Meze Audio POET - Official impressions and discussion thread

[Dragonmilenario]

Hi @Resolve

I hope this message finds you well. I wanted to share some thoughts on your analyses, aiming to be as polite and fair as possible.

First off, I really appreciate how you communicate and express information. English isn't my first language, but I enjoy listening to you. You have a wealth of knowledge and you're a fantastic communicator, so please take my following critique as constructive.

I'll start with an anecdote to illustrate my point. Last year, during my annual medical check-up at work (common in Europe), some parameters in my blood test were off, and I was advised to retake some tests. I'm quite scared of blood, so a friend recommended a private clinic known for being gentle and respectful when drawing blood. They truly lived up to their reputation. The next day, I received my results via email (for which I paid a substantial amount), and everything was within normal range—a perfect blood test with all parameters correct. Despite this, a doctor from the clinic called me to explain the results (a 3-minute call) even though it was clear from the graphs that everything was fine, and they wanted to charge me quite a bit for this obvious and unnecessary phone call. Eventually, I reached an agreement with them, as they understood that the call was unnecessary to explain results that were clearly reflected in the charts.

This story reflects your approach to audio analyses. You often spend too much time explaining a graph that we all understand and sometimes is even included in the box. Why waste your amazing communication skills talking about things we can already see and know?

In this analysis and even in a clearer way in the one for the Liric II (my favorite headphones, and I've been lucky to try dozens and dozens), you didn't discuss anything beyond the frequency graph. I don't get it.

There are so many parameters that make headphones good or bad, which you don't discuss, such as clarity, coherence, musicality, layer separation (in which the Liric II and probably the Poet excel), spatial effect, stereoscopic presence of the sound, driver speed, mechanical recovery capacity... Truly, you waste your communication talent on the graph that we all see and know.

I hope you see this as constructive criticism. We listen to you and follow you for your knowledge because we want to know how a specific piece of equipment sounds, not for you to explain the graph and repeatedly mention that the 17-18Khz peak of the Meze doesn't sit well with you and that potential customers of this headphone can't hear because after 30 it isn't audible or minimal.

Your reviews of the Poet and the Liric II are not up to the level of your communicative skills and knowledge.

Warm regards, and I'll continue listening to you as always, but I expect more from you.

 

[razoredge]

It's becoming increasingly predictable which media channels or analysts will favor certain headphones based on the brand. It's disheartening to see the 'professional' community reach this point.

Even before reading the reviews, I could already tell that Resolve or Super Review wouldn't be fans of the Poet… And here's my prediction: DMS and Golden Ears won't like them either. You can call me Nostradamus.

And the most curious thing is seeing people of a certain age, who can't possibly hear frequencies above 16kHz, complaining about peaks in the 17-18kHz range...

yeah I find it extremely suspicious that anyone who is willing and able to pay $2K for headphones would be able to hear anything above 14KHz and complain about peaks there. Also suspicious that all of these measurer/reviewers now apparently have their own branded frequency target curves, as if that is something you could just create out of thin air and judge every headphone against. its a mess.

 

[Resolve]

I hope you see this as constructive criticism. We listen to you and follow you for your knowledge because we want to know how a specific piece of equipment sounds, not for you to explain the graph and repeatedly mention that the 17-18Khz peak of the Meze doesn't sit well with you and that potential customers of this headphone can't hear because after 30 it isn't audible or minimal.

I'm not sure where folks are getting the 17-18khz peak ideas from. For me it's a fairly wideband elevation above around 13khz (and this is quite audible. For reference, my hearing goes up to around 18khz still). As to the rest of it... we don't focus primarily on measurements, we focus primarily on frequency response at the eardrum, which is something that's not easily measured for humans. If the effect is "detail, resolution, dynamics" and the typical audiophile lexicon, the cause is highly likely to be frequency response at the ear drum in situ, since it can also be heard to be exactly that, once you're able to parse these effects in terms of various FR features. But I'm not even opposed to talking about the effects, I just don't like using the typical audiophile lexicon because it means different things to different people. Any sense of shared language with regards to these terms is a complete illusion, and this is something you realize very quickly when speaking to people at canjams, and how they attach meaning to the things they're hearing.

 

[Dragonmilenario]

I'm not sure where folks are getting the 17-18khz peak ideas from. For me it's a fairly wideband elevation above around 13khz (and this is quite audible. For reference, my hearing goes up to around 18khz still). As to the rest of it... we don't focus primarily on measurements, we focus primarily on frequency response at the eardrum, which is something that's not easily measured for humans. If the effect is "detail, resolution, dynamics" and the typical audiophile lexicon, the cause is highly likely to be frequency response at the ear drum in situ, since it can also be heard to be exactly that, once you're able to parse these effects in terms of various FR features. But I'm not even opposed to talking about the effects, I just don't like using the typical audiophile lexicon because it means different things to different people. Any sense of shared language with regards to these terms is a complete illusion, and this is something you realize very quickly when speaking to people at canjams, and how they attach meaning to the things they're hearing.

Thank you for the response. I don't agree that the frequency response is significant in determining the quality of headphones. There are models on the market that measure extremely well for $100 or $150, and we seem foolish spending thousands if it”s so.

The virtues of most headphones we can call Hi-Fi cannot be captured in a frequency curve, is just a small part of the overall quality of headphones.

It's as reductive as basing the value of a Michelin-starred restaurant on the nutritional information of the food.

 

[Resolve]

Ahh but see this is only the traditional view of FR. You're judging it based on a graph, not how sound is propagating at your eardrum. "Measures well" is based on one head and one set of ears, it gives you the picture of that FR, and ideally relative to the HRTF of the measurement system. It does not give you the picture of the sound at your eardrum relative to your HRTF. This really is the missing link between what we see on the graph and what we experience subjectively, apart from any other psychoacoustic priming like openness or whatever, and once you're able to think of it in terms of FR, the technicalities language stops making any sense. I know it's a big step to try and think of it that way, but the more you dig into it the more it all points to this.

 

[Kevintj604]

Ahh but see this is only the traditional view of FR. You're judging it based on a graph, not how sound is propagating at your eardrum. "Measures well" is based on one head and one set of ears, it gives you the picture of that FR, and ideally relative to the HRTF of the measurement system. It does not give you the picture of the sound at your eardrum relative to your HRTF. This really is the missing link between what we see on the graph and what we experience subjectively, apart from any other psychoacoustic priming like openness or whatever, and once you're able to think of it in terms of FR, the technicalities language stops making any sense. I know it's a big step to try and think of it that way, but the more you dig into it the more it all points to this.

Don’t know why you and I are both up at 5:00am responding to HeadFi threads instead of sleeping but it is what it is I guess. Taron better be paying you overtime.

I’m curious as to the other intangibles that can’t be explained by FR as well. You’ve referred to the audiophile lexicon so I’m curious how properties like punch and slam, soundstage, resolution, imaging are explained in frequency response when we all seem to have the same experiences subjectively in hearing those properties but all have different eardrums.

If everything boils down to FR at each specific individuals HRTF wouldn’t we all perceive the properties of the headphones differently?

 

[Reputator]

The decision to not evaluate huge swaths of a headphone's character because you aren't happy with the language for describing sound is... certainly a choice.

I've been watching Resolve's reviews for several years (before The Headphone Show when it was just Resolve) and I have a deep respect for his work, not just reviewing but also for the hobby in general. That said, I can 100% understand the frustration a viewer would have watching a review of his today.

I watched the POET review, and it is focused almost entirely on timbre. At one point he says they are engaging, and that's maybe the one-time deviation from that. Contrast that with an older review of his like of the HEDDphone ONE, where he would describe things like how there's a sense of impact throughout the frequency range. Talk of spaciousness, decay, separation, etc that used to be more present in his reviews are now rare. I guess those words are just too confusing, huh?

But those are important metrics for the vast majority of prospective buyers. Yeah timbre might be the most important, but it's not the full picture. It's like if you reviewed food only describing the flavor, but giving no mention of the texture.

 

[Resolve]

Don’t know why you and I are both up at 5:00am responding to HeadFi threads instead of sleeping but it is what it is I guess. Taron better be paying you overtime.

I’m curious as to the other intangibles that can’t be explained by FR as well. You’ve referred to the audiophile lexicon so I’m curious how properties like punch and slam, soundstage, resolution, imaging are explained in frequency response when we all seem to have the same experiences subjectively in hearing those properties but all have different eardrums.

If everything boils down to FR at each specific individuals HRTF wouldn’t we all perceive the properties of the headphones differently?

Yes, and yes we do experience many of these things differently. And not only that, the way we attach meaning, where our attention goes, it's all specific to the individual. I won't go into it here as much to not derail the thread, but we have gone into it at length over on our forum.

 

[Resolve]

The decision to not evaluate huge swaths of a headphone's character because you aren't happy with the language for describing sound is... certainly a choice.

I've been watching Resolve's reviews for several years (before The Headphone Show when it was just Resolve) and I have a deep respect for his work, not just reviewing but also for the hobby in general. That said, I can 100% understand the frustration a viewer would have watching a review of his today.

I watched the POET review, and it is focused almost entirely on timbre. At one point he says they are engaging, and that's maybe the one-time deviation from that. Contrast that with an older review of his like of the HEDDphone ONE, where he would describe things like how there's a sense of impact throughout the frequency range. Talk of spaciousness, decay, separation, etc that used to be more present in his reviews are now rare. I guess those words are just too confusing, huh?

But those are important metrics for the vast majority of prospective buyers. Yeah timbre might be the most important, but it's not the full picture. It's like if you reviewed food only describing the flavor, but giving no mention of the texture.

But this is precisely the problem. I may have attached meaning in that way, but it's going to lead to a lot of false positives and false negatives when inevitably other people don't attach meaning in that same way. I can't expect people to be able to peer into my brain. Not only that, I do not wish to contribute to what I believe to be a false narrative, that there is some magical property of transducers that's not being measured. And, even if there are, these description can only ever refer to the goodness/badness assigned by the subjective experience of the individual. So to be clear, I'm fine reporting those subjective qualities, I just don't want to contribute to the "technicalities as separate from FR" line of thinking, since I've come to learn that is false to a large degree.

 

[razoredge]

I'm not sure where folks are getting the 17-18khz peak ideas from. For me it's a fairly wideband elevation above around 13khz (and this is quite audible. For reference, my hearing goes up to around 18khz still). As to the rest of it... we don't focus primarily on measurements, we focus primarily on frequency response at the eardrum, which is something that's not easily measured for humans. If the effect is "detail, resolution, dynamics" and the typical audiophile lexicon, the cause is highly likely to be frequency response at the ear drum in situ, since it can also be heard to be exactly that, once you're able to parse these effects in terms of various FR features. But I'm not even opposed to talking about the effects, I just don't like using the typical audiophile lexicon because it means different things to different people. Any sense of shared language with regards to these terms is a complete illusion, and this is something you realize very quickly when speaking to people at canjams, and how they attach meaning to the things they're hearing.

@Resolve, I like that you guys measure all these headphones and present your data to viewers. I genuinely think that is a good thing. But your comparison to the tungsten gets at part of what people here are finding frustrating. From 20 Hz to 6 kHz, these 2 headphones measure damn near identical. in the high treble, by your own admission, the tungsten measure worse since it has a broad and large (5+ dB) peak from 6 kHz to 10+khz. The poet, on the other hand measures smoothly until 12 kHz or so, and only past that point does it go to the higher side. But you say the Poet is "several steps behind" the tungsten in performance because your ears are shaped differently and so you heard the high treble differently from how you measured it. That is all well, but why measure anything above 6 kHz then? if so much of the variation in performance is going to be individual to each person in these parts of the frequency response, why show it? why use it to justify whether something is good or not? you'd only be making statements about your own ears at that point, and not about the headphone in question.
The accusations about brand favoritism, I think, are about your use of this (supposedly objective) data to support certain conclusions when the data agrees with your subjective experience (which is generally biased for all of us), and to deny its relevance when it doesn't match your subjective experience (as in this poet review). Sadly, I think what an earlier commenter said about being able to predict which reviewers will like a given headphone just by looking at a measurement or even brand (which typically, but not in all cases, correlates to tuning which correlates to measurement) is correct. It was easy to predict that you guys would like the Empyrean II but not the Poet, for example.
I hope you can answer my question of why you continue to show and use measurements in the high treble if their correlation to actual subjective performance is so varied between individuals? I am genuinely curious. Thanks!

 

[Dragonmilenario]

For example, I switched from the Meze 109 Pro to the Liric II (I used both for many months and naturally stopped listening to the 109s).

There is no measurable reason to explain my change. I didn't switch to get better timbre or a frequency response closer to what is considered "ideal."

The improvement between the two is related to intangibles such as musicality, layer separation (how each instrument is perceived separately and not all together), bass that doesn't blur other sounds, quick and with an astonishing decay, much more natural and realistic voices, and a sense of being in the music that a lower-level headphone cannot offer, even if it has a more "correct" timbre or frequency response.

Where is all of this in Headphones.com's analyses? Not everything is about timbre; it's only a part. A good timbre can be achieved for $100, you don't need an analyst to telling you that.

 

[Resolve]

@Resolve, I like that you guys measure all these headphones and present your data to viewers. I genuinely think that is a good thing. But your comparison to the tungsten gets at part of what people here are finding frustrating. From 20 Hz to 6 kHz, these 2 headphones measure damn near identical. in the high treble, by your own admission, the tungsten measure worse since it has a broad and large (5+ dB) peak from 6 kHz to 10+khz. The poet, on the other hand measures smoothly until 12 kHz or so, and only past that point does it go to the higher side. But you say the Poet is "several steps behind" the tungsten in performance because your ears are shaped differently and so you heard the high treble differently from how you measured it. That is all well, but why measure anything above 6 kHz then? if so much of the variation in performance is going to be individual to each person in these parts of the frequency response, why show it? why use it to justify whether something is good or not? you'd only be making statements about your own ears at that point, and not about the headphone in question.
The accusations about brand favoritism, I think, are about your use of this (supposedly objective) data to support certain conclusions when the data agrees with your subjective experience (which is generally biased for all of us), and to deny its relevance when it doesn't match your subjective experience (as in this poet review). Sadly, I think what an earlier commenter said about being able to predict which reviewers will like a given headphone just by looking at a measurement or even brand (which typically, but not in all cases, correlates to tuning which correlates to measurement) is correct. It was easy to predict that you guys would like the Empyrean II but not the Poet, for example.
I hope you can answer my question of why you continue to show and use measurements in the high treble if their correlation to actual subjective performance is so varied between individuals? I am genuinely curious. Thanks!

It's a good question tbh. Rig ears by their very nature aren't going to be the best proxy for individual human ears, given how much they can vary from person to person. And this becomes particularly relevant above 3khz or so where the pinna effect variation becomes more substantial.

I have several different answers as to why it still matters that we show this information:

1. It's an additional data point for how a human being could hear these products. Note the wording here is not "how all humans hear these products".
2. Despite the fact that there is variation for pinna effects, we do all typically have heads and ears, and so there will also be trends. The measurement gives you an indication of that trend.
3. There are certain design choices that make the measured results more or less predictive. So closed back headphones would be the least predictive given leakage effects and also high acoustic impedance, while very open, positionally consistent headphones will be much more easily predicted by the rig pinna. We want to ensure that if there is a best case scenario, it's accurately reflected as such.

But yeah, the general read of measurement data as "this is how the headphone performs" is an extremely limited perspective and should be checked at every opportunity. It's just one condition, one data point in addition to your own subjective experience. It's not telling you "this is how it will sound to you", it's telling you "this is how the rig heard it". And if I'm telling people about my experience, it makes perfect sense to supplement that with data for an additional viewpoint.

The point with measurements is that it's the additional perspective, and a visualization that indicates it's not anything goes because we are all humans with heads and ears. But at the same time, we shouldn't treat it as indicative of "the sole truth about how this headphone performs". Change the condition to a different head and it may perform differently.

 

[waveSounds]

I always find the attempt to add objectivity to something as subjective as one's musical preferences (as that what audio is satisfying at the end of it all) fairly amusing.

Perhaps it could be said that classical music is objectively better than happy hardcore, but there will always be those who prefer happy hardcore.

Same thing with headphones. Measurements will only ever tell what something measures like. I suppose this is good from a standardisation point of reference for some, but it is pretty insignificant outside of that.

You have people who adore audio products that measure 'awfully', according to what some people have decided to good measurements should look like, and who's to tell them they're wrong?! If they like how it sounds then wether it measures like a cardiogram report literally means nothing in terms of enjoyment. No?

I don't think their use is unnecessary, but in the domain of personal audio primarily for entertainment purposes they should be a footnote on the matters of enjoyment. Yeah, you might say that how a set of cans makes a reviewer feel cannot be transferred to others, but... that's the same case with measurements! It's just far less interesting to read about how graph go up here and uppy-down there, versus how it emotionally connected you to music. I don't buy headphones because they measure well, I buy them because they allow me to better enjoy my music with. Maybe I'm weird in that respect.

Now, for the love of God, can someone please tell us how the Poet makes you feel.

Edit - This wasn't aimed at you @Resolve, I've been watching your content for a long time and think you're a good reviewer, the whole conversation prior was just a useful segue for me to express my viewpoint.

 

[Resolve]

Now, for the love of God, can someone please tell us how the Poet makes you feel.

Bold of you to assume I'm capable of feeling things...


But since you asked, it makes me feel... Beguiled. Or maybe it's the music doing that, I don't know.

 

[Reputator]

Driver speed and excursion force are two very important aspects that aren't measured by a sine wave. Talking about speed and time domain effects is no less useful than relaying your own subjective experiences about timbre.

Anyway, I don't really agree that reviews being predictable based on FR is a growing problem, or even a problem at all. Listening preferences don't tend stray THAT drastically among most people, and there’s plenty of data to support that. Timbre is also generally the most important metric among the majority of listeners. So yes, a headphone with a FR as wonky as the Audivina can't realistically be expected to result in evenly split reviews. It's going to result in most reviews coming out against it, with an expectedly small number of outliers that enjoy it. Yes we're all different, but most of us aren't lightyears apart from each other.