[gammi]

Just got my in ear microphones today. Did a sweep left and right, then moved the mic around a little, and did another left and right sweep. Sub-bass drop off is less on head, only drops about 3dB at 20Hz and 2dB at 30Hz. As expected, there is the classic STAX hump around 1k. Starts ramping up around 650Hz and peaks at about a 4-5dB increase at 1225Hz. I'm attributing the weirdness at 6-8k to odd mic effects. Harder to get accurate results up higher.

Overall it's impressively flat. Reinforces the idea that all these headphones need is a minor bass shelf and a few dB trimmed around 1k but that may be a taste thing, some people liking it, some not. That 7-8k dip is from the mic, as pointed out earlier in this thread.

Edit: Added Susvara comparison. Depending on where you volume match, Susvara has about 2dB more sub-bass. Could also say they have the same sub-bass but X9000 has more mid-bass. Either way it's accurate to say X9000 has a little bit of a mid-bass hump while Susvara is flat in the bass. Otherwise they track each other pretty closely but X9000 has a vastly smoother curve. Lots of weird sawtooth going on in the Susvara reading that's not in the X9000. I saw the sawtooth readings with Susvara when I measured with EARS too so it's definitely part of the headphone, not the measurement rigs/my head. Wonder what causes that.

Note, I did not touch the in ear mics at all when swapping from X9000 to Susvara.

Overall if someone showed me these two FR charts without me knowing which was which, I would pick the X9000 and just add a few db bass shelf 20-40Hz.

 

[SolarCetacean]

The Susvara's sawtooth response is due to resonances in the driver. A number of Hifiman planar headphones have those.

I wonder how Stax develops their headphones, because when I've reviewed the various measurements posted, it seems like the in-ear measurements show a much smoother response than the head-and-torso simulators (KEMAR/GRAS/B&K). And I tried using AutoEQ with a KEMAR measurement and the results of that were horrible. Manual sine sweeps that I've done also sound much closer to what the in-ear measurements say than what the HATS say (though I don't have in-ear mics for my own measurements). For whatever reason, the X9000 seems to confound the KEMAR and B&K rigs.

 

[gammi]

The Susvara vs X9000 comparison would up turning into a whole shootout between nine different high end headphones, so I made a new thread for it.

https://www.head-fi.org/threads/gammis-mega-headphone-shootout.972886/

 

[Svperstar]

I'm attributing the weirdness at 6-8k to odd mic effects.

Was gonna say that is one brutal 6k peak

 

[SolarCetacean]

I hear a small 6K peak, mainly because there is an audible dip in the response at 5.6K and 7K, so the 6K region exhibits a local peak. It's not a 6K peak jutting out of an otherwise smooth response, so it's not that harmful to the sound. To my ear, it adds a bit of crispness and bite to the sound.

 

[gammi]

Was gonna say that is one brutal 6k peak

Sorry, I should have been more specific, I don't think the 6k peak is the mic, rather the dip after it down to about 8k.
I had someone tell me previously that this mic has a dip at 8k and sure enough everything I tested had a dip around there.

While most headphones had a peak at 6k, it was less magnitude and varied more, implying actual differences in the headphones and not a mic issue.
Also I don't hear the 8k dip on a sweep.

 

[bnb110]

I had someone tell me previously that this mic has a dip at 8k and sure enough everything I tested had a dip around there.

Oh I mean removing this part would eliminate the 8 KHz dip, it's able to produce a fairly accurate result toward 20 KHz.

The “6 KHz peak” is part of the ear gain, if you measure speakers with this microphone you will get a similar peak.

 

[biscottino]

I see that little is said about the 700t, but while waiting for more, I wanted to give it a chance.
Once it has been run in for a long time, at least 2/300 hours, the amplifier begins to melt, the sound is not as lush as an all valve amplifier, but it is well balanced, clean, precise and articulate.
I might mistake it for a solid stage amp, although it never gets to the sharp side.
I feel that something more in terms of tone could have been given, and so armed with many screwdrivers, I opened it and proceeded to change the tubes, from the standard tung sol I put some Toshiba tubes, please note that they are Verdione GT from the 6sn7, and so after a fairly long bias calibration, I dedicated myself to listening with the x9000.
The sound has become delicate and creamy,
the Toshiba is not a valve dedicated to detail, but it has a warm tone, like Mullard so to speak, this touch of warmth and humami, I dare say, has dramatically changed the way the x9000 plays, which now seems heavier and fuller in the notes, filling the spaces better, the scene in the center is fuller and there is more impact in the notes, especially the low ones, they are lower and fuller.
For now I can say that I like it, it's a new point of view, I've had the Valkyria for two years, and in terms of piano reproduction it has no equal, with this valve the timbre of the x9000 is very similar to it, the notes seem more suspended in the spaces , yes, I know, it's just an equalization, but it's very effective, I have other versions of 6sn7 and will definitely try something else.

 

[icnaiwhimor1975]

Surprised to see MiniDSP EARS can get such smooth response

*DSP has many implementations and some good software ones give good results

 

[gammi]

Oh I mean removing this part would eliminate the 8 KHz dip, it's able to produce a fairly accurate result toward 20 KHz.

The “6 KHz peak” is part of the ear gain, if you measure speakers with this microphone you will get a similar peak.

Yeah, I understood, I just didn't remove that part so it still has the 8K dip and I was letting everyone know that the mic (as I used it) creates a dip around 8k.
Didn't feel like cutting that off just yet, then the issue becomes getting it to stay seated in my ear. I suppose the best solution for that is some sticky putty on the backside.

 

[xshen79]

I found people sometimes having different opinions regarding same headphone matching same amps. Considering STAXs are more sensitive to sources than other headphones, might be better to include source when sharing your thoughts on synergy between any amp and headphone? Just some 2 cents of mine.

 

[gammi]

I found people sometimes having different opinions regarding same headphone matching same amps. Considering STAXs are more sensitive to sources than other headphones, might be better to include source when sharing your thoughts on synergy between any amp and headphone? Just some 2 cents of mine.

Novem for X9000. Bliss for Susvara.

I did mention the amps earlier but amps won't affect FR graphs much anyway unless it's literally not powerful enough to drive the headphone. Any modern amp has a frequency response of like +/-1dB for the whole range. Amp differences show up elsewhere in things like transients.

Honestly I think it's better not to mention the amps. That only serves to open the door to discussions about how the amps are coloring the sound (when they aren't)

 

[xshen79]

I did mention the amps earlier but amps won't affect FR graphs much anyway unless it's literally not powerful enough to drive the headphone. Any modern amp has a frequency response of like +/-1dB for the whole range. Amp differences show up elsewhere in things like transients.

Honestly I think it's better not to mention the amps. That only serves to open the door to discussions about how the amps are coloring the sound (when they aren't)

Just trying to understand why sometimes people describing same combo with different featuring.

 

[MrHaelscheir]

Oh I mean removing this part would eliminate the 8 KHz dip, it's able to produce a fairly accurate result toward 20 KHz.

The “6 KHz peak” is part of the ear gain, if you measure speakers with this microphone you will get a similar peak.

I would note that I also get that dip/null in my own in-ear measurements using mics (see https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152) and https://audiosciencereview.com/foru...phone-measurements.18451/page-61#post-1956895 (post #1,210)) that don't have that encircled concha mount, so I'd say something else about the mic placement relative to the ear is causing that cancellation; I haven't seen how you position those Sound Professionals mics with those concha mounts removed. I have yet to fashion probe mics and see whether those nulls are still present near the eardrum (or haven't myself encountered literature on at-eardrum HRTF measurements), though IIRC, they may serve a purpose in sound localization. I don't know the details of why the GRAS ears have similar nulls while the B&K ears don't.

 

[bnb110]

I would note that I also get that dip/null in my own in-ear measurements using mics (see https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152) and https://audiosciencereview.com/foru...phone-measurements.18451/page-61#post-1956895 (post #1,210)) that don't have that encircled concha mount, so I'd say something else about the mic placement relative to the ear is causing that cancellation; I haven't seen how you position those Sound Professionals mics with those concha mounts removed. I have yet to fashion probe mics and see whether those nulls are still present near the eardrum (or haven't myself encountered literature on at-eardrum HRTF measurements), though IIRC, they may serve a purpose in sound localization. I don't know the details of why the GRAS ears have similar nulls while the B&K ears don't.

Here is what I did to the mic and how I place it.
The removement can't be undone but I think X9000 owners is not likely care about an 100$ mic.
I can't take a good photo of my ear so I use an artifical ear for example.

The raw measurement of X9000 (DAC: V380 AMP: 727A)

The 5.5KHz dip and 9.5KHZ does exist when I run a manual sweep without the mic.
I think the 9.5KHz null depends on ear, with GRAS X-9000 only has the 5.5KHz dip.

The raw measurement of Susvara (DAC/AMP: V380)

No dip/null except 2KHz (Hifiman House Sound).
I'm not sure about other mics but MS-TFB is able to measure the Susvara without nulls after mount removed.