Friday, December 5, 2008

Neutral vs. Natural: A Thought.

The question of finding a 'neutral/natural' sounding IEM* usually comes up in many forums in regular interval (*this probably applies to all headphone as well). While the question might seems very easy to answer, I often find it to be anything but simple. Reason? I just can't agree that 'neutral' and 'natural' being used interchangeably to describe the same kind of sound - well, at least not in the IEM world. So what do 'neutral' and 'natural' mean to me and why do such similar terms affect your path on finding the right sound? These all go back to the root of headphone design and the fundamental of what makes headphone so different from loud speaker that most of us are more familiar with.

To understand this topic, you need to understand some essential differences between loud speaker and headphone. A reputable loud speaker manufacturer usually like to advertise their product for having a flat frequency response (FR). The important of having a flat FR is that it means the speaker won't add or subtract sound from the amplifier's signal. It will just sound as what it is intended to sound like. In the case of a non-flat FR, the speaker is said to 'color' the sound by increasing or decreasing the sound pressure level on certain region (i.e. extra bass or treble that is not supposed to be there). People who are into Hi-Fi usually like to avoid coloration as much as possible. After all, they want high fidelity, not high fiddle-ity. Here is where neutral meets natural: The speaker is said to be neutral (flat) sounding when there isn't any coloration (non-flat FR), which means it is also most truthful to the recording thus it is closest to what sounding natural. In the headphone world, however, neutral does not equal to natural. When you use speaker, the sound travels through the room, interacting with the wall, furnitures, your body, and your outer ears before reaching your eardrum. Every interaction during the process affects the final FR you hear, meaning it is not likely that the FR measured in your ear canal will be as flat as the FR of the speaker . If the FR is not longer flat, why do we called it natural sounding? That's because interaction is part of the natural process of how sound travels through space. When using headphone (especially IEM), many of those interaction do not take place. If headphone has a flat FR, you will probably hear a flat FR, which is not normally what you will hear when you are listening to speaker. To compensate, headphone manufacturers will tuned the FR of their headphone so it will better resemble the FR after the speaker's sound travels through the room. A common example will be the extra bass response most headphones have. lower bass note are often felt by the skin and heard by the ears at the same time, so the impression of hearing low bass note is a mixture of auditorial and tactile sensation. Since headphone transducer are much smaller (and move much less air), headphone user often 'not feeling' enough bass when the FR is flat. To compensate, headphone manufactures give their headphone an extra bass response so more air is moved and the user can better 'feel' the bass note as they would like when they are using speaker . This is why coloration is more favorable than being neutral (flat FR), thus coloration is what sounding more natural to the users, not flat, neutral FR.

So what kind of FR actually sounds more natural to an IEM user? There isn't a clear, straight answer I can give you. To add to the complex issue of what kind of tuning results in a more natural sound, we still have to consider the fact that each of us hears thing differently - we all have our very own FR curve due to aging and hearing damage accumulated over time. When you add the difference in musical taste to the big mix, there is impossible to tell what would sound most natural to you. However, there is one thing we can be sure of: Neutral and natural are not created equally in the IEM world.

16 comments:

Steven Wayne said...

"tentacle sensation" - really?

Or did you mean tactile?

Tai / ClieOS said...

Indeed.

Anonymous said...

Fabulous !!

Lauri said...

If you have aged and have "funny" dips in your hearing, they will be there as much if you listen to speakers as if you listen to headphones. This is not something that can really be compensated without the compensation sounding unnatural since this is the way you hear sounds in the real world.

Room acoustics and all that is another thing. But frankly I would prefer headphones to leave these compensations out as well, letting the acoustics of the recording come through, whether they reflect the recording environment or post-processing, like reverbs or delays. This is the reason why the more reasonable audiophiles consider acoustic improvement of the room more important than the equipment they use to play the sound: they want to take the room acoustics away so that the acoustics of the source material can shine through. I do not want my headphones to add such distorting acoustics.

But there is one thing a headphone, especially a canalphone, should do to colour the sound: that is to take into account the way your head and particularly your pinnae modify the sound before it enters the ear canal. These actually exert a tremendous influence, and this is also something that is very individual. Each one of us has a brain that has learned the specific peaks and troughs caused in the FR becasue of our heads and pinnae, and these are the things that allow us to locate sounds in 3D space rather than just on a plane - i.e. they give us a sensation of a sound coming from above vs. below or from ahead vs. behind. Of course, people's ears do share some similarities, but no two people have the same exact ears - which leads to each person having slightly different dips and troughs in the FR of sounds reaching their ear canals. In other words, there is no one correct way to colour the sound of canalphones: some will be a better mach for you personally than others, better reflecting the way your pinnae and head affect the sounds you hear out there in the real world.

Unknown said...

Honestly this is a poor article as it doesn't have any grounds to back itself with. Headphones can be made to sound neutral (it's called Headphone-related-transfer) , research by Moller and H have found the compensations needed to make it so. The only true mystery is the missing 6db effect (we perceive bass levels lower on headphones than speakers despite the same sound pressure response), boosting all of the bass isn't a good way to approach this phenomenon as you mess with the mids and tone in the process. A boost below 100hz may be a good way to do it, which will mean a truly dedicated woofer that just hasn't been done in any IEM. Or increase THD levels in the lower regions as that increases the perception of more bass without messing with the sound pressure.

The use of the term "natural" in this article just sound like an excuse for a subjective sound preference that goes against the truly natural, real nature of a recording. If one likes more bass that's fine, but "natural" is a problematic way to describe it...

Unknown said...

References
http://rinchoi.blogspot.com/2010/05/headphone-equalization.html

http://www.ingentaconnect.com/content/dav/aaua/2008/00000094/00000001/art00014

F. V├Âlk, and H. Fastl, "Locating the Missing 6 dB by Loudness Calibration of Binaural Synthesis," presented at the 131st Audio Eng. Soc. Convention, preprint 8488 (October 2011).

Tai / ClieOS said...

Actually, it is exactly the problem with HRTF (which is averaged estimation, not tailored to individual listener), loudness curve (which depends a lot on the listner's preferred listening volume) as well as the individual difference between listener's hearing ability (golden ears vs. hearing damage) that are what this article is trying to point out. You can of course trying to 'average' everything out and build an ideal FR curve for a headphone, as Etymotic has done so in their ER4 series over 20 years ago using research data - but you also can not deny not everyone is your 'average Joe' and unlikely everyone will agree that ER4 is the most neutral and natural (or should I say 'accurate') sounding IEM they have heard. Measurement is great tool for setting standard, but that's for measuring equipments and not for human ears.

As I have said in the article, it is just a thought - not scientific research on published journal by a long shot. It is intended to warn many forum novices that many seasoned audiophiles (me included) like to use (and mix) these terms to their own liking, because they only listen with their ears and not by any scientific standard or tools. As you have so adequately pointed out - it has been long that we (the audiophile community, as well as different audio companies) use subjective sound preference to define these terms, and it has somehow became an acceptable habit. I am more than welcome if you like to take it to the big companies that all like us to believe they have the most natural and neutral sounding headphone every made using some researches, but I am just one guy expressing his opinion on how these terms might not be the most accurate things to go by in the current world we are living in.

Unknown said...

The Moller study studied the HRTF of many individuals and created an average, it's no guess or estimation. It's a recent 2008 study that actually has a very similar curve to a previous one by Killion
http://asadl.org/jasa/resource/1/jasman/v81/iS1/pS75_s4?bypassSSO=1

So since you're going to have fixed sound when tuning an IEM, why won't you want to have the most universal and objective form to go by? It's not perfect, but it's the closest we can get unless it's really tailored for that individual, that's just not possible in a universal headphone. It's the reason why the HE90, 009, high end Senns,etc. take the research in mind.

Loudness curve is a factor to consider, but the peaks and valley are still there, it won't make a highly colored IEM, neutral. It takes a very quiet listening level for considerable change but even then you're dealing with compressed dynamic range.

Ear perception will have to do more with range will it not? With this in mind, it won't be a huge factor. Also, even untrained listeners know neutral is better in controlled settings.
http://seanolive.blogspot.com/2012/05/more-evidence-that-kids-even-japanese.html

You don't have to guess of what's accurate or form personal opinions. The true answer for that will be to compare the headphone's response to a reference, research that has been done by Etymotic. Why are measurements tailored for a human ear's average resonants, with a simulator head and human factors in mind, if it isn't for human ears? We can't listen to each other's listening, this is the next best thing.

Natural usually means boosted bass as it's main factor. Most people don't know the true nature of the recording so just make broad guesses and there many biases that occur when purchasing a product.

Tai / ClieOS said...

I don't disagree with most of your points. But the truth is, even among big brands, they can't fully agree on what the best FR curve for a headphone is. They understand the studies well enough and actually making a near 'perfect' curve isn't too difficult either (as demonstrated by Etymotic 20 years ago). But if we already know what the supposed perfect curve is, why isn't every company making their headphone sounds just like that? What isn't every high-end headphone sounds the same way as it is supposed to sound like? Because ultimately the guys who are designing these headphone and the guys who are marketing these headphone realize they can't sell measurement and plotted curve to listener. They need to sell the sound, and as accurate as they can measure and plot it, it still need to be attractive to the listeners' ears. Taking Etymotic MC series as an example - highly accurate to the same Etymotic reference curve as used in their ER4 series, yet it sounds uninspiring to many ER4 owners who have so used to and loved that FR curve. I don't think we can disagree that FR curve isn't everything that makes up a headphone.

Despite our ability to measure everything so accurately, opinion is still opinion. People who are discussing headphone are unlikely going to be scientific about it and start showing off graph and research to prove their headphone is superior. At the end, there will be (and have found to be) people that dislike HD800 and ER4S because they don't sound natural to them (and you can find these discussion on Head-fi.org easily), and some of them are well respected members in the community. The danger is when novices are reading their impression/review without knowing their preference and begin to take those definition of neutral/natural as their own and imagine there is somehow an universal standard that everyone has already agreed upon. Because there is none, at least not in the real world of headphone makers and listers. If there is one, it is still in the ivory tower of science and hasn't made it way to most people out there yet.

Unknown said...

Unfortunately some brands still take a free-field equalization in mind, an approach that has been debunked by Killion as well I believe, just a matter of using outdated data.

Not all companies are created equally, it's those that want the best peformance that keep the most established, recent research in mind. There is no perfection, my problem is with lack of well established approach. HD800, SR009, K701, etc. the're not perfect to the curve, but close enough to be taken seriously in the realm of performance. It may seem as easy as just following a curve but it's not. Even with the Moller curve, the missing 6db phenomenon has to be taken in mind. Bose surprisingly took a somewhat clever approach to compensate for this without boosting bass in the AE2.
http://rinchoi.blogspot.com/2012/06/bose-ae2-part3-in-depth-analysis.html

Neutral signatures still have their notable differences, thus everything won't be sounding the same but still a lot closer than all these so called overpriced IEMs that are highly colored, yet having a high price-tag.

Also, most companies tune by ear and don't have the resources to fine-tune frequencies as a company like Sennheisser. Also, some want to color the sound in favor to make a certain approach some sort of trend, but it's a consumer approach, not one that tries to maximize performance. Thus, of course they can't sell measurements as well as established research, only high-end professional oriented products do.

That's the problem with the MC series, it's a handicapped HF5/ER4. Taking those two out of the picture, it's truly the best, most accurate product in it's class. Thing is, compared to an HF5, the dynamic driver has less transient speed, treble bandwidth and midrange transparency (due to bigger midbass). It's not a bad sounding IEM at all though, the helmholtz resonator is probably the most innovative application on a dynamic as it really shows finesse, lack of resonance. Preference is one thing, accuracy/performance is another.

That latter paragraph furthers my point, sure that's obvious, people will think something colored sounds better to them (they can't claim more neutral without ground though), but that's their own personal matter. If trying to establish a universal hierachy to which any reader can go by, well established rules have to be made, accuracy is the first step.

Tai / ClieOS said...

Even Sennheiser is not tuning their IEM using the ideal curve - just listen to their top-end IE and MX series. I was amazed by technology and effort Eytmotic put into the MC series when I read its patent application, but hearing it makes me think that it has been overly compensated (or should I said overly dampened) to force a FR curve that isn't supposed to be on that transducer. You can call it my personal bias and I won't deny it, but MC series is by far the worst sounding IEM Etymotic ever made even compared to the dated ER6i, left alone calling it best-in-class. On the other hand, Phonak, another hearing aid company, has done an excellent job on introducing the similar 'ideal curve' on their first product, the PFE1xx series. But even they have to change their tactic when people start asking for more bass with the PFE0xx Perfect Bass and PFE232. As I have said, being ideal doesn't equal to making money, or having happy customer in the matter. In the end, the majority of music listeners care very little about accuracy, or what it means to be neutral/natural. They just want something that sounds better than their previous earphone. It would be too easy to search for the FR curve published on HeadRoom and tell people which is the best simply by looking at graphs. But that hasn't really work either. They help, but you can't look at graph and tell whether you'll like it or not, even if that IEM is claimed to be highly accurate.

Accuracy, at the end, is as much an marketing word for headphone makers as it is with 'neutral' or 'natural'. It is the same word that has been over-used on headphone boxes and ads. If you want people to use accuracy as a standard for evaluation, you not only have to make headphone makers all abide to one single definition of accuracy, but you'll also have to educate most readers to understand what accuracy is about and how it plays into their decision making. Would be really nice if you can do it, but we all know that is next to impossible when there are already too many novices wanting a quick answer to the 'which headphone is the best?' question and plenty of companies will love to sell you their 'most accurate' headphone ever without having the claim certified by ISO.

Unknown said...

Like Sony (XBA series being completely different from the EX510-1000 series), Sennheisser has different development teams. The IE8 is actually quite diffuse field overall except for it's bloated midbass. Have no info on the MX, but I wouldn't be surprised if it's also linear like the PK1/2, except with the typical issues of an earbud.

The MC series is actually not dampened much, though not as undampened as the ER4S. How can you assume what's "right" for the driver. Dynamics aren't as great as BAs in the IEM world in terms of transparency, not a fault of the tuning, which got as close as it can for the moment. It's still their elementary approach, the EtyKids seems like a good step-up. Haven't heard the ER6i, but that is their less accurate IEM from what they tested. ER6 on the other hand is a monster, too bad that was discontinued (dated IEM, still outstanding performance).

Best in class as in the most accurate, sure is IME. As long as the HF series can't be attained for the same price. I'll consider the PFEs in a higher price bracket and yes, they're quite accurate as well and a bit better, my only problem is their higher treble resonance.

You can educate listeners but biases and conformity won't allow everyone to like the best performing product, yes that's true.

It's not silly marketing if it's been proven by 3rd parties over and over. Some companies do their homework, some just lie or mask the facts.

Tai / ClieOS said...

You will be surprised some of these IEM you are referring are from OEM and not actually developed by these big brands at all. The are too many assumption that the Chinese only make clones of the real stuff, but not many know that some of them are actually the behind-the-scene creator, almost like Foxconn to iPhone.

Ever wonder why some of these companies that never open a factory in China can keep pumping out 'made-in-China' products? Unless you are counting OEM as 'different development teams', I really won't put too much stock into the idea of actual R&D work being done by the companies themselves, including some of the high-end models as well. Of course sometime they do indeed developed in-house, but a lot of the time you will also see these companies hunting down new products on the Spring Hong Kong Electronic Fair as well - though that's story for another day, I just can't stop but wondering who is actually doing the R&D and who is just rebranding stuff from the OEM.

Unknown said...

Perhaps. If it's the case of the IE8, I wouldn't be surprised. I do know for a fact that most IEMs are OEMs from China, I try to find the connection whenever I can as manufacturers try to mask this. Perhaps Sennheisser does tell them to tune the sound in a certain way or maybe they don't any of the work at all as the resulting sound coincides with their research.

I know that's not the case for Sony as they do in fact have two R&D teams and in both cases the drivers are made in-house. For Sennheiser's higher-end stuff I believe they design and engineer in-house, acoustics done in Ireland and put together/assemble in China. For the IE8 and lower end Senns, it could just be all done in China and they bought rights or something of the-like. Wouldn't surprised me but the end-result shows compared to a HD25 or HD800, it's really not a top-tier IEM by any means (though it still has it's inevitable fans).

Either way point being if manufacturers want to be taken seriously they need to their research. Sennheiser and Beyerdynamic have the potential to release something worthy of the high-end due to their knowledge, just hasn't been done. Etymotics are the Stax of the IEM world, bet people would respect them more if they had a ridiculous price-tag like the K3003.

Tai / ClieOS said...

Indeed.

Just adding a bit more info on Sony - their Japanese team has been the driving force of the company's headphone business for as long as I can remember (from the early EX to the current XBA series), but I am glad to say their European team is catching up as well, mainly due to the Sony's recent dedication to the development of higher end smartphone. It might seen strange at first, but smartphone users are rapidly becoming the most important group of music listeners. That's why we have seen many higher end IEM now come with mic+remote. Without a doubt we will see more.

Anyway, there will be an interesting article coming from a contributing author (as an industry insider) on either the end of August or early September that is related to R&D of an audiophile oriented product. It will be something you don't usually read on any website, I can assure of that, and we'll have a glimpse on how the whole process works.

Unknown said...

I meant to say that Senn and Beyer haven't released a top-tier in the IEM world despite the potential due to headphone knowledge.

I am one of those users myself, smartphones are getting better at being sources as well. Before the iPhones were the only good sounding smartphones, now Android based smartphones have good sounding ones in the Galaxy 3 and Galaxy Note and a few others from other companies.

Looking forward to it

Post a Comment

Amazon's Deal

Disclaimer: All trademarks and logos in the website belong to their respective owners. Beside getting free review samples, I don't work for or get paid by anyone to write anything on this website, or anywhere else in that matter. Also, free review samples are never sold for any financial gain. I do buy gears and review them, but for simplicity you (the reader) should always assume what I review is free sample in nature (and thus comes with all the bias). The website does have Google Ads and Amazon Associates enabled (which I have no direct control over their content) - though I don't write review for a living, nor does the ads generates enough money to cover my breakfast (in fact, not even one breakfast per week). Listening to music and playing with audio gears are purely hobby for me. In short, I am just an audiophiles who happens to have his own blog. Not a journalist who happens to be an audiophile. Oh, and excuse my writing as I am not a native English speaker and can't afford a proofreader. Also, just because I don't write in a negative tone doesn't mean I don't write down the negative aspect of a gear. Please read them carefully.

Important: All postings are my own personal opinion only and should not be treated as absolute truth. I do get things wrong just like everyone else. Always do your own research!

Lastest Posts

Latest Comments

 
Copyright 2008-2014 In Ear Matters. Powered by Blogger Blogger Templates create by Deluxe Templates. WP by Masterplan