CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS

  • December 31, 2015, 10:38:16 AM
  • Welcome, Guest
Please login or register.

Login with username, password and session length
Advanced search  
Pages: 1 ... 3 4 5 6 7 [8] 9

Author Topic: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots  (Read 26714 times)

0 Members and 1 Guest are viewing this topic.

Solderdude

  • Grab the dScope Kowalski!
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +206/-4
  • Offline Offline
  • Posts: 907
  • No can do skipper, the dScope was terminated
    • DIY-Audio-Heaven
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #70 on: January 13, 2013, 10:04:18 AM »

I'm not familiar with headphone electrical science, how does the source impedance factor into this exactly?

This question has several answers.

Firstly... the drivers are probably LEVEL matched as impedance matching says absolutely nothing about the drivers efficiency.
The efficiency depends on: magnet field strength, distance between 'coil'  diaphragm and magnets, current through the coil, (voltage across it divided by the impedance, resistance in this case), number of 'windings'.

What is weird is that there is such a spread between the resistances. It might be due to the production method but personally think it is odd.

Given the fact that they measure different yet sound equally loud says that the efficiency of the drivers is not under tight tolerances and they 'match' drivers probably by measuring them and selecting drivers closely level matched.
This could mean the impedance (due to variances in the production process) could vary.

This is not a problem when a low output resistance amplifier (source resistance) is used but due to voltage division COULD cause L-R balance issues when driven from a higher output R amplifier.
A non- issue really as planars should be driven from low resistance amplifiers anyway.

The resistance you measure = impedance for planars. It is different for 98% of all 'normal' dynamic headphones though.
In this case the measured DC resistance is the lowest impedance and in general will be above that value .
Logged
Use your ears to enjoy music, not as an analyser.

paranoidroid

  • Able Bodied Sailor
  • Powder Monkey
  • ***
  • Brownie Points: +6/-0
  • Offline Offline
  • Posts: 27
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #71 on: January 13, 2013, 10:14:16 AM »

That makes total sense. They're level matched and not impedance matched. And it probably just turns out there is an impedance variation due to production process..
Logged

jerg

  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +41/-2
  • Offline Offline
  • Posts: 466
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #72 on: January 13, 2013, 10:16:32 AM »

I'm not familiar with headphone electrical science, how does the source impedance factor into this exactly?

This question has several answers.

Firstly... the drivers are probably LEVEL matched as impedance matching says absolutely nothing about the drivers efficiency.
The efficiency depends on: magnet field strength, distance between 'coil'  diaphragm and magnets, current through the coil, (voltage across it divided by the impedance, resistance in this case), number of 'windings'.

What is weird is that there is such a spread between the resistances. It might be due to the production method but personally think it is odd.

Given the fact that they measure different yet sound equally loud says that the efficiency of the drivers is not under tight tolerances and they 'match' drivers probably by measuring them and selecting drivers closely level matched.
This could mean the impedance (due to variances in the production process) could vary.

This is not a problem when a low output resistance amplifier (source resistance) is used but due to voltage division COULD cause L-R balance issues when driven from a higher output R amplifier.
A non- issue really as planars should be driven from low resistance amplifiers anyway.

The resistance you measure = impedance for planars. It is different for 98% of all 'normal' dynamic headphones though.
In this case the measured DC resistance is the lowest impedance and in general will be above that value .

Thanks for the explanation. So it seems that the impedance imbalance would be fairly benign under normal use then.
Logged

Solderdude

  • Grab the dScope Kowalski!
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +206/-4
  • Offline Offline
  • Posts: 907
  • No can do skipper, the dScope was terminated
    • DIY-Audio-Heaven
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #73 on: January 13, 2013, 10:22:39 AM »

Thanks for the explanation. So it seems that the impedance imbalance would be fairly benign under normal use then.

correct... when driven from low output R amplifiers.

Higher output R amplifiers could have a slight influence on L-R balance but little on tonal balance.
Well a higher output R seems to calm the 10kHz peak down a bit.

Would be cool if Purrin could show these effects in FR and waterfall by simply repeating a measurement with an added output R of say 47 to 100 Ohm.

I have not yet built my own rig, nor do I have access to an HE500 anymore.
Logged
Use your ears to enjoy music, not as an analyser.

AstralStorm

  • Speculation and Speculums
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +250/-164
  • Offline Offline
  • Posts: 559
  • Warning: causes nearby electronics to go haywire
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #74 on: January 30, 2013, 09:50:02 PM »

Interesting. Seems like these are seriously dark/muffled/cupped with the stock leather pads. Kinda similar from memory to my older DT1350, except way better everywhere. Especially the 1800 Hz hole affects the sound, making for weird behind-the-head soundstaging.

Have a yummy correction curve. Of course my pair of HE-500 is not yet burned in as per manufacturer's recommendation, but this shouldn't change anything major.
With or without the correction, there seem to be no audible resonances. Well maybe there's some slight extra grit with electric guitars and low trumpet overblow, I'm not entirely sure. Could be something near 4k.
Bass is hard hitting and not really distorted at all, very fast. Treble has very nice body to it yet is not lacking sharpness at all. Mids are just superb, best I've heard out there. Generally powerful presentation but not lacking in finesse and micronanodetail when equalized. It makes TWFK and old Beyer T70 sound like a gritty mess. Modded equalized RE-ZERO are somewhat close, but far longer delay, reverberated and not as powerful bass impact.

I'll follow up with velour and quite a bit later with home-made Jergpads.

Reference 500 Hz, matched loudness vs the IEMs, calibrated as usual using a set of radio voice recordings (approx. 40 sones at -6 dBFS = not loud). Driven off Leckerton UHA6-mkII, OPA2209, gain switch on, dial around 2/3 position.


I'd love to send them in for the measurement, but I want to compare them head to head vs Paradox first.

On another note: Hifiman cannot make cables. The HE-500 cables are horrible PVC or PET (or some other crud stiff transparent plastic) that causes nasty plasticky microphonics, aggravated by the braid. It is also very long. Fortunately unlike IEM's it's replaceable.
« Last Edit: January 30, 2013, 10:31:52 PM by AstralStorm »
Logged
For sale: Hifiman HE-500; Paradox; Brainwavz B2. PM me if you would like to buy them.

jerg

  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +41/-2
  • Offline Offline
  • Posts: 466
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #75 on: January 30, 2013, 10:36:27 PM »

Interesting. Seems like these are seriously dark/muffled/cupped with the stock leather pads. Kinda similar from memory to my older DT1350, except way better everywhere. Especially the 1800 Hz hole affects the sound, making for weird behind-the-head soundstaging.

Have a yummy correction curve. Of course my pair of HE-500 is not yet burned in as per manufacturer's recommendation, but this shouldn't change anything major.
With or without the correction, there seem to be no audible resonances. Well maybe there's some slight extra grit with electric guitars and low trumpet overblow, I'm not entirely sure. Could be something near 4k.
Bass is hard hitting and not really distorted at all, very fast. Treble has very nice body to it yet is not lacking sharpness at all. Mids are just superb, best I've heard out there. Generally powerful presentation but not lacking in finesse and micronanodetail when equalized. It makes TWFK and old Beyer T70 sound like a gritty mess. Modded equalized RE-ZERO are somewhat close, but far longer delay, reverberated and not as powerful bass impact.

I'll follow up with velour and quite a bit later with home-made Jergpads.

Reference 500 Hz, matched loudness vs the IEMs, calibrated as usual using a set of radio voice recordings (approx. 40 sones at -6 dBFS = not loud). Driven off Leckerton UHA6-mkII, OPA2209, gain switch on, dial around 2/3 position.


I'd love to send them in for the measurement, but I want to compare them head to head vs Paradox first.

On another note: Hifiman cannot make cables. The HE-500 cables are horrible PVC or PET (or some other crud stiff transparent plastic) that causes nasty plasticky microphonics, aggravated by the braid. It is also very long. Fortunately unlike IEM's it's replaceable.

I do hope you try my earpad mod soon and post some impressions.

As for cabling, the HE6 coppers or the old HE500 coppers were much better than this new silver-plated copper crap. Really soft and flexible, and minimal microphonics compared to the new silvers.
Logged

AstralStorm

  • Speculation and Speculums
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +250/-164
  • Offline Offline
  • Posts: 559
  • Warning: causes nearby electronics to go haywire
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #76 on: January 31, 2013, 07:38:49 AM »

Interesting. Velours are less linear, but bass is quite a bit tighter still, but less hardhitting. (regardless of eq) Perhaps the impact is actually some harmonic distortion?
Crunch at 4k is gone. Sound is a tiny bit less reverberated, slightly more precise everywhere. Blacker background?
Sounds closer to balanced armatures with these on - but way more refined.

Same conditions, same loudness.  :)p8 I've even checked surrounding temperature.


Sound with leather pads on sounds silky or perhaps liquid, while with velours sounds... veloury, refined.
The latter presentation is way more monitor worthy.

Edit: Actually, the lack of "hit" was due to slightly too deep 37 Hz cut. Fixed and it's clearly superior in all regards - well, it does still feel like a tiny bit less impact. Way more listenable unequalized as well - less dark.
« Last Edit: January 31, 2013, 08:18:33 AM by AstralStorm »
Logged
For sale: Hifiman HE-500; Paradox; Brainwavz B2. PM me if you would like to buy them.

AstralStorm

  • Speculation and Speculums
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +250/-164
  • Offline Offline
  • Posts: 559
  • Warning: causes nearby electronics to go haywire
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #77 on: February 01, 2013, 11:14:48 PM »

More experiment: now with Beyer T70 velours. (same as T1 and DT990 I think) Had to trim the mounting plastic ring using nail scissors to accomodate these.
These velours are made of thicker cloth, are softer and flatter (thinner). Not to mention they don't have the dust protection screen.

The sound has a body inbetween leathers and velours. Lack of the dust protector improves air quality quite a bit. (The eq boost actually now can bring them to flatness. Previously I had to cheat in a slow rolloff.)  Way, way sharper than even TWFK, but this isn't fake sharpness - does not sound like odd order harmonic distortion. It isn't overly liquid smooth like leathers, but doesn't have the slight grit of stock velour either.
Sound without equalization is quite similar (from memory) to GR07, but way, way more refined and airy.
Lower mids and bass have a bit more body with these than with stock velours. Soundstage is slightly narrower in a good way, forward positioning is better than leathers or stock velours. No slight shout or reverb unlike leather pads.
With equalization, these are the winner for now. Equalized the sound is like equalized modded RE-ZERO, but even more neutral and much better note weight and power. Less reverb, more precision. :)p1

Here's the correction curve. The very narrow notch is quite dependent on placement on ears in all cases. This one at 8500 Hz is with the headphones pushed slightly forward on the head. Ranges from 8200 to 9500 Hz, but is trivial to reproduce reliably on the head.



The notch is probably the acoustic shadow of sorts of the huge mounting wire in the middle front of the driver.

Huh, these are now merciless vs mastering issues and encoding artifacts, a true monitor quality. They do keep all the qualities of the recording now.
« Last Edit: February 02, 2013, 12:13:14 AM by AstralStorm »
Logged
For sale: Hifiman HE-500; Paradox; Brainwavz B2. PM me if you would like to buy them.

jerg

  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +41/-2
  • Offline Offline
  • Posts: 466
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #78 on: February 02, 2013, 03:46:05 AM »

More experiment: now with Beyer T70 velours. (same as T1 and DT990 I think) Had to trim the mounting plastic ring using nail scissors to accomodate these.
These velours are made of thicker cloth, are softer and flatter (thinner). Not to mention they don't have the dust protection screen.

The sound has a body inbetween leathers and velours. Lack of the dust protector improves air quality quite a bit. (The eq boost actually now can bring them to flatness. Previously I had to cheat in a slow rolloff.)  Way, way sharper than even TWFK, but this isn't fake sharpness - does not sound like odd order harmonic distortion. It isn't overly liquid smooth like leathers, but doesn't have the slight grit of stock velour either.
Sound without equalization is quite similar (from memory) to GR07, but way, way more refined and airy.
Lower mids and bass have a bit more body with these than with stock velours. Soundstage is slightly narrower in a good way, forward positioning is better than leathers or stock velours. No slight shout or reverb unlike leather pads.
With equalization, these are the winner for now. Equalized the sound is like equalized modded RE-ZERO, but even more neutral and much better note weight and power. Less reverb, more precision. :)p1

Here's the correction curve. The very narrow notch is quite dependent on placement on ears in all cases. This one at 8500 Hz is with the headphones pushed slightly forward on the head. Ranges from 8200 to 9500 Hz, but is trivial to reproduce reliably on the head.



The notch is probably the acoustic shadow of sorts of the huge mounting wire in the middle front of the driver.

Huh, these are now merciless vs mastering issues and encoding artifacts, a true monitor quality. They do keep all the qualities of the recording now.

Are your parametric EQs based solely on your perceived frequence response?
Logged

AstralStorm

  • Speculation and Speculums
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +250/-164
  • Offline Offline
  • Posts: 559
  • Warning: causes nearby electronics to go haywire
Re: HiFiMAN HE-500 Frequency Response and CSD Waterfall Plots
« Reply #79 on: March 09, 2013, 11:55:46 PM »

Yes, the parametric eq are 100% by ear and are well reproducible - the nasty sharp dip can be slightly wider and shift within 1khz due to placement on head. The roudness in the earlier eqs are due to the more precise estimation of the peaks taking lots of time.

I can confirm that Paradox is nearly perfeclty linear with some bass "round" boost combined with slight slow subbass roll (-2 dB), wide small sparkly 6-10k boost and a relatively small (6 dB) high shelf beyond about 11 kHz. Everything else within 1 dB. The sound is laid back, bass boosted, small space but refined and energetic. Highest fidelity of the three.
RE-400 on the other hand have some 4k boost (3 dB?) and some slight highest end shelf (also -3 dB), also wideband bass boost, with 0 at 30 Hz, -3 dB at 150 Hz. The sound is bass boosted and slightly splashy. Also highly energetic, bit smoothed out.
HE-500 have some 5-6kHz boost (3 dB), missing 8kHz (-2 dB + peak) energy and recessed 1-2kHz (-1, -2 dB) and are grainier than both of the above, suggesting distortion or unevenness in FR. Bass is "noisy". Lowest fidelity of the three.
Not comparable: TWFK Brainwavz B2 - noticeably bandlimited, midbass boost (+2 dB), mids cut at 1, 2kHz (-2, -4), major highs boost centered on 5.5kHz (+8 dB). Highly splashy and flattened soundstage, noticeable mids distortion and some high mids ringing. Fidelity similar to HE-500, but different issues.

I'm not sure what's happening with the Hifiman's highest end though. Paradox are about as bright when unequalized, yet more linear - both are a bit dark, but in a very different way. HE-500 is closer to neutral, but less even - dips are detectable. Paradox are somwhat dark and tiny bit bass boosted, but are smoother and more linear, even when equalized. Bit cleaner all the way. (HE-500 with velours with screen removed.)

I might have to send this pair to Purrin for measurement. It is possible that my anatomy is just shifting the resonances upwards noticeably or my pair has higher membrane tension.
« Last Edit: March 10, 2013, 11:13:03 AM by AstralStorm »
Logged
For sale: Hifiman HE-500; Paradox; Brainwavz B2. PM me if you would like to buy them.
Pages: 1 ... 3 4 5 6 7 [8] 9