Correct, with some experience it is the best test to check the tonal balance and if there are some peaks in it, an experienced audio engineer once said all loudspeaker listening tests should start with it, it would save a lot of time.
Just out of interest, how does the untrained user know what pink noise is meant to sound like ? Especially If the only transducers they have are a pair of headphones with wonky FR.
They won't, that is why I wrote some past experience is needed.
My bad!
I'm sorry, but the paper linked here has to do with perception of sound quality of music, apparently nothing to do with ability to discern specifics about a headphone's FR. Wider bandwidth music sounding better is orthogonal to the effectiveness of tones vs. PN for EQ.
:shrug: ?
It builds on and corroborates the findings of this paper. I suggest reading this blog post by the author of those papers, Dr Sean Olive.
Now you're getting it.
You've provided zero evidence of this. Whether you could tell them apart with sine sweeps would be irrelevant anyway if you couldn't tell them apart with actual music (which pink noise is spectrally analogous to).
Which means it's not a good test signal for discriminating and reliable judgements of transducer sound quality, as the scientific research by Sean Olive et al above found.
That's not what he said.
It's not 'flaws' in your hearing, it's differences in everyone's hearing (individual ears' 'frequency response') and differences in perception of a transducer's frequency response when listening to sine sweeps vs broadband sound (music or pink noise).
That's not what's being said. False positives (rather than false negatives) are the main problem when using sine tones:
How do you propose I provide evidence of myself thinking something would be easier ... for myself? "Easier" is subjective no matter how much evidence I submit.
That's not what the article you linked actually says. It says pink noise and alt rock are better than less harmonically rich music. That's all. Tones aren't mentioned there.
I get what he's saying and he's completely right to say it depending on context.
Subjective discrimination claims can be tested via blind ABX tests. (Btw looks like I edited my post for clarity before you posted this reply, have a read of it again, your claim is moot anyway.)
That was in reply to your indifferent shrug to my comment that your friend using alt rock music to tune does not necessarily mean it has the smooth broadband spectrum the science says produces the most discriminating and reliable sound quality judgements.
It doesn't look like you do from the below.
If they're completely different to the normal program material (music) that, you know, people actually listen to using the headphones, then they're irrelevant.
We're talking about linear, not nonlinear distortions here (the former contributing the vast majority to perceived sound quality).
I've already told you several times (backed by comments from Sean Olive and Oratory) the trap that you're falling into: false positives, which I clearly explained with a concrete example in this post, which you pretty much ignored. You're just not listening.
No, I totally grasp your arguments, but they're at least half based on arguments that Olive hasn't actually made, (at least not in the sources you've linked) and you're misunderstanding my use case, I think. I'm not actually sure what we disagree about.
As someone who learned to EQ over the last year or two, my biggest takeaway isn't the SQ improvement: rather, it's the education and ear-training that comes from learning to do it well, as someone not trained in electronics or audio engineering. I also liked a lot of music but had no musical training.
He literally calls listening to sine sweeps a bogus test.
Well at least you're now only talking about yourself and no longer prescribing others to follow your misguided methods:
Many people use that EQ strategy, it doesn't mean it's works for music. You could for instance identify two treble peaks at 4kHz and 7kHz. But with musical content it's perceived differently, your auditory system fills in the gaps. If you went on with your method, letting it dictate EQ moves you might have EQ'd away good detail for arbitrary reasons because the sine sweep gave you a false sense of certainty.
Are you asking:
So true!
That is a nice feature. I find that if I actually get the response reasonably flat, everything sounds good. Some tracks obviously better than others.
Well, not if you know what peaks are supposed to be there and which aren't. Which, according to everyone arguing about this, including myself, knowing this is part of SOP for headphone tuning.
It's so clear that you made a example unrelated to my example, and a referenced a counter argument which is different from mine. All that we can agree on is that many people feel it's right do it. That's not very well enough to say optimizing sine sweeps is valid for music. You talk about correcting high Q "flaws" on the basis of knowledge, why stop there? Why no take into account the realities about your auditory system handling musical content and test signals differently?
Except you don't, because you don't know the exact shape of your equal loudness contours.
Because as I've told you multiple times, you don't know if you're 'correcting' the headphone's frequency response or your ear's. And pink noise being a signal more revealing of low-Q than high-Q deviations is exactly what you want, because this reflects the ear's greater sensitivity to low-Q than high-Q deviations, as found in the scientific research of Dr Floyd Toole. Then as you say there's ringing which can be audible with high-Q positive-gain EQ filters, with suggestive evidence from Jaakko Pasanen (of AuotoEQ fame) pointing towards filters with slopes above 12 or maybe 18 dB/octave decreasing sound quality preference in blind tests.
You're still not getting it. You'd have no idea how much of that 10 kHz spike is down to the headphone and how much is down to your ear. And whether and how much it shifts in amplitude and frequency with slightly different headphone placement (which can be a big effect, especially at high frequencies like 10 kHz). And if you're tuning the headphone for others, whether the spike is due to the interaction between that specific headphone and your specific ear, which could be completely different for others.
I mean, I do, everyone who tests with music does, it's just a lot faster and easier to do certain EQ tasks with tones than music.
I think "no idea" is overstating it somewhat. You're right in principle, in practice I think such a big deviation in your personal loudness contour would be easy to discover by comparing notes with another listener.
Indeed, that's a massive factor and one I try to check for pretty constantly while I'm evaluating a set of cans. However, my personal subjective perception is that swings in high frequency content depending on placement are less severe on ears than they are on a measurement jig. This makes sense since our ears are evolved to do something useful above 10khz and the internal structures of the ear are pretty different than those of a test jig.
Agree, I wouldn't try to flatten anything in the upper treble by ear if it was "for work". And again to be clear, I relied on music more than tones for EQ at work, too. And the opinions of other people... and pink noise, even.
