Dan Clark Audio Stealth Review, Interview, Measurements

[DarginMahkum]

That's a strawman, as that is not what he said at all.

The problem with Amir is, his biases are affected by what his measurements say, and consequently the positivity of his review. He has an entire forum that is supposedly based around "science", but approaches science and the meaning of it in a grossly misleading way. For example, he specifically states, when measuring the frequency response of a headphone, whether it "complies" with the Harman target curve. He treats the curve as being the only correct frequency response a headphone can have. Anyone who has actually read the research, will know that there was considerable variation in both peoples' preferences for the amount of bass, as well as the amount of treble, and that the curve itself is an average of those preferences.

The idea of the sciences is to better understand things, not a means to declare absolute truth, which is more akin to a religion. Amir treats the Harman curve as if it's an absolute truth that must be complied to, or the headphones are faulty, and should be EQ'ed to match the curve. However, what the Harman research has lead us to understand is, that, just as some people prefer the greater mid-bass of the Expanse, and other people prefer the tuning of the Stealth, there is a great variation in what people prefer when it comes to the tuning of headphones. A great deal of this probably relates to the music people listen to, as there is considerable variation in the balance of frequencies between different genres, albums, and even tracks in an album, but also the considerable variation in the frequency sensitivity of peoples' ears.

Scientific experiments are designed in such a way as to help us answer questions about a specific thing in a manner that reduces the possibility of other factors, including the cognitive and other biases of the person conductive them, so that the results can be useful. If a reviewer, having seen measurements of a product, lets himself believe he (or she) is hearing a certain thing because they saw something in a measurement, for the reason I just stated, they are doing the exact opposite of science!

What a waste of time and energy...

 

[ObjectiveSubjectivist]

I said I don't need them, but now I certainly do want them...

If you decide to get them, please report back how they compare against Stealth.
I'm curious

 

[ThanatosVI]

A good Review always includes Subjective opinions on how the reviewer likes the headphones as well as objective data like measurements.

If sibjective hearing of a reviewer is in total contradiction to the measurement, then you know that his hearing ability is not that trustworthy

 

[ThanatosVI]

If you decide to get them, please report back how they compare against Stealth.
I'm curious

I don't have a Stealth, only Expanse.
If I get them it's still a while off tho

Currently I'm still waiting on my delayed Empyrean II and around January is a big hit for my wallet that will take many months to recover from...

 

[Currawong]

If you don't accept your ears are easily tricked by expectation bias and the like you're a prime target for snake oil salesmen.

Likewise, you have to accept that your brain can be fooled by flawed, yet seemingly logical arguments, as a result of one's lack of knowledge of a subject.

 

[plumpudding2]

Likewise, you have to accept that your brain can be fooled by flawed, yet seemingly logical arguments, as a result of one's lack of knowledge of a subject.

True, and the subject of audio engineering is also a very confusing one so without scientific training it is not easy to separate fact from fiction.
My mind was blown when I realized that our recordings aren't band-limited because mathematically speaking a signal can be finite only in either time or frequency domain.
This is getting a bit off-topic though so I'll stop here

 

[Daniel Johnston]

Though measurements are a small part of the characteristics of a headphone => as Harman response target, just reflect 70% of the listeners taste, basically, if you are among the 70%, I think, it's interesting to know and see if the headphone/iem you want to buy is concerning you or not ! I don't think any headphone IEM too far from harman target response can please me.
I imagine, that most people can't try any headphones before buying them in their life as they buy them via internet for more than a decade now... Maybe it's, for me, even a full none sense to try a headphone in a crowded public place (jam or shop).

Raw Measurements seems still like the easiest way to, at least, get this information and order one sample...
Probably, many advices from reviewers can also be a good way to help to choose a headphone. but, something like 70% of the reviewers which are positive on a headphone, that means probably that Harman target is not really far behind the concerned model.

For the 30% who don't like the harman response curve => good luck .. they are in a super niche of a niche market.. and probably some reviewers are also not concerned with harman target frequency response

In a way, I am glad that DCA decided to tune their Stealth to a Harman target, as I can from this target, adapt, with some EQ with (poweramp EQ), to my own taste, which can be with a bit less and clearer low frequencies => without any scientific or not measurements (just my own taste). I would say, I could buy a very good headphone, a bit too pricey, but the best I owned in 40 years with the HEDDPhone one.
So, I feel DarginMahkum has still an interesting point of view, though I don't consider ASR as a scientific website and I hate the binary "Yes or No" recommendation. I don't like too the standard "reviewers" measurements placed in the end of most articles as if it was the peak point and conclusion the article or the video !!!
I always start, like probably many, by reading the charts and then guess if it is worth to read or watch the whole review (sometime 30' or more).

ASR is maybe like for IEMs, with so many users (not scientist at all too), spending time to try to measure every IEM with unreliable mics... https://squig.link/ ..we get some basic info which should be offered to potential users in first place (like Chinese are doing for a while now with IEM and publish the chart directly on the box).
Ultrasone, in Germany, delivers each high-end headphone they sell, with their measurements of the exact headphone you got in the box !!! probably as a proof of their quality control.

To be fair, HeadFi is a niche of a niche. People spending >$500USD is niche as well. I can see using Harman tuning as a baseline for headphones up a certain price point. What that is, I don't know.

If one considers spending multiple thousands of dollars on a headphone, there is no amount of reviews or measurements that substutes for a live demonstration. Luckily manufacturers like DCA have good return policies. Places like Headphones.com have awesome return windows.

 

[number1sixerfan]

That's a strawman, as that is not what he said at all.

The problem with Amir is, his biases are affected by what his measurements say, and consequently the positivity of his review. He has an entire forum that is supposedly based around "science", but approaches science and the meaning of it in a grossly misleading way. For example, he specifically states, when measuring the frequency response of a headphone, whether it "complies" with the Harman target curve. He treats the curve as being the only correct frequency response a headphone can have. Anyone who has actually read the research, will know that there was considerable variation in both peoples' preferences for the amount of bass, as well as the amount of treble, and that the curve itself is an average of those preferences.

The idea of the sciences is to better understand things, not a means to declare absolute truth, which is more akin to a religion. Amir treats the Harman curve as if it's an absolute truth that must be complied to, or the headphones are faulty, and should be EQ'ed to match the curve. However, what the Harman research has lead us to understand is, that, just as some people prefer the greater mid-bass of the Expanse, and other people prefer the tuning of the Stealth, there is a great variation in what people prefer when it comes to the tuning of headphones. A great deal of this probably relates to the music people listen to, as there is considerable variation in the balance of frequencies between different genres, albums, and even tracks in an album, but also the considerable variation in the frequency sensitivity of peoples' ears.

Scientific experiments are designed in such a way as to help us answer questions about a specific thing in a manner that reduces the possibility of other factors, including the cognitive and other biases of the person conductive them, so that the results can be useful. If a reviewer, having seen measurements of a product, lets himself believe he (or she) is hearing a certain thing because they saw something in a measurement, for the reason I just stated, they are doing the exact opposite of science!

Very, very well stated. Often, people are misapplying and misunderstanding science while claiming their application of it as superior and absolute. That's what much of it really comes down to (and I say that as someone that works with scientific researchers on a week to week basis unrelated to audio). I will absolutely say that I totally get folks that get annoyed with people that only rely on subjective listening impressions and that shun measurements and more objective data points entirely. But that's not much of what I see here and elsewhere. It's push back on relying on "objective" data as source of absolute truth despite the limitations and heavy component of subjectivity in the hobby.

And the bolded part, we simply do not get into enough. We share so many strongly held impressions without often talking about what we listen to, which really is such a huge factor.

 

[SolarCetacean]

Harman curve is one point of the technical aspect only (speed, low frequencies clarity/details, distortion with busy music behaviour,... etc are not often represented at all in charts) ...
Maybe, it is easier and safety to adapt the sound with a poweramp EQ than to buy something totally out of a standard frequency response..

I disagree, I'd rather buy something non-standard but interesting and then EQ it to Harman if needed. It just doesn't work the other way starting from Harman and then trying to make something interesting. There are infinite combinations of FR balances, and only some of them sound good. And different FR features can complement each other, so that it only sounds good when both are there. So it's nearly impossible to craft a brand new sound via EQ that's also good because you make one change which could sound bad on its own, but can sound good if you also made another change at another set of frequencies. The state space is simply far too large. Plus EQ doesn't take into account all of the other factors that make a headphone sound the way that it does, so you can't EQ one headphone to sound exactly like another.

The Harman target was developed with EQ, so it's far easier to EQ to compliance there. Though the target is smoothed to 1/3 octave, so there's a lot that it actually doesn't say about the fine grained tuning. And recall that the preference portion of Harman refers to the bass and treble shelves. The midrange of Harman is based on the measured on-head response of flat speakers. The target is good if you want your headphones to have the tonality of flat speakers, but if you don't want that, it offers no suggestions. Dr Olive himself mentioned that his research doesn't take spatial effects into account, so the Harman target says nothing on that front. I think people really overstate what preference is encapsulated in the Harman target.

Basically, if you start with something interesting, then EQ to Harman, you get two tunings in that headphone. If you have a Harman headphone, you can't really get another tuning via EQ other than worse Harman. I would much rather get a headphone that offers a different perspective on the sound, e.g. interesting spatial properties, out of the box, and then if I wanted the flat speaker tonality, I EQ it to Harman.

 

[Ciggavelli]

I generally like the Harman curve, but I don't want all my headphones to sound the same. It completely circumvents the idea of personal preference. I do understand that for the sake of measurements, you need to have a one common go to, but that's not what this hobby is about. People like tubes for a reason, even though they are simply distortions. Only fitting everything into a specific curve leaves out the human side. Some people like X, while others like Y.

 

[DarginMahkum]

 

[mrspeakers]

all reviewers should be forced to review HP's without any measurements...give out the product to reviewers before the product goes to market without any measurements so reviewers will actually have to listen and render a judgment...measurements are a crutch other than perhaps power requirements so one knows what kind of amp is needed....I would love to see this amir guy forced to review products simply with his ears...he would still use price but would have to be more honest

The problem with Amir is, his biases are affected by what his measurements say, and consequently the positivity of his review. He has an entire forum that is supposedly based around "science", but approaches science and the meaning of it in a grossly misleading way.

The idea of the sciences is to better understand things, not a means to declare absolute truth, which is more akin to a religion. Amir treats the Harman curve as if it's an absolute truth that must be complied to, or the headphones are faulty, and should be EQ'ed to match the curve. However, what the Harman research has lead us to understand is, that, just as some people prefer the greater mid-bass of the Expanse, and other people prefer the tuning of the Stealth, there is a great variation in what people prefer when it comes to the tuning of headphones. A great deal of this probably relates to the music people listen to, as there is considerable variation in the balance of frequencies between different genres, albums, and even tracks in an album, but also the considerable variation in the frequency sensitivity of peoples' ears.

Scientific experiments are designed in such a way as to help us answer questions about a specific thing in a manner that reduces the possibility of other factors, including the cognitive and other biases of the person conductive them, so that the results can be useful. If a reviewer, having seen measurements of a product, lets himself believe he (or she) is hearing a certain thing because they saw something in a measurement, for the reason I just stated, they are doing the exact opposite of science!

It's always the same with every review; on ASR the measurement purists flame anything subjective, while there are also people here and other forums that reject or just don't care about measurements.

Me, I think all approaches are reasonable. Pure subjective, measured, and mixed are all fine. I do suggest that reviews that lean on measurements or are truly measurement-driven would benefit from creating subjective impressions first. That said, some measurement-driven reviews don't even include subjective impressions...

For the record, Amir listened to the headphone and shared his private impressions with me BEFORE he measured, and they were exactly the same he shared after he measured.

As to the Harman curve itself, as I've said many times I liken it to the Pirate's Code, more of a guideline really. I do think the approach headphones.com uses of showing the range of values to the curve is helpful because the Harman was an average of a wide range of results and it's helpful to understand that when visualizing the response curves to see where it falls in the range.

In the end, I think people who take an absolute view on this should just accept that what works for them may not work for others, and in this hobby it's a bit less about finding absolute truth and just enjoying yourself.

 

[ObjectiveSubjectivist]

Both sides are not free of sin.

 

[DarginMahkum]

Both sides are not free of sin.

Normally, these discussions about measurements, ASR, tubes etc. should have been long settled after being repeated by the same people hundreds of times, and a common sense should have taught us all that, we are free to choose what we read and believe as right and there's enough data to make everyone happy - that's if we're ready to show the slightest effort to find them. If you don't like ASR, well, just choose something else over it. If you don't like another subjective reviewer, well, just ignore that reviewer. Let them be the way they want to be.

But, apparently it is a very difficult thing to achieve.

 

[DMITRIY R]

As to the Harman curve itself, as I've said many times I liken it to the Pirate's Code, more of a guideline really. I do think the approach headphones.com uses of showing the range of values to the curve is helpful because the Harman was an average of a wide range of results and it's helpful to understand that when visualizing the response curves to see where it falls in the range.

Do you plan to release special versions of “Not Harman” headphones (For example, Stealth NH, E NH)? Considering how much people spend debating the correctness of this or that frequency response, this could be the right decision. Den Clark Audio products would immediately capture the supporters of both warring camps.