grapefruit
100+ Head-Fier
- Joined
- Jan 28, 2012
- Posts
- 100
- Likes
- 14
I am a systems neuroscientist who moonlights as a sound engineer in an established studio for fun + cash.
Everything I present below is standard knowledge for sound engineers, but applied to audiophile topics. I encourage other sound engineers to weigh in. I hope to undercut popular descriptors we use and propose we revise how we talk about gear. In doing so we can fully realize which vendors are doing good work and which are actually producing very dishonest sound.
I will just touch on a small bit, and if this gets traction, I will continue with other concepts and maybe include some sound demonstrations.
Point 1: True neutral (and harman neutral) sounds like crap.
All the rich detailed information captured is imperceptible by audiophile standards and must be brutalized into existence. Don't believe me? Bring an omnidirectional calibration mic to a sound deadened room and listen.
Point 2: Natural dynamic range, which far outstrips sound and reproduction formats, sounds like crap
A "natural," "well mastered" record will be around 0.00004x as dynamic as how it actually sounds. Just like how looking at a screen, no matter how hi res, will never appear like looking through a window because no backlight is as bright as the sun.
Point 3: The stronger the sound signature, the worse the accuracy and fidelity.
This is not a hot take but it is somehow not appreciated at the forefront of peoples minds. In terms of neutrality and accuracy, most music out there has a pretty offensively mangled sound signature simply because putting sounds together on a record is extremely difficult. Gear that imparts clarity and neutrality onto everything is lying to you point blank.
Thus "neutral precise accurate sound" among hi quality gear is a generous ****-on of all fidelity.
The ER-4s is a great example.
The mainstream standard of accuracy. It's been a favorite of mine for over a decade long before I knew anything about sound. I know how it sounds by heart: a loss of mud and sibilance, a locked-in forwardness of sound and separation, complemented with heaps of detail you never heard before.
Turns out you can produce a signal chain to etymotic-ify anything. Audiophiles probably think it's eq. It aint.
It starts with removing dynamic range. You will notice that etymotics have little impact. That's because all louder things are "pushed down" to a somewhat equivalent level to quiet sounds. This has the psychoacoustic effect of bringing things forward, solidifying them, and heightening detail. By tuning the enclosure, you can selectively compress and eliminate different parts of the sound spectrum. Etymotic is brutally good at removing sibiliant frequencies without losing air.
Then comes the clear, airy, nonsibilant highs. To remove sibilance regardless of source material, you cut the 6-8khz range down, or potentially eliminate small chunks of it completely (called notch filtering). Then you prevent sound in that region from moving up or down in loudness through compression—and the way speaker designers do so is pretty unforgivable.
Third is exaggerating resonances and smearing timing to "tune"
So how are frequencies selectively cut (voluntarily or involuntarily) without an EQ? Next to driver selection, the most common way is manipulating the enclosure such that there is a partial delay in the sounds that are reintroduced before the complete sound exits. By introducing pipe resonances in the 6-8k range of the delayed signal, when you add it back to the exiting soundwave, it will attenuate that 6-8k range also prevent it from rising no matter how loud the sound is.
Essentially, all enclosure tuning introduces gross delays in timing and resonances, which manufacturers market as "neutral."
Altogether, I'd like to invite people to talk more about compression and understand how ****ed up our favorite gear is. Next I will update with additional info on sweetness, PRaT, soundstage, and separation.
Jeez. I feel better. Would like to hear your thoughts?
Everything I present below is standard knowledge for sound engineers, but applied to audiophile topics. I encourage other sound engineers to weigh in. I hope to undercut popular descriptors we use and propose we revise how we talk about gear. In doing so we can fully realize which vendors are doing good work and which are actually producing very dishonest sound.
I will just touch on a small bit, and if this gets traction, I will continue with other concepts and maybe include some sound demonstrations.
Point 1: True neutral (and harman neutral) sounds like crap.
All the rich detailed information captured is imperceptible by audiophile standards and must be brutalized into existence. Don't believe me? Bring an omnidirectional calibration mic to a sound deadened room and listen.
Point 2: Natural dynamic range, which far outstrips sound and reproduction formats, sounds like crap
A "natural," "well mastered" record will be around 0.00004x as dynamic as how it actually sounds. Just like how looking at a screen, no matter how hi res, will never appear like looking through a window because no backlight is as bright as the sun.
Point 3: The stronger the sound signature, the worse the accuracy and fidelity.
This is not a hot take but it is somehow not appreciated at the forefront of peoples minds. In terms of neutrality and accuracy, most music out there has a pretty offensively mangled sound signature simply because putting sounds together on a record is extremely difficult. Gear that imparts clarity and neutrality onto everything is lying to you point blank.
Thus "neutral precise accurate sound" among hi quality gear is a generous ****-on of all fidelity.
The ER-4s is a great example.
The mainstream standard of accuracy. It's been a favorite of mine for over a decade long before I knew anything about sound. I know how it sounds by heart: a loss of mud and sibilance, a locked-in forwardness of sound and separation, complemented with heaps of detail you never heard before.
Turns out you can produce a signal chain to etymotic-ify anything. Audiophiles probably think it's eq. It aint.
It starts with removing dynamic range. You will notice that etymotics have little impact. That's because all louder things are "pushed down" to a somewhat equivalent level to quiet sounds. This has the psychoacoustic effect of bringing things forward, solidifying them, and heightening detail. By tuning the enclosure, you can selectively compress and eliminate different parts of the sound spectrum. Etymotic is brutally good at removing sibiliant frequencies without losing air.
Then comes the clear, airy, nonsibilant highs. To remove sibilance regardless of source material, you cut the 6-8khz range down, or potentially eliminate small chunks of it completely (called notch filtering). Then you prevent sound in that region from moving up or down in loudness through compression—and the way speaker designers do so is pretty unforgivable.
Third is exaggerating resonances and smearing timing to "tune"
So how are frequencies selectively cut (voluntarily or involuntarily) without an EQ? Next to driver selection, the most common way is manipulating the enclosure such that there is a partial delay in the sounds that are reintroduced before the complete sound exits. By introducing pipe resonances in the 6-8k range of the delayed signal, when you add it back to the exiting soundwave, it will attenuate that 6-8k range also prevent it from rising no matter how loud the sound is.
Essentially, all enclosure tuning introduces gross delays in timing and resonances, which manufacturers market as "neutral."
Altogether, I'd like to invite people to talk more about compression and understand how ****ed up our favorite gear is. Next I will update with additional info on sweetness, PRaT, soundstage, and separation.
Jeez. I feel better. Would like to hear your thoughts?
Last edited: