance problem, than you get more bass from a different copy of the same file. Maybe there was a grounding problem or something.

Viewed another way, if there was an interesting way to just plain old resample a signal (read: treat it as coming from a different digital source) that resulted in (sometimes) pleasant tonality changes, you might expect to see VST plugins for musicians that made use of that effect. There are no such plugins to my knowledge. You can resample something back and forth from 44.1 to 88.2 and back 1000 times and you're not going to get more bass. There are plenty that implement different types of dithering and resampling, for "final touch before exporting" work on mixes, but none are touted as changing FR in a meaningful way.

This suggests that there is a problem and either the DAC is designed wrong, or the effect is coming in somewhere else, or it just wasn't a real effect at all.

 

[Davesrose]

If you're replying to me, I've never heard anyone say Benchmark DACs are faulty or that they intentionally color sound. My observation was two different digital transports going to the same DAC (playing the same digital file).

 

[KeithEmo]

Errrr.... yes and no.

Saying "plain old resampling of the signal" is an oversimplification.
In general, altering the sample rate involves some filtering.
If you look at the software you're using to convert between sample rates, you'll USUALLY find options for what sort of filtering to use (in R8Brain, or Izotope, you can choose both the sharpness and the cutoff frequency of the filter).
And, yes, you will hear slight differences in the output when you change those settings (those differences are usually pretty subtle).
And, yes, some of those differences involve audible differences in high-frequency response (which can affect "the tonal balance").
And, yes, some pros DO prefer one converter over another "because of how it sounds".

If you want to see a comparison of a whole bunch of sample rate converters.... check out this website:
http://src.infinitewave.ca/

You'll see that many sample rate conversion utilities do deliver virtually identical outputs...
But a surprising number actually DO cause significant alterations to the signal...

I would agree that, with a DAC, assuming that the data remains the same, then the resulting analog audio should sound identical.
However, apparently, on some DACs, the various TYPES of inputs have considerably different performance... particularly when it comes to rejecting jitter present on the incoming source.
(Feel free to argue for another ten pages about whether the difference should be audible or not... but you can readily measure it.)

I would also add that, subjectively, many people I know agree that Sabre DACs tend to emphasize high frequencies, resulting in a sound that seems more detailed, yet slightly unnatural. I'm not at all interested in arguing about "how we must be imagining it" - nor do I know anyone who has done proper tests to reasonably confirm or deny it. HOWEVER, in their early product literature, ESS (who makes Sabre DACs) claimed that, rather than design their DACs for the most accurate response, they used focus groups to "choose the output filter that most people preferred the sound of". They at least claimed, at that point, not that their DACs were accurate, but that they "sounded different and better". They have notably omitted this claim in recent years.

If there really is an audible difference when you're just changing the digital source and the analog stage is the same, then there's something wrong. Especially if the difference is in tonality. More likely some weird impedance problem, than you get more bass from a different copy of the same file. Maybe there was a grounding problem or something.

Viewed another way, if there was an interesting way to just plain old resample a signal (read: treat it as coming from a different digital source) that resulted in (sometimes) pleasant tonality changes, you might expect to see VST plugins for musicians that made use of that effect. There are no such plugins to my knowledge. You can resample something back and forth from 44.1 to 88.2 and back 1000 times and you're not going to get more bass. There are plenty that implement different types of dithering and resampling, for "final touch before exporting" work on mixes, but none are touted as changing FR in a meaningful way.

This suggests that there is a problem and either the DAC is designed wrong, or the effect is coming in somewhere else, or it just wasn't a real effect at all.

 

[gregorio]

... [1] but my expectations were not confirmed when these digital sources sounded different.
[2] Also, the Mcgurk effect is not relevant as I'm comparing sound during approximately the same time and with consistent stimuli. Your video is showing the change of audio perception from visual perception (and that if you close your eyes, you may hear something differently then when you're watching the video).

1. What "expectations" your judgement of your conscious expectation or all your actual expectations?
2. The video demonstrates how audio perception is changed by a sub-conscious expectation (caused by a visual stimulus), even when your conscious expectation was for there to be no change/difference. How do you know the difference you heard in your test was not the same process, a sub-conscious expectation outweighing both what you actually heard and your conscious expectation?

G

 

[Zapp_Fan]

Errrr.... yes and no.

Saying "plain old resampling of the signal" is an oversimplification.
...
I would agree that, with a DAC, assuming that the data remains the same, then the resulting analog audio should sound identical.
However, apparently, on some DACs, the various TYPES of inputs have considerably different performance... particularly when it comes to rejecting jitter present on the incoming source.
(Feel free to argue for another ten pages about whether the difference should be audible or not... but you can readily measure it.)

I would also add that, subjectively, many people I know agree that Sabre DACs tend to emphasize high frequencies, resulting in a sound that seems more detailed, yet slightly unnatural . I'm not at all interested in arguing about "how we must be imagining it" - nor do I know anayone who has done proper tests to reasonably confirm or deny it. HOWEVER, in their early product literature, ESS (who makes Sabre DACs) claimed that, rather than design their DACs for the most accurate response, they used focus groups to "choose the output filter that most people preferred the sound of". They at least claimed, at that point, not that their DACs were accurate, but that they "sounded different and better". They have notably omitted this claim in recent years.

Good point, I was definitely oversimplifying, but to keep it a little more focused, @Davesrose was saying that he heard a noticeable change in low frequencies, which really should not happen regardless of the output filter choice. If I understood correctly, he was saying that the bass tonality changed just from the digital source, not even switching between DACs. I could easily believe that a DAC's filter was deliberately colored, but the effect was supposed to have originated before the DAC.

Restating my point: there is no reason to think that feeding the same file into a DAC from a phone vs. computer should affect anything that's easily noticeable, let alone low frequencies, unless something has gone wrong.

And I do acknowledge there are valid artistic choices to be made between resampling algorithms depending on the input, I was more trying to make the point that it's not a strong effect (and should amount to almost nothing in consumer use) except in wild edge cases.

 

[Davesrose]

Restating my point: there is no reason to think that feeding the same file into a DAC from a phone vs. computer should affect anything that's easily noticeable, let alone low frequencies, unless something has gone wrong.

1. What "expectations" your judgement of your conscious expectation or all your actual expectations?
2. The video demonstrates how audio perception is changed by a sub-conscious expectation (caused by a visual stimulus), even when your conscious expectation was for there to be no change/difference. How do you know the difference you heard in your test was not the same process, a sub-conscious expectation outweighing both what you actually heard and your conscious expectation?

G

My original goal was to see if using the iPhone for my source would be a good system, or if I needed to make it larger with having a computer. My conscious expectation was that the sources would sound the same (as they both were using the same USB adapter going to the Benchmark's coaxial port). My focus at that point of time was just listening to the same track of music (there wasn't a preconceived notion or change in stimuli).

@Zapp_Fan: there definitely was a difference in tonality between the computer and iPhone going to my DAC. On further reflection as to why that is: perhaps the source OS is applying a DSP with the USB adapter (as its seen as a soundcard). Maybe it's just a matter of things being more complicated and the DAC not getting an unaltered bit for bit copy of the file.

 

[Zapp_Fan]

My original goal was to see if using the iPhone for my source would be a good system, or if I needed to make it larger with having a computer. My conscious expectation was that the sources would sound the same (as they both were using the same USB adapter going to the Benchmark's coaxial port). My focus at that point of time was just listening to the same track of music (there wasn't a preconceived notion or change in stimuli).

@Zapp_Fan: there definitely was a difference in tonality between the computer and iPhone going to my DAC. On further reflection as to why that is: perhaps the source OS is applying a DSP with the USB adapter (as its seen as a soundcard). Maybe it's just a matter of things being more complicated and the DAC not getting an unaltered bit for bit copy of the file.

Yeah, the most straightforward explanation I can think of is that the OS applies some kind of filter for some reason. Or that you just heard a difference for psychological reasons, or your headphones shifted on your head slightly, etc.

 

[bigshot]

Definition (or resolution) - The ability of a component to reveal the subtle information that is fundamental to high fidelity sound.

What aspect of sound reproduction is responsible for that? Frequency response? Distortion? Timing accuracy? Proper dynamics? Everything? Because if the answer is "everything" it isn't a very useful term when you are looking to find solutions to problems.

A lot of audiophile terms are vague and only serve as generalized praise. That may be fine for advertorial where they're trying to sell you something, but it's completely useless for identifying *why* one thing is better than another. All it tells you is someone's subjective conclusion. It doesn't tell you anything about how the person arrived at that conclusion. If I say this amp has audible distortion in the 1kHz to 4kHz range, or if I say the response is boosted in the upper mids, you're going to have a pretty good idea how it sounds... nasty and strident. But if I just say it sounds nasty and strident, you really don't know at all how it sounds.

If someone is going to be serious about sound reproduction, I would expect them to try to describe things with precision, so people know exactly what they're saying. But in audiophile circles, writers usually are more interested in being an "authority" and having their conclusions accepted without question than conveying information.

Nothing new is likely to be said in the revival of this thread, so I'll just repeat what I've said before: the issues of variability and unreliability of perception and memory which plague casual sighted listening comparisons are also a problem for blind comparisons.

This isn't a revived thread. It's pinned because that first post is vitally important to what we talk about here in Sound Science. And as it has been explained to you in the past, this thread has absolutely nothing to do with perceptual error. It's about audio fidelity- accurately reproduced sound. We use various techniques to reduce the effect of perceptual error so we can determine the objective facts. "Can a difference be heard or not?" is the basic question. ABX testing is very good at determining that. We don't enter into subjective questions like "Do I like this better?" or "Does this audio component make me feel better?" Each person answers those questions for themselves. We just want to know if the signal going in sounds the same as the signal coming out.

 

[Phronesis]

If I say this amp has audible distortion in the 1kHz to 4kHz range, or if I say the response is boosted in the upper mids, you're going to have a pretty good idea how it sounds... nasty and strident. But if I just say it sounds nasty and strident, you really don't know at all how it sounds.

There's a contradiction here ...

 

[bigshot]

My original goal was to see if using the iPhone for my source would be a good system, or if I needed to make it larger with having a computer.

Here is the answer to that question...
https://kenrockwell.com/apple/iphone-6-plus.htm#measurements
https://kenrockwell.com/apple/iphone-6s-plus-audio-quality.htm

There's a contradiction here ...

One is a specific way of describing a problem and one is a non specific way that begs a subjective conclusion.

 

[Davesrose]

Here is the answer to that question...
https://kenrockwell.com/apple/iphone-6-plus.htm#measurements
https://kenrockwell.com/apple/iphone-6s-plus-audio-quality.htm

My topic was about using the iPhone as a digital source for an external DAC: not the audio quality of the included amp.