Headphone Frequency Response vs Price

[Wombat]

Paper: https://asa.scitation.org/doi/full/10.1121/1.4984044

 

[Grave]

I have not actually read this but I have heard about the results. Is anyone surprised? Many expensive headphones have insane frequency responses.

 

[Sal1950]

How you gonna sell product if yours doesn't sound any different?
Audiophool imaginary differences can only go so far.
Some guys want to hear real sparkle on the top or boom in the bass.

 

[Timbo2]

"However, the variance in low-frequency response seems to decrease with increasing price, indicating an improved bass response measurement consistency across headphones in the higher price range. It is however unclear whether this improved consistency with a higher retail price is the result of better headphones or better repeatability of measurements with more expensive models."

I can totally believe the report. Consistent with my personal experiences it seems the bass response seems to be worse in cheap headphones. My guess is proper bass reproduction adds more to unit costs than engineering the high end. But I'm not disagreeing with any of the conclusions that they found.

 

[solderdude]

The title of the article says it all and totally agree:

No correlation between headphone frequency response and retail price

 

[openbaffle3278]

One of my favorite analyses on the subject:

"The State of the Flagships: An Under-Engineered Mess"

Cool article, I'd like the get the data. I like the idea of doing FPCA/Functional Data Analysis to look at where deviations from target response is. Would be interesting to correlate perceived sound quality with the FPCs.

Thanks for sharing.

 

[audiobill]

How you gonna sell product if yours doesn't sound any different?
Audiophool imaginary differences can only go so far.
Some guys want to hear real sparkle on the top or boom in the bass.

I would think the more serious user would prefer balanced sound. Boosted bass and treble would be more for kids I think. Price does not dictate quality.

 

[svart-hvitt]

Clearly, headphones are still in their wild-west stages of their evolution. Absence of science and standards explain variation.

 

[mitchco]

Some good science summarised here that points to 14 AES papers on the subject area.

 

[svart-hvitt]

Some good science summarised here that points to 14 AES papers on the subject area.

I didn’t say there is no science in headphone area. What I wrote, is partly an Olive quote. The full quote goes like this:

«A recent study by Jeroen Breebaart measured 283 headphones and found there was no correlation between their retail price and measured frequency response: the best objective indicator of how good the sound [20]. Clearly headphones are still in their wild-west stages of their evolution. Without sound guidance from science and standards, their performance and sound are highly variable. Fortunately, this is rapidly changing».
Source: https://e265b8fd1ff9a586c366-1acdfa...-for-Choosing-Loudspeakers-and-Headphones.pdf

Where I, cynically or realistically, differ from Olive is on the notion that «this is rapidly changing». There is still no evidence of convergence, standardization. In fact, headphones producers have little to win by standardizing their products. Remember, standardization means commoditization, which means lower price and profit.

If someone can come up with evidence that the article in the OP is already dated, that standards are about to emerge, then provide that evidence.

Having said that, let’s not forget that you can have science in a field and still see products coming out as if science were not present. It’s naive to think that science will only be used in a benevolent way. However, science can also be used in a malignant way. @restorer-john has written elsewhere on this forum that producers of amplifiers used to make circuits with in-built curves, and the flatter curves were only present in the more expensive ones. This is a strategy that works, that is science based, that takes advantage of non-informed users to maximize profits through a jungle of products.

Headphones are still wild-west and there’s no sheriff in headphone town.

 

[solderdude]

I think there are a lot of serious brands all trying hardest to make a headphone (and drivers) as well as they can.
You just run in so much variables that it is impossible to build a 'perfect' headphone that is universally liked by everyone.
That is regardless of any standards..

I am quite certain that if a single 'standard' (let's assume O-W would be a standard) would exist and all would use the exact same HATS and compensations that all the products based on that would
A: still sound different and
B: Some won't like them and would have to EQ them.
C: Some would be begging for 'better' standards

Just buy one of the damn things that sounds closest to what one prefers (as one does with speakers as well) and then EQ it to how you want it.
A lot of people are not afraid to use (inherently flawed) room correction DSP's but refuse to use EQ for headphones.

Standards also bring limitations.

 

[svart-hvitt]

I think there are a lot of serious brands all trying hardest to make a headphone (and drivers) as well as they can.
You just run in so much variables that it is impossible to build a 'perfect' headphone that is universally liked by everyone.
That is regardless of any standards..

I am quite certain that if a single 'standard' (let's assume O-W would be a standard) would exist and all would use the exact same HATS and compensations that all the products based on that would
A: still sound different and
B: Some won't like them and would have to EQ them.
C: Some would be begging for 'better' standards

Just buy one of the damn things that sounds closest to what one prefers (as one does with speakers as well) and then EQ it to how you want it.
A lot of people are not afraid to use (inherently flawed) room correction DSP's but refuse to use EQ for headphones.

Standards also bring limitations.

As science uncovers truth, this truth becomes a baseline, a reference, a benchmark. Is it possible to combine a position where one is against - or skeptical - towards standards, while at the same time claiming one is for science? Isn’t it a contradiction?

Even if the headphone problem seems difficult, should one give up on standards?

Standards and science are not exactly the same thing, but they are interlinked. So I remain extremely skeptical towards those who claim standards are a bad thing, well aware of the fact that there are bad standards and good standards.

 

[mitchco]

@svart-hvitt now see the context, thanks. Agree it is still the Wild West, even though there is good research for both headphones and loudspeakers as to what sounds good. Loudspeaker manufacturers can acquire the research that Olive and Toole did to ensure they produce a decent sounding loudspeaker. Not sure what the uptake is, but I suspect it is the same as headphones. I have had only one set of loudspeakers come through my listening room that were not overly bright right out of the box... Most have been so bright that without reducing the HF output, I can't listen to them for more than five minutes.

 

[solderdude]

Standards aren't a bad thing.
One could say the sound output of a driver should follow the waveform of the applied signal within a certain percentage (or dB) in aspects such as amplitude or time.
The question then remains HOW that is measured and how many percent (or dB) deviation is allowed on that particular test rig and call it a day.
That still does not mean all people will like it even when scientists say a headphone should sound a certain way.
How 'tight' limits are set and upheld by manufacturers could mean one could attach a 'sign of approval it is meeting certain criteria' still begs the question if these were sold better.

The AQ Nighthawk is one mans dream headphone and anothers dark, fat, bassy, ugly sounding monstrosity, One mans analytic HD800 is another mans cold, bass shy and sharp monstrosity. One mans 'accurate' is another mans 'boring'. How does one capture this in a standard ?

Surely it is much easier to apply standards to electrical equipment than it is to acoustic equipment because acoustics are very complex (even in a headphone) and dependent on multiple factors such as Pinna size and shape and seal to name just a few things.
Must there also be standards for comfort ? Minimum pad size ? minimal room for the ears ? driver angle ? contact pressure ? weight ? pad materials ? Impedance ?
Or should it only be frequency response measured in a certain way ?

I am all for standards but chances are some standards are more limitations than anything else.

 

[svart-hvitt]

Standards aren't a bad thing.
One could say the sound output of a driver should follow the waveform of the applied signal within a certain percentage (or dB) in aspects such as amplitude or time.
The question then remains HOW that is measured and how many percent (or dB) deviation is allowed on that particular test rig and call it a day.
That still does not mean all people will like it.

Surely it is much easier to apply standards to electrical equipment than it is to acoustic equipment because acoustics are very complex (even in a headphone) and dependent on multiple factors such as Pinna size and shape and seal to name just a few things.
Must there also be standards for comfort ? Minimum pad size ? minimal room for the ears ? driver angle ? contact pressure ? weight ? pad materials ? Impedance ?
Or should it only be frequency response measured in a certain way ?

I am all for standards but chances are some standards are more limitations than anything else.

The implication of what you wrote is that you think the OP article Breebaart (2017) is irrelevant (because they didn’t control for the jungle of factors you mentioned).

I think the Breebaart (2017) article is just what the doctor ordered: It’s documentation that we’re in the middle of a headphone jungle and we need a guide to help us out of it. Science is that guide. And the natural sciences (almost) always lead us to standards.

 

[solderdude]

The research is fine, any research is fine.
The question is what it will lead to ?
Will that research pave the path or even lead to correlation between headphone frequency response and retail price ?
Will headphones that apply to a certain standard(s) all really sound closely the same ?
Can proposed tolerances actually be met ?
Is +/- 1dB or +/-3dB over the considered audible range actually possible to make ?
Would a headphone have to meet 'hires logo' types of frequency response ?

 

[svart-hvitt]

The research is fine, any research is fine.
The question is what it will lead to ?
Will that research pave the path or even lead to correlation between headphone frequency response and retail price ?
Will headphones that apply to a certain standard(s) all really sound closely the same ?
Can proposed tolerances actually be met ?
Is +/- 1dB or +/-3dB over the considered audible range actually possible to make ?
Would a headphone have to meet 'hires logo' types of frequency response ?

Yes, the tolerances, error margins you mention are examples of the type of convergence I expect to see in presence of science.

It could be that headphones should have a couple of switches (two or three switches will yield many different choices for tailor-made sound) for adapting to the individual’s anatomy.

To me, it seems like Breebaart (2017) suggests we’re deep in the jungle still. But I agree with Olive, Toole etc. that standardization is the way out of the jungle.

——- EDIT ——-

Such switches I referred to above have been implemented in speakers for decades already. Which suggests it’s a good idea for headphones as well.

 

[openbaffle3278]

So in the overview PPT, they mention the JBL Everest 700 Elite as headphone, which matches their curve. Being an active headphone, I would assume they are taking advantage of heavy EQ. Ive never heard them though. FR of the 700 Elite is very close to the ideal curve.

 

[solderdude]

I have heard the 700 Elite on a few occasions. I would not like to see that becoming a 'standard'.
On top of that... threre is MUCH more to fidelity than tonal balance.

The switches on speakers are easy to make. In passive speakers they are just attenuators of a tweeter for instance. In active speakers it is much easier to do.
There are a few BT headphones already that have such 'tuning options' by usage of apps.
The PXC550 to name just one. They did not address the treble peak it has though but can tune the overall balance in various ways.
I don't see this happening in the higher end though.
Consumers will want their amplifiers in the chain.

I am sure though that even when there is a standard for a certain fidelity there still won't be any relation between price and frequency response.

The only way to achieve 'perfect' frequency response is active EQ (DSP) when you have that tonal balance preferences are easy.
There is already such things on the market. It's what the EARS are for. Too bad there is so much variance in the product.
Even there.. you can't polish a turd. You need something as close as possible to perfect and correct that.

I have various headphones that I can EQ tonally 'flat' yet they all still will sound different and also here there is no relation between price and SQ.

 

[openbaffle3278]

On top of that... threre is MUCH more to fidelity than tonal balance.

That is why I like the multi-point "score-card" from the measurements that was provided in the PPT I linked to above:

• Bass linearity of +-5 dB from 20hz to 100hz.
• 100dB distortion should not exceed 0.8% beyond the sub-bass frequencies.
• 100dB distortion should not exceed 1% at 30hz.
• Frequency response curve should be very smooth with any resonances being very minor -- no major dips.
• Very small to absolutely no dip at 70hz-150hz.
• Air-level treble should be no more than -15dB relative to the mid-range.
• No "wiggle" in the impedance graph.
• Nearly perfect channel balance.
• The headphone is open or semi-open.

The measurements of the 700 are here. BTW, I was looking at slide 128 in this presentation: "Breaking the Circle of Confusion" where the "creation of the art" is done on a JBL Pro M2 and presented on a 700 Elite.