Discussion in 'Headphone Measurements' started by purr1n, Sep 28, 2018.
Yeah, go for it. The treble can be boosted a little.
Curious about how this compares to the Stax L300/L500/L700 paired with either the 353 or 007 amp/energizer (mostly L300 with 353 amp given price points).
IMO a lot of music didn't sound "quite right" with the L300 - everything seemed to decay way too fast (ex: drum cymbals wouldn't ring) amongst many other unique nuances. As expected bass was lacking as well although the "quality" of the bass itself was very good. 700s were pretty nice with either amp mentioned above but for the same or less money I liked other dynamic or orthos + amp combos more. Both L300/L700 really excelled at other aspects though - at this price especially, if Koss 95x + amp/energizer doesn't have the nuances of the L300 + 353 or 007, I'm very tempted.
edit: Limited to 360... 316 sold as of right now... uhhhhhhh
You are not the only one. My wife dislikes e-stat like sound, she hates all stats and also the thin membrane HFM's: HEX, HEK, and Susvara. So you get the idea. Says there is too much unnatural separation resulting in a too analytical too left brained sound that is good for analysis but not musical enjoyment.
The one saving grace is that the Koss ESP95X maybe isn't quite as fast as the STAX Lambdas; but in the end she vastly preferred the slower HD800 (modded), Code-X, and Verum over the Koss.
To me, the ESP95X is a refreshing change that doesnt break the bank. Its frequency response* and timbre suit my preferences very well - whereas there are maybe two or three Lambda models (from the 1990s to now) I find tonally acceptable. I've gone to the deep end with stats with a DIY T2 before and I don't feel a need to go back there again.
As an aside, the SRM-323 sucked ass. I doubt the SRM-353 is much better. STAX seems to be making shittier and shittier electronics. The latest TOTL STAX amp released not too long ago was a colossal POS considering the price.** Need to move up to the next price bracket from SRM-3x3 for a notable improvement IMO.
I have a SRM-323 lying around and will see how it goes wirh the ESP-95X to be sure. I do know that the Koss will scale with better amps.
* Too much has been made of the shelved bass below 50Hz. A speaker that does 50Hz before rolloff is considered good / decent. In this case, its not rolloff even, it's a shelf down 5db. Unless you are into EDM or trance, not much content in sub bass in most recordings.
** This STAX amp was so forgettable that I don't remember the model number. There is a review / quick impressions from me here on SBAF where I held back a lot. Can't even find the link.
The problem with estat bass isn’t light diaphragms, rather it’s the small stator-to-stator distance which limits excursion. This supposeddly is somewhat fixed on round driver Stax phones, but increased stator distance means you need a very powerful amp to drive them properly.
Then there’s THD, or rather the lack of. When I worked for SW, we had some interesting discussions on system THD and it’s impact on mixing decisions for music. Basically the theory was that having an extremely clean system output could get you a “too naked” signal, as the engineer would hear it “enriched” by the harmonic content of his/her monitoring chain. Adding extra harmonic richness isn’t unusual for mixing, the question remains what the engineer is shooting for and how far off is the listener with the gear at hand.
This is quite interesting. So in a sense a measured dose of harmonic distortion is actually desirable in most cases as it enriches music with aesthetically pleasing sonic colourations, or is it that there are recording engineers who count on systems having THD that passes a certain threshold, ergo being too clean actually runs counter to that objectivist maxim of pursuing "what the artist intended people to hear"? Fun.
I can understand people preferring dynamic drivers for their having actual body in music, the HEK et al certainly felt like instruments just had no impact, that it's for left-brainers who want to pick everything apart, but the relaxed presentation does deserve a place in audio.
If anything, it goes to show that FR alone won't really tell you poop about how something will sound, though it is good for painting a picture of something with broad strokes.
Well, this is hardly proven knowledge, so take it with a sizable grain of your favorite condiment. Basically the mechanism is that even the systems used by sound engineers have some THD and they tweak the mix until it sounds great. One of the measures can be harmonic richness of the instruments. Basically, if you have less inherent THD on your system than the one the engineer used, you'd be getting a tad different sound. At least, that's the theory.
AFR does tell the most of any other 2 axis measurement. The IR does tell more, but good luck Fourier transforming that in your head. CSD essentially tells the most. What many people do, is they fall into the reification fallacy with measurements. AFR is focused indicator to describe a real physical system that has potentially infinite complexity. In a minimum-phase system (like all transducers) AFR discrepancies almost always indicate a presence of some phenomena which has it's footprint in both time and frequency domain. You then work to eliminate the phenomenon, not to just straighten some measurement. All minimum-phase phenomena can be fully remedied by minimum-phase filtering, if they're correctly measured to reverse to make a perfect filter. Exceptions are perfect destructive interference (rare), non-minimum phase phenomena and occasions when system cannot gracefully take the corrective signal.
Two cups of mustard seed ready, then. Nah, it checks out for me, I've observed that some measure of 2nd-order harmonic distortion does tend to be less obnoxious on the ears, and may even be preferable for some. I do think this is kinda funny, as it adds yet another dimension of complexity to the pursuit of "artist's intent" that many advocate; I've long since given up and just want a nice-sounding system that doesn't suck too badly with any one genre now, haha.
Sounds like a discussion worth having, to be honest, though one would need a great deal of experience on the production side of things to contribute; back to lurking for me then
Impulse response (that is what IR refers to, yes?) measurements are definitely cool, but your point about how much of a pain it'd be to mentally adjust these stands. I didn't mean to imply that FR wasn't a greatly useful way to visualise how a component sounds, mind, only that in certain sectors of the internet you do get those who claim other measures such as CSD are useless. Logic doesn't check out for me, but people will believe what they will (eye roll).
Square waves seem like a great way to focus on specific problem areas, the limitation thereof being (it seems to me) that you don't get as complete a picture of the entire frequency spectrum, or as comprehensive a view of a component's sonic attributes. I remember @purr1n found some interesting things using burst response not too long ago (https://www.superbestaudiofriends.o...burst-response-hd800-sr-207-hd650.3688/page-9), supplemented by stuff from @JohnM, but it's apparently a massive PUTA* PITA to deal with so that kinda sucks. Wonder how some better-received cans would compare using this measure?
Also, "reification". That's not a word I've heard in recent years, thanks! Love getting to exercise my vocab
I came across this while looking stuff up to make sure I wasn't talking out of my ureter. Logic checks out for me, but there is of course an off-chance someone who knows their stuff might be able to offer a good counterargument. AKG K712>Sennheiser HD600 is an opinion that may start fires around here, haha: https://docs.google.com/document/d/1ohBFU92JJUJcAocbwO6biedlrrrmDb7yh1dd57BkhiM/edit
... Hella off topic. sorry.
*This was an actual typo that made my laugh out loud so I let it be and just crossed it out.
I'm not experienced enough yet to know if I dislike all e-stats (especially given cost ratios), but I have also tried the SR007-mk2 and SR009. On paper performance was really good. But the frequency response of the SR009 was only about 6/10 for my preferences. SR007-mk2 had a 7.5/10 frequency response for me (and idk if they are actually different when graphed), but to me all the other "on paper" performance stuff wasn't as "technically good". (I don't have 1 ideal freq response).
Also tried the VOCE for ~5min in a non-ideal setting - the only thing I got there is that I need to try it again. Have yet to try a Sonoma M1 or Koss. No interest in spending $30k+ on HFM Shangri-la or Sennheiser HE1 or Orpheus short of winning the lottery.
I'd concur with you on the SR007mk2 and SR009 FR. I had an SR009 and sold it. The SR007mk2 with your FR rating of "7.5/10" was keepable for me. The ESP950 and HFM Jade (modded) actually got more time than the SR007mk2.
VOCE seems interesting, but again I don't want to go down that road. The $300 5% downpayment that I'd put toward a VOCE and BHSE, I'd rather put toward buying / selling interesting DACs to satisfy my curiosity (there seems be to a lot of good boutique stuff now) or outright buying a Google WiFi mesh x3 pucks for the house. Priorities.
For anyone that calls this a "hobby" for themselves, having various technologies in the headphones is pretty par for the course. Dynamics, biocellulose, planar, and estats represent the tech. Getting the Koss for ~$500 used, or the MD for $499 new is pretty much a no brainer. It's the best, most inexpensive way to listen to estats . Hell, some people here spend more on cables. You can even get/make a Stax adapter and take the headset to meets to plug into the "big boy" amps and truly experience how they scale.
That said, I would recommend listening exclusively for at least three days to any new headphone, especially when it's a new technology for you. We all know that these aren't the headphones for heavy bass genres, so if that's your thing, just move on. If, ever, you like music that doesn't require a subwoofer to hear the melody, get some of these and add it to your rotation, especially for jazz, folk, classical, and even older rock.
These headphones are fine with bass heavy tracks. Alicia Keys Unplugged was absolutely fine. Bass impact was good and no worse than say an HD800 powered by Magni 3 with a Topping DAC.
Where the Koss falls behind is with the deep low fundamental notes which are muted. Most people probably wouldn't notice because they are used to hearing the second fundamental from their car stereos or HD650.
My wife dislikes stats because of none of the above. She dislikes them because of their transient response characteristics, of which there are no handy standard measurements.
As for the e-stat low harmonic distortion angle being too clear or naked, I don't necessarily buy that. Dynamic speakers with big woofers and horn compression drivers have barely any THD when used in a home or studio environment. Also, the Koss ESP95X distortion measurements are pretty good, on par with dynamics, but nothing close to the performance of TOTL stats and orthos.
Back to transient response: I think there is something different to how a lightweight high surface-area tensioned surface accelerates and deaccelerates compared to high-mass low surface-area higher excursion surface supported by a spider and surrounds.
And yes, I could imagine an alternate universe where Quad ESL (instead of Altec/JBL) dominated the industry during the heyday of audio 70-80 years ago, and people now bitching about the slow or smeared sound of dynamics instead - the audio engineers would have mixed and mastered for ESL after all.
But that's Man In High Castle stuff.
Funny you mention quad. They have some headphones now too. Planar. ERA-1 ~$800
And a headphone amp. PA-1 $1800
Thanks, that makes sense. Would you include the Utopia in the heft, slam, and tactility group? I may never have experienced bass like you're describing
Utopia's got some slam in the midbass. Below 40Hz is Utopia's weakness - all the 40mm Focals run into excursion limits - where we start to note weird distortion. This does not happen often and very rarely is there a lot of energy below 40Hz in most recordings, including recordings we think there is a lot of bass.
While I do maintain the ESP95X's bass is fine, it's not ever going to slam or punch on the midbass like Utopia, HD650 or HD800 or have the low end heft and sustain of HE-6/5 or LCD2C (assuming capable amplification).
Keep in mind that ESP950 (stock pad version) sounds midbassier with less low-sub bass than the ESP95X. In other words, the ESP95X's bass is less boomy and rolled off.
Seal is going to play a role, so my observations may not apply to others.
They look at least promising.
Thanks! Will iFi have them at RMAF?
Nope. But we'll have something really big from AMR with us.
I've kinda realized that I should be mixing "cleaner", as my system is very very clean, and I mix it so it sounds nice and thick and dirty. So on a normal syatem you'll get a really filthy sound. Which I like, but still.
Separate names with a comma.