Hifiman planars for music production

[Goschie]

I've been researching a lot about what to look for in a planar magnetic when it comes to music production. What I have found is interesting. At the very least, the Hifiman planar magnetic headphones perform in a very similar way when it comes to how well you can eq the imperfections out of them. The graphs people have been generating all show that that they can be flattened to a reference curve without adding any instability/distortion. They all rate pretty well, and get close to the reference, but all have different characteristics that can be eliminated with eq'ing.

The fact they are planar magnetic makes their driver respond faster to change, at least this is according to other experts, and some people claim they all sound equally good and have a "fast" bass response. I have read that people believe the advantage of having a fast bass response is that they can hear low frequency elements on top of each other in space much easier, and this would be an advantage in music production, especially for layering. Bass drum layering is big in EDM, and it can be challenging fit it all together in the mix.

However, the biggest difference between some of these planars, even though they sound very similar just for the fact that they are planars, is sound stage and imaging. What I mean is that they do well in stimulating your ears into perceiving a similar depth and width you get from speakers, but they differ the most in this department. This begs an important question: Is there any point in focusing on other OBJECTIVE characteristics besides sound stage and imaging if these headphones can all be so close to each other with some simple eq'ing? It would seem that decent planar magnetics all behave so similarly, that geometry of the headphone is mostly what matters to give you speaker like qualities for better mixing.

 

[GaryH]

I'd suggest reading the excellent explanation below of what determines soundstage in headphones. Basically it comes down to frequency response, because that's what sound is - propagating pressure waves with a frequency and an amplitude. And in turn EQing obviously affects amplitude of the response at certain frequencies, as does the size, shape and angle of the headphone drivers/earcups via how much of the pinna of your ear the sound they emit interacts with. HifiMan headphones tend to have fairly large, angled pads so do well here. Good channel matching is also important for imaging and soundstage, and HiFiMan headphones tend to do well here too.
https://www.reddit.com/r/oratory1990/comments/i9pfx1/_/g1jf75o As for 'fast bass', this is an oxymoron really. By definition bass is low frequency, so must have a 'slow' oscillation cycle. And if a headphone can reproduce high frequencies up to the audible limit of 20 kHz (which pretty much every single one can), that's as fast as your ear can perceive, so as fast as you need your headphone's transducer to move. What I suspect people mean by 'fast bass' is determined by a combination of frequency response and low distortion at low frequencies, the latter being a forte of most planar magnetic drivers, including HifiMan's. See here for further explanation/debunking of some audiophile terms such as 'speed'. Actually I'd suggest just reading through any posts you feel are relevant in headphone engineer Oratory's Reddit 'blog', in particular 'How should a good headphone sound?' and 'Will two headphones sound the same if they have the same frequency response?'. In terms of buying advice, I'd just get one of the less expensive HifiMan's such as the HE4XX (which actually have a better stock sub-bass response than the Sundara) if you're going to EQ them anyway.

 

[Goschie]

I'd suggest reading the excellent explanation below of what determines soundstage in headphones. Basically it comes down to frequency response, because that's what sound is - propagating pressure waves with a frequency and an amplitude. And in turn EQing obviously affects amplitude of the response at certain frequencies, as does the size, shape and angle of the headphone drivers/earcups via how much of the pinna of your ear the sound they emit interacts with. HifiMan headphones tend to have fairly large, angled pads so do well here. Good channel matching is also important for imaging and soundstage, and HiFiMan headphones tend to do well here too.
https://www.reddit.com/r/oratory1990/comments/i9pfx1/_/g1jf75o As for 'fast bass', this is an oxymoron really. By definition bass is low frequency, so must have a 'slow' oscillation cycle. And if a headphone can reproduce high frequencies up to the audible limit of 20 kHz (which pretty much every single one can), that's as fast as your ear can perceive, so as fast as you need your headphone's transducer to move. What I suspect people mean by 'fast bass' is determined by a combination of frequency response and low distortion at low frequencies, the latter being a forte of most planar magnetic drivers, including HifiMan's. See here for further explanation/debunking of some audiophile terms such as 'speed'. Actually I'd suggest just reading through any posts you feel are relevant in headphone engineer Oratory's Reddit 'blog', in particular 'How should a good headphone sound?' and 'Will two headphones sound the same if they have the same frequency response?'. In terms of buying advice, I'd just get one of the less expensive HifiMan's such as the HE4XX (which actually have a better stock sub-bass response than the Sundara) if you're going to EQ them anyway.

I am just going off of what people are describing, as I am new to headphones, but as a physicist, some things they are saying seem intuitive. I think what they mean by "fast" is that if you pretend the driver is a simple harmonic oscillator, it is changing its mode "faster" than other headphones. Or another analogy is a rotating wheel. If you apply a torque, you accelerate the wheel to a new rpm. How quickly you can get the wheel to the new rpm is what they are describing as fast. Maybe I'm wrong, but that is how they make it sound.

 

[Jimbob54]

I'd suggest reading the excellent explanation below of what determines soundstage in headphones. Basically it comes down to frequency response, because that's what sound is - propagating pressure waves with a frequency and an amplitude. And in turn EQing obviously affects amplitude of the response at certain frequencies, as does the size, shape and angle of the headphone drivers/earcups via how much of the pinna of your ear the sound they emit interacts with. HifiMan headphones tend to have fairly large, angled pads so do well here. Good channel matching is also important for imaging and soundstage, and HiFiMan headphones tend to do well here too.
https://www.reddit.com/r/oratory1990/comments/i9pfx1/_/g1jf75o As for 'fast bass', this is an oxymoron really. By definition bass is low frequency, so must have a 'slow' oscillation cycle. And if a headphone can reproduce high frequencies up to the audible limit of 20 kHz (which pretty much every single one can), that's as fast as your ear can perceive, so as fast as you need your headphone's transducer to move. What I suspect people mean by 'fast bass' is determined by a combination of frequency response and low distortion at low frequencies, the latter being a forte of most planar magnetic drivers, including HifiMan's. See here for further explanation/debunking of some audiophile terms such as 'speed'. Actually I'd suggest just reading through any posts you feel are relevant in headphone engineer Oratory's Reddit 'blog', in particular 'How should a good headphone sound?' and 'Will two headphones sound the same if they have the same frequency response?'. In terms of buying advice, I'd just get one of the less expensive HifiMan's such as the HE4XX (which actually have a better stock sub-bass response than the Sundara) if you're going to EQ them anyway.

The only real downsides to the 4xx are for non US buyers having to import from Drop as its exclusive and that I find the pas felt a bit like cheap old carpet on my skin . I prefer softer velour or leather/ pleather for longer sessions .

 

[Goschie]

The only real downsides to the 4xx are for non US buyers having to import from Drop as its exclusive and that I find the pas felt a bit like cheap old carpet on my skin . I prefer softer velour or leather/ pleather for longer sessions .

What did you think of the imaging and soundstage? Do you think you could mix with it? Or is it too unnatural to accurately pick out sound location and character accurately?

 

[Jimbob54]

What did you think of the imaging and soundstage? Do you think you could mix with it? Or is it too unnatural to accurately pick out sound location and character accurately?

Never mixed anything!

I don’t perceive the 4xx as the best I have for imaging and soundstage. The Sennheiser HD800 takes that honour and it’s not a close race .

Difference is, even second hand 800s are £300+ more than new 4xx .

 

[solderdude]

I am just going off of what people are describing, as I am new to headphones, but as a physicist, some things they are saying seem intuitive. I think what they mean by "fast" is that if you pretend the driver is a simple harmonic oscillator, it is changing its mode "faster" than other headphones. Or another analogy is a rotating wheel. If you apply a torque, you accelerate the wheel to a new rpm. How quickly you can get the wheel to the new rpm is what they are describing as fast. Maybe I'm wrong, but that is how they make it sound.

Headphone drivers follow the applied voltage. There is no 'rotating wheel' effect. If there were you could not reproduce treble and bass at the same time.
Of course, there is a difference between large excursions (bass) and very small excursions (treble) on top of the larger excursions.
Having very low distortion in the bass (planars often have this) ensures the smaller wiggles (higher music frequencies) are not amplitude modulated by the bass because of non linear behavior.
For music production one often listens at higher than 'normal' SPL so planars work better.

Also look at some Audeze, yes heavier in weight, need more lower treble correction but are MUCH better behaved in the upper mids/treble when it comes to ringing.

Imaging is headphone and brain related. Well balanced (as well as L-R balance) and properly EQ'ed and angled drivers with a substantial distance from the ear helps.

 

[Goschie]

Headphone drivers follow the applied voltage. There is no 'rotating wheel' effect. If there were you could not reproduce treble and bass at the same time.
Of course, there is a difference between large excursions (bass) and very small excursions (treble) on top of the larger excursions.
Having very low distortion in the bass (planars often have this) ensures the smaller wiggles (higher music frequencies) are not amplitude modulated by the bass because of non linear behavior.
For music production one often listens at higher than 'normal' SPL so planars work better.

Also look at some Audeze, yes heavier in weight, need more lower treble correction but are MUCH better behaved in the upper mids/treble when it comes to ringing.

Imaging is headphone and brain related. Well balanced (as well as L-R balance) and properly EQ'ed and angled drivers with a substantial distance from the ear helps.

I dont really understand what you mean by "no 'rotating wheel' effect". Certainly, the fundamental has a maximum rate of hz/s change. The wheel was just an analogy.

 

[solderdude]

Or another analogy is a rotating wheel. If you apply a torque, you accelerate the wheel to a new rpm. How quickly you can get the wheel to the new rpm is what they are describing as fast.

The wheel was just an analogy.

Just mentioning that headphone drivers do not even remotely work anything like this so an analogy it fails. Drivers do like to vibrate at certain frequencies though. Hifiman headphones especially at higher frequencies which is clearly seen in CSD. Audeze is one of the exceptions. Also look at DCA headphones.
The question is about audibility of it though as the resonances are usually very narrow so chances are mostly they aren't exited and in music signals the frequencies don't stop suddenly but ring out and often take longer than the driver itself.

 

[Goschie]

Just mentioning that headphone drivers do not even remotely work anything like this so an analogy it fails. Drivers do like to vibrate at certain frequencies though. Hifiman headphones especially at higher frequencies which is clearly seen in CSD. Audeze is one of the exceptions. Also look at DCA headphones.
The question is about audibility of it though as the resonances are usually very narrow so chances are mostly they aren't exited and in music signals the frequencies don't stop suddenly but ring out and often take longer than the driver itself.

I dont mean to be rude, but I really dont think you understood the analogy. If you do not think that there is some limit to how fast the material in the speaker can change its fundamental frequency, then you are missing out on what I am trying to communicate. I'm not trying to be rude, its just I know that something like that must be involved, especially during an impulse or rapid change in pitch.

 

[Dealux]

I am just going off of what people are describing, as I am new to headphones, but as a physicist, some things they are saying seem intuitive. I think what they mean by "fast" is that if you pretend the driver is a simple harmonic oscillator, it is changing its mode "faster" than other headphones. Or another analogy is a rotating wheel. If you apply a torque, you accelerate the wheel to a new rpm. How quickly you can get the wheel to the new rpm is what they are describing as fast. Maybe I'm wrong, but that is how they make it sound.

Except that such descriptions are complete fabrications. Your brain has no way of knowing what it's actually hearing. The most we can say is that what we perceive is primarily frequency response (where distortion is negligible). New information or better information would also have to be conveyed by frequency response in some way but we do not know exactly how accurate measurements are vs how we actually perceive it.

Also, if slow drivers actually existed they'd have a very poor ability to produce HF content which, as you probably know, is not the case in virtually any headphone. The speaker itself is not limited in that sense but there are other factors that can limit performance such as dampening, driver enclosure design, etc.

I think what audiophiles perceive as "speed" is the relative absence of resonances in the treble (or simply put linearity) but in my experience this terms is greatly misused. Many people describe the HD800 as fast despite its fairly grainy and harsh presentation. In fact I've tested a bunch of cheaper headphones that sound "faster" than the HD800. The DT990 has smoother treble response overall and a more clear and detailed presentation (though this is a controversial claim) but audiophiles hate it (perhaps rightly so) because it has worse spikes in the treble than other headphones. So it's still a subjective metric. It has nothing to do with actual physical properties of the driver.

 

[JohnYang1997]

I dont mean to be rude, but I really dont think you understood the analogy. If you do not think that there is some limit to how fast the material in the speaker can change its fundamental frequency, then you are missing out on what I am trying to communicate. I'm not trying to be rude, its just I know that something like that must be involved, especially during an impulse or rapid change in pitch.

Speed = frequency. Fast = high frequency. Frequency response exists for just that. A material can have different resonant frequency. It will not change a fundamental frequency or frequency of the input signal.
Leading edge of time domain transient responses means high frequency. Plain and simple.

 

[Dealux]

Speed = frequency. Fast = high frequency. Frequency response exists for just that. A material can have different resonant frequency. It will not change a fundamental frequency or frequency of the input signal.
Leading edge of time domain transient responses means high frequency. Plain and simple.

Which planar do you like/recommend?

 

[JohnYang1997]

Which planar do you like/recommend?

Sundara 2020 is pretty good(one if the most accurate planars). I also like Ananda quite a lot (knowing some of the flaws in the frequency response). If you consider PM3 as planars, then PM3 too(one of the best closed back IMHO).

 

[Dealux]

Sundara 2020 is pretty good(one if the most accurate planars). I also like Ananda quite a lot (knowing some of the flaws in the frequency response). If you consider PM3 as planars, then PM3 too(one of the best closed back IMHO).

Would you take the Ananda over the Sundara? Also, what about Arya?

 

[LightninBoy]

However, the biggest difference between some of these planars, even though they sound very similar just for the fact that they are planars, is sound stage and imaging. What I mean is that they do well in stimulating your ears into perceiving a similar depth and width you get from speakers, but they differ the most in this department. This begs an important question: Is there any point in focusing on other OBJECTIVE characteristics besides sound stage and imaging if these headphones can all be so close to each other with some simple eq'ing? It would seem that decent planar magnetics all behave so similarly, that geometry of the headphone is mostly what matters to give you speaker like qualities for better mixing.

Interesting question. Sure, given low distortion, decent tracking to target curve, and ability to apply EQ - that would make FR variations between headphones less meaningful and so you could prioritize other performance aspects. However, soundstage is not an objective characteristic in headphones, in that we don't have a measurement that can predict it.

My other concern around focusing on sound stage would be that the better sound stage perceived in headphones such as the HD800 may actually be an artificial artifact that doesn't correlate with stereo speaker soundstage.

Oh yeah ... and don't mix with headphones.

 

[JohnYang1997]

Would you take the Ananda over the Sundara? Also, what about Arya?

If I have to choose one then Ananda. I have not heard Arya for a long time. But it sounded thin and has a peak around 13khz to my ears. No idea about newer version.

 

[fieldcar]

While I enjoy my monolith M570 and 4XX's atmospheric soundstage, I wouldn't use them for mixing & mastering work. They are far too peaky for mixing. If I didn't use my Jbl 306Pmkii's, I would be going with an etymotic ER2SE, ER4XR, or ER2XR/ER3XR(if you want a bit more bass). The name of the music-production game is minimal peaks and dips, which is why we use flat measuring studio monitors/speakers.

However, nothing beats an open back for auditioning. I can't remember who said it, but it went something like this. "Listen to your mixes everywhere, bluetooth speakers, cars, and all of the most popular headphones, as that's what your audience will be using".

So, just go with a sundara or ananda, but I'd get some etymotics on the side to seal the deal. In the USA, adrorama has them all on sale frequently.
https://slickdeals.net/newsearch.php?q=etymotic&searcharea=deals&searchin=first

 

[Goschie]

Interesting question. Sure, given low distortion, decent tracking to target curve, and ability to apply EQ - that would make FR variations between headphones less meaningful and so you could prioritize other performance aspects. However, soundstage is not an objective characteristic in headphones, in that we don't have a measurement that can predict it.

My other concern around focusing on sound stage would be that the better sound stage perceived in headphones such as the HD800 may actually be an artificial artifact that doesn't correlate with stereo speaker soundstage.

Oh yeah ... and don't mix with headphones.

Maybe not 10 years ago. I'd wager we definitely can today. But more importantly, I currently have no choice. lol

 

[Blake Klondike]

Thanks for this interesting discussion! The world of pro audio is just prone to nonsense as the consumer world, and I don’t see these kinds of discussions happening around that gear. In a world where you can choose between a $50,000 microphone or a $99 microphone, these kinds of discussion based in measurement and blind testing are very important, especially for people like me with No EE training!