Have you ever wondered why there are so many sarcastic jokes in this Hi-Fi community? The 'giant killer' nwavguy’s combo? The 'funny' audio science review (ASR) ranking? Science loses its power in the audio world? Scientists are considered to be the evilest group in the HiFi community? Some of us even claim the best headphones were made 30 years ago? Do you ever get a feeling that you live in The Matrix of audio?

Yes, you come to the right place and we are now pretty much talking. You can also send me PM, I am willing to offer you pills.

"You take the blue pill—the story ends, you wake up in your bed and believe whatever you want to believe. You take the red pill—you stay in Wonderland, and I show you how deep the rabbit hole goes. Remember: all I'm offering is the truth. Nothing more."

Premier:
⦿ Spoiler: truth sometimes can be very offensive. This is a collection of advice I have given other audiophiles and is completely my opinion. Please do not take any of this the wrong way, as some of this can tend to sound obnoxious or self-centered, but I can tell you this based on my own experience of buying and reselling tons of high-end audio stuff you might never see. But still, this is just my opinion.

⦿ I have/had uber-expensive headphone and speaker rigs (owning >100K USD worth headphones/speakers at this moment), but not by any means am I a rich people. I drive a $2K Toyota. I still use the crappy camera on my vintage iPhone 6. You can probably see from here how much I am into this hobby.

⦿ Over 90% of my music collections are classical music. If you focus on different genres, we might end up with drastically different conclusions for the same headphones/speaker.

⦿ Recordings from un-amped instruments and vocals (i.e. classical music) benefit much more from better transducers. Personally, I will stop at $500 (used) range headphones if I only listen to pop music (e.g. soundtrack, hard rock, EDM, metals, etc). Buying high-end playback devices for pop music is philosophically wrong because you are mostly listening to synthesized/fake notes generated from computers. However, it might be the reason that people listening to pop music like to spend a crap ton of money on upstream electronics. Because they don’t have a true reference on what real instruments sound like, so they’d like to spend money as long as there is a difference (no matter good or bad).


Embodiments:

0. You can't tell the difference (in a blind test) between two devices if they have identical measured frequency responses (measured with a load like a headphone). However, the frequency response of two devices might not necessarily be the same if you can not tell the difference in a blind test.

0.5. The frequency response is a low dimensional projection from a high dimensional space based on Fourier theory. In English, that means any perceivable changes in the time domain for example sound stage/speed/detail/timbre should alter the frequency response. However, this doesn't mean you can exactly tell the sound of a system from its frequency response. You can only guess how it sounds distantly from the graph. This is because interpreting the behavior of frequency responses is simply too complicated and not very linear. Plus, the definition of speed/soundstage/timbre, etc varies from person to person.

1. Maybe more than 99% of audiophiles never experienced any true blind test session. Some of them just don't want to. But more importantly, a loudness level compensated double-blind test is impossible if you don't have proper tools. This is because level matching can never be done by anyone's ear. True story, I let one of my friends adjust the loudness level from -40.1dB to -40dB (to match another DAC) using my measurement equipment. It literally took him 5 minutes to adjust the last tiny bits of the volume knob. Just think about it, how come this can be done by human ears.

2. The only conclusions that make sense are those based on blind tests with matched loudness levels. Casual listening without strictly level matched AB comparisons will only end up with a crap load of BS. Humans perceive sound differently at different sound levels. That means for the same system, the soundstage/speed/details/transient, etc are not the same at different loudness levels. Please spend 2 minutes and check the concept of equal loudness contour: https://en.wikipedia.org/wiki/Equal-loudness_contour

3. I am not a cable believer. Don't get me wrong. Cables can make a difference. But most of those hi-end cables won't make your signal cleaner. In fact, I am not even a believer in those fancy flamboyant super high-end upstream electronics (DACs/AMPS/Players/Turntables, etc). IMHO, good upstream electronics would only contribute to less than 1% (can be as small as inaudible) of the sound quality. Moreover, an integrated chip-set as small as your fingernail from any of those big tech companies, like TI and ADI, can produce literally identical sound as some $30000 DACs (please refer to my equipment inventory and you know what I am talking about). If you happen to have one of them, don't worry, these are the best DACs because they reveal everything from your digital files, nothing more, nothing less. I honor them as 'standard DAC' (the same applies to amplifiers, cables, servers, and any upstream electronics. I simply call them 'standard gears'). I have a lot of them from $40 to $4000. My place is always open for blind tests, send me a PM and you are always welcome. I actually made a lot of new friends by opening my place for blind listening. Here are some statistics, I have tested 53 audiophiles (as of 12/31/2020). None of them can tell the difference between my $40 standard DAC and $4000 Standard DAC. Most of them came to my place with absolute confidence (that they can tell the difference). This number simply means it will be a medical miracle if you can tell my $40 standard DAC and $4000 standard DAC. Well, not saying you can’t buy them, paying a premium sometimes makes us more like men, after all.

4. If you find two upstream electronic devices (like DACs/amps/players) sound very different in blind test sessions even if you matched their levels, it means at least one of them is poorly designed, no matter if it is R2R, Multibit or Delta Sigma whatsoever.

5. There is a very simple way to own yourself a bullshiiiit-detector. Buy a cheap dummy head like Mini-DSP E.A.R.S. and measure frequency responses. They are not industry standard, but they can help you to test if your amp/DAC/cables made any difference. Swap any upstream device (e.g. from amp A to amp B) while keeping the headphone not moving and adjust the SPL to the same level. According to my experience, 99% (actually 100%) of the time, if the frequency responses look identical, then you won't hear any difference in blind tests.

6. Amplifiers for electrostatic headphones sound slightly different from one and another (they have measurable differences in frequency response). Probably it is because ES headphones are not popular enough to draw attention from those really serious people (like TI, ADI, NI, etc). Small market, after all.

7. What determines 99% of the sound quality are the two transducers. The first one is the microphone which does the recording job, another is your headphone/speaker. You have no access to the former transducer in most cases. So spend your money on the latter one.

8. Earpads, damping, and enclosures (or anything related to the physical driver) changes the sound drastically for headphones. They lead to measurable frequency response alterations. They are far more important than cables, amplifiers, or DACs (whereas the frequency responses are usually identical). For speakers, the enclosure, damping, and room environment are as important. Again, they are far more important than cables, amplifiers, or DACs.

9. The headphone transducer technology stopped advancing around 30 years ago. The speaker technology matured even earlier. Every industry insider knows this (check what speakers Nelson Pass has in his home on YouTube). Just like cars, they are yesterday's technologies. Yeah, there is a glorious time for the audio industry. There were some super-smart people working on audio while dinosaurs still walked around on earth (probably in 1920 - 1940, audio at that time was the AI as in 2020). For a more concrete example, headphones like the HD580 beat some $3000-$5000 headphones nowadays. Don't be fooled by the market!

10. Different from the upstream electronics, there is no 'right' or 'standard' sound for headphones and speakers, absolutely no! I know audio manufacturers are trying their asses out to make you think they are working on some 'true to the source' sound. For example, they call them 'flat response' or 'Harman target curve' or whatever.

It's simply because there is a wireless channel between your ears and the transducers. Anyone who works in wireless industries (audio/radio) should understand the wireless channel could only be modeled stochastically (whereas the upstream electronics can be described deterministically without any problem). The only reasons a manufacturer wants to advertise his 'target curve' is either because he wants to brainwash you and sell his taste or he is just innocent and doesn't understand how a wireless signal works.

11. The sound of the best headphones in the market is always based on the music preference of the crowd (of course the primary goal for a company is to make money, not to be a saint), not the so-called 'flat response' or 'target curve'. Therefore, designing headphones/speakers in my opinion is an art instead of engineering. The 'target curve' for headphones or speakers are not the same at different times.

Today’s music like EDM, hard rock and Electronics targets strong bass and they basically don’t care about timbre that much. Today's hi-end headphones are designed for this type of music. They outperform any headphones made in the past when playing recently produced music. However, the major music lovers back in the 80s liked un-amped acoustic/classical music. They care about timbre/transient/air. That's why the best headphones (as in 2020) for classical music are still those made 30 years ago. So a person (like me) might find the best headphones made in the 80s if she likes the music produced in the 80s.


A list of my favorite headphones:
1. Sony MDR-R10 (bass light version). The best all-rounder, sweet sweet fresh harmonious sound (not as sweet as live concerts though). Fast. Best bass department among all headphones. Playing stringed music is sublime!

2. Sennheiser HE90. The most dynamic headphone. Lightning-fast. 'Everything comes back to life' sound. Silky smooth and glorious treble at the same time, bravo!

3. (Tied) Stax-Omega. Godlike soundstage. Simply out-of-head floating imaging. The best of all headphones (include K1000, Myshpere, SR1A) in the soundstage department. Super fast response like your supercharged Shelby GT500. Pure and unsweetened sound, lovely!

3. (Tied) Stax - 007mk1 with ECL-02 earpads. They sound like a mix between R10 and HE90. The timbre reproduction from these variants is spot on! The ECL-02 earpads completely remove the last bit of smear and boomy from the notorious bass find in the entire 007 series. Unfortunately, these magical sounds could not be found in 007MK2s even replaced with the same ECL-02 earpads.

5. Hifiman Susvara

6. Kennerton Thror

7. Stax SR-009

8. RAAL SR1A

9. Sennheiser HE60

10. Sennheiser HD580 (early printed logo version)

11. (Tied) AKG K1000

11. (Tied) Sennheiser HD800

Headphones I do not like:
1. Sonoma M1 Electrostatic Headphones

2. Abyss AB1266/Diana

3. Focal Utopia

4. Stax SR-007mk2

5. Mr. Speaker Voce

I am not saying the headphones above are bad. They might be perfect for some pop music. I don't like them because the tonality of these headphones got some serious issues. For example, I found the stringed instruments on Utopia sound ok, but the piano sound is too wooly. The problem for Abyss is exactly on the opposite, the stringed instrument lacks texture and sweetness on the Abyss.

My headphone colormap, some examples (based on my years of experience dealing with un-amped instruments):

Dark: LCD2, LCD3, 007MK2, HD650
Slightly dark: Focal Utopia, 007MK1, Abyss 1266, Voce, Ether/Ether Flow
Neutral: HD600, HD800, HE60, HE90, SR - Ω, Susvara
Slightly bright: HD580, R10, Thror, SR - 009, K1000, SR1a
Bright: Lambda Signature, TH900, RS1, SR325i

-Thanks for reading! Last updated - 12/31/2020