General Headphone Audio

Ultimately that's down to Apple to resolve I guess, Windows has had WASAPI since Windows 10, someone needs to nudge Apple in the right direction as they have the display side of things perfected with HDR and everything (where Windows falls flat on its face).
 
One thing the last few weeks made me aware of (if I wasn't already) is that whether it's a unix-like OS, or Windows 10/11 every OS seems to be imperfect when it comes to how they deal with audio.
 
I think Windows has it better than the rest since it has native support for WASAPI Exclusive, none of the other OSes offer anything similar. The USB delay thing to be fair is down to the device vendor to implement either on hardware or in the driver control panel, but smartphones are a different story, whilst Android does offer low level interactivity to the audio stack, like you can plug in a USB DAC and a compatible app will talk directly to the DAC bypassing the OS audio stack entirely, the app vendor needs to support this function so this feature would be landing on the shoulders of Spotify, Tidal etc.

So for once, Windows is actually the winner here when it comes to respecting bit-perfect audio.
 
Hit another fun bug on the Fiio k15, whenever I play a flac file in my library that is 48 or 96khz it seems to disable the eq but still show it as enabled on the display. Swapping profiles re-enables it for the track.
 
I think Windows has it better than the rest since it has native support for WASAPI Exclusive, none of the other OSes offer anything similar. The USB delay thing to be fair is down to the device vendor to implement either on hardware or in the driver control panel, but smartphones are a different story, whilst Android does offer low level interactivity to the audio stack, like you can plug in a USB DAC and a compatible app will talk directly to the DAC bypassing the OS audio stack entirely, the app vendor needs to support this function so this feature would be landing on the shoulders of Spotify, Tidal etc.

So for once, Windows is actually the winner here when it comes to respecting bit-perfect audio.

There is work going on, or other complete audio equivalent systems/API's on Linux such as Jack (very much orientated for audio production like ASIO), gstreamer etc. I don't think there is quite the same need to bypass the audio stack on Linux (i.e. it's not as messy). In theory both OS can and do sent the same audio. But it's clear Linux misses out from a complete lack of support from Vendors (more than just audio) and lack of software.
 
Last edited:
I did have 16bit 44.1khz as default but for the next few months i might go with 24/48khz until my Apple Music sub runs out.
 
One thing the last few weeks made me aware of (if I wasn't already) is that whether it's a unix-like OS, or Windows 10/11 every OS seems to be imperfect when it comes to how they deal with audio.

Windows is terrible under the hood when it comes to audio, they depreciated a lot of the convenient methods for using audio with the Win32 API as well (though MCI and stuff like sndPlaySound are pretty limited) and the replacement APIs in typical MS fashion of late are a horrendous convoluted mess.
 
Last edited:
Windows is terrible under the hood when it comes to audio, they depreciated a lot of the convenient methods for using audio with the Win32 API as well (though MCI and stuff like sndPlaySound are pretty limited) and the replacement APIs in typical MS fashion of late are a horrendous convoluted mess.

Indeed, I mean this is beyond my comfort zone and more comparison points comparing the different OSes would be interesting. However I do understand why Windows might be preferred.
 
And here we go, the very reason why measurements are mostly pointless in a general scale, they are exactly as I assumed a while back, good to baseline against headphones on the same review just to highlight where one differs to the other, but for what audio sounds like to human ears? There is only one measurement that matters, each person's individual ears. No two people will ever hear the exact same sound coming from headphones or speakers for that matter, which explains why some measurement thumpers will outright deny anyone who trusts their ears alone, aka ASR forums for the most part.


So whilst measurements might be scientific, so is biology, and biology cannot be given a fixed value. So like a piece of art, measurements go out the window when referring to how something sounds to the actual ears as its' different to every single ear on the planet.
 
And here we go, the very reason why measurements are mostly pointless in a general scale, they are exactly as I assumed a while back, good to baseline against headphones on the same review just to highlight where one differs to the other, but for what audio sounds like to human ears? There is only one measurement that matters, each person's individual ears. No two people will ever hear the exact same sound coming from headphones or speakers for that matter, which explains why some measurement thumpers will outright deny anyone who trusts their ears alone, aka ASR forums for the most part.


So whilst measurements might be scientific, so is biology, and biology cannot be given a fixed value. So like a piece of art, measurements go out the window when referring to how something sounds to the actual ears as its' different to every single ear on the planet.

Counter to that though even though due to ear shape giving variation between people measurements are still useful for each individuals to compare responses between headphones.
If I for example open squiglink, while I may not hear the responses exactly as represented and may perceive it slightly differently to others, how i perceive it should impact all measurements(done to same standard) in the same way.
Measurements for headphones only ever gave a general idea of profile anyway, as you say good for comparison but also good for eq and tuning.

The drama in that video doesn't really change anything imo just because one youtuber had an opinion and another disagreed and provided some badly presented measurements.
The research that goes into it acknowledges the variation between people and does its best to compensate for it and represent it.
Im pretty sure that the guys developing headphones/speakers are measurement thumpers aswell.
 
Last edited:
And here we go, the very reason why measurements are mostly pointless in a general scale, they are exactly as I assumed a while back, good to baseline against headphones on the same review just to highlight where one differs to the other, but for what audio sounds like to human ears? There is only one measurement that matters, each person's individual ears. No two people will ever hear the exact same sound coming from headphones or speakers for that matter, which explains why some measurement thumpers will outright deny anyone who trusts their ears alone, aka ASR forums for the most part.

So whilst measurements might be scientific, so is biology, and biology cannot be given a fixed value. So like a piece of art, measurements go out the window when referring to how something sounds to the actual ears as its' different to every single ear on the planet.

I didn't see the original video, but did however watch DMS's response and have seen the above. It seemed to me Super Reviews made some mistakes based on mis-understanding of how measurements work which DMS called out that out in a fairly calm manner. Reviews will always have an element of subjectivity, but I do think there needs to be a baseline in terms of benchmarks you can understand, even if that includes some not-ideal measurements and how they are presented.

My other takeaways are that again LTT is an entertainment channel with OTT reactions, Youtube thumbnails are still awful and oh the drama...
 
There was another comment chain on reddit iirc where it was talked about re: standardised measurements, that I agree would be more useful but as passion for sound mentions in his video, not every reviewer will have the same rig, and the best rigs which cover the widest FR can cost up to ~$40,000! With the average for a more restricted one being around 9k.

Of course though manufacturers will be interested in measurements, they have to be in order to outline specs pages like we see from amp makers all the time, that's fine as they are actually making the products, the community however does seem to focus a bit too much on it a lot of the time and typically that devolves into outright denial within those pockets that someone can hear something differently to what a graph has shown they should be hearing - Classic case in point from recent times from ASR and reddit threads where apparently all amps sound the same and there's no way you would hear even a small difference between two similarly priced amps that run on two different architectures.
 
Amp/dac measurements and headphone frequency response measurements are very different beasts though. And as we've gone over before, there are people in denial on both sides of the argument, some people just swear blind you can/cant hear a difference and refuse to accept/consider anything to the contrary.
 
Last edited:
Amp/dac measurements and headphone frequency response measurements are very different beasts though. And as we've gone over before, there are people in denial on both sides of the argument, some people just swear blind you can/cant hear a difference and refuse to accept/consider anything to the contrary.

Agreed, I don't think after watching DMS's video I'm about to tie a bow on the topic and declare the death of measurements as easily as MRK is.

Measurements are still extremely important, especially for speakers, headphones, and room treatment/response, IEMs are a little more flakey to measure for obvious reasons.

But pertaining to what MRK has mentioned I do agree to a point regarding different measuring rigs etc, it's part of the reason I think GoldenSound was spot on regarding the X9's headphone EQ profiles as we don't even know how they was measured so the baseline of those measurements are up in the are tbh and something you can really rely on.
 
Last edited:
This is nothing new, I always never put much value on measurements like how some people have done and always said it's nice to be aware of them, but as we all know, what a graph shows is not what human ears actually hear a lot of the time and relying on just a graph to form an pre-expectation is counter productive rather than just trusting your ears and then just being aware of a standardised measurement like system specs on general hardware. This is why I never delve into remeasurements, as I just don't care. The LCD-5 is a prime example of this where measurements are amazing on paper, but they sound like trash without EQing.

As for the headphones profile EQing bit, well that applies to other brands like FiiO too as they too have a large headphones database and target curves, with no info on the rigs used there either. The principle of this I mentioned before anyway, to get a fairly decent baseline that sounds OK, then tweak the PEQ to match your preference and save it as a custom one. Saves having to start from scratch which saves time and makes life simpler. This is exactly what I did with the LCD-5 with both the K13 R2R and X9. No other reviewer has expanded on this principle.

If you solely rely on the preset curves then yes the above will be a factor to keep in mind, but as always, presets won't sound the same to everyone, so need fine tuning on a per-person basis anyway.
 
Last edited:
I disagree that you have a "fairly decent baseline" when we are completely blind to knowing the process and equipment used to measure the headphones for the X9's profiles, I wouldn't want to base anything off those profiles, it's like building a house without knowing nothing about the foundations, how can you rely on that? and I say that as someone that thinks measurements are important to point, of course HRTF is also very important as I've always mentioned, really the only way to go is learning how to EQ from scratch to match your own personal preferences and HRTF.

Where measurements are very handy is for example using an audio precision apx555 analyzer to check how an amplifier is performing, also room response, and comparative measurements, as Passion for Sound says,

"Measurements are still very valuable, especially if you can compare the measurement of the headphone or earphone you're looking at with the measurement of another headphone or earphone that you know very well so long as it's been measured on the exact same rig and with all the same compensation applied"

and that has been my use case for measure for decades.
 
Last edited:
No you have misunderstood what I said there.

By decent baseline I am referring to an EQ profile that sounds good enough to start with to the ears, and then fine tune from there and resave as a custom one. I did not say decent baseline from a standards point of view as there is no such thing currently, only an interpretation of what might be considered good on paper based on known rig configs etc.
 
Last edited:
But how can a set profile sound good "enough" to everyone when said persons HRTF comes into to play which effects everyone differently, that's my point, it's not a reliable foundation, especially when there is no understanding of how they even got to that baseline (equipment, measuring rig etc), you might as well just put "jazz", "classical", "Rock" presets on there, it might sound good to some people but it's a very poor foundation, if you are looking to get the best out of your headphone the EQ'ing from scratch is really the only way to go imo.
 
Last edited:
I assume all their preset headphone profiles are just trying to pull everything more towards Harman, but without any details I could be wrong and have no idea. Personally, I would rather not base a custom profile on that, I would rather just start with the stock sound, use some frequency measurements to get an idea of where its generally spiked or recessed in the range and take it from there based on how it sounds.
 
Last edited:
Back
Top Bottom