As one who listens to and attends lots of symphonic music, I would note that it would be good to make sure that the X9000 seals well on your own head since the circular cups are so large and are not particularly contoured. For my head, no matter the positioning, I was feeling uncomfortable gaps at the front and back of the pads incurring increased pressure felt on the top and bottom of the pads, this likewise incurring a measurable decrease in sub-bass which might be needed for visceral orchestral bass or fullness. That was the deal-breaker for me.
As one who listens to and attends lots of symphonic music, I would note that it would be good to make sure that the X9000 seals well on your own head since the circular cups are so large and are not particularly contoured. For my head, no matter the positioning, I was feeling uncomfortable gaps at the front and back of the pads incurring increased pressure felt on the top and bottom of the pads, this likewise incurring a measurable decrease in sub-bass which might be needed for visceral orchestral bass or fullness. That was the deal-breaker for me.
Just want to add that while a broken seal does decrease sub-bass, it also increases mid-bass.
Can test this by lifting up the bottom of the X9000 by the cable connectors and hearing bass increase.
So it depends on the sub-bass content of the track if this is a good thing or not.
I own both of those and now that I have an "end game" amp debating picking up the X9000 and seeing if I like it, hearing people say it is bass shy or deficient scares me off some.
https://autoeq.app is a great site to download EQ profiles and it looks like they have measured all the major Stax besides the X9000.
Does anyone have a good X9000 EQ profile for EqualizerAPO?
I could just dial in +db in the bass region on my RME ADI-2 but its not the same as an EQ profile built off of measurements.
I own both of those and now that I have an "end game" amp debating picking up the X9000 and seeing if I like it, hearing people say it is bass shy or deficient scares me off some.
https://autoeq.app is a great site to download EQ profiles and it looks like they have measured all the major Stax besides the X9000.
Does anyone have a good X9000 EQ profile for EqualizerAPO?
I could just dial in +db in the bass region on my RME ADI-2 but its not the same as an EQ profile built off of measurements.
The post I linked on the previous page has an EQ profile based on the chart in the post above this one.
Also I disagree with those saying it's bass deficient. As shown in the above chart, it does roll off below 40Hz, but not much music goes that low anyway. 40-80Hz it has mildly more bass than Susvara. Also the quality of the bass is excellent. For example, I don't think I've heard bass guitar sound better on anything vs the X9000, to my ear. Strings in general sound superb on the X9000.
All depends on the type of music you listen to. The roll off might not matter much at all. Or if electronic, modern pop or similiar is in your listening rotation it could be a deal breaker. But that is why I keep multiple headphones for different music.
I've played around with a Harman AutoEQ profile courtesy of the X9000 measurement from Kyle at Den-Fi (@Ishcabible I think). Found here. That page has an AutoEQ function to EQ a given curve to a selected target.
Preamp: -9.3 dB
Filter 1: ON PK Fc 32 Hz Gain 9.8 dB Q 0.500
Filter 2: ON PK Fc 35 Hz Gain -1.3 dB Q 2.000
Filter 3: ON PK Fc 53 Hz Gain 1.2 dB Q 0.800
Filter 4: ON PK Fc 67 Hz Gain -2.5 dB Q 2.000
Filter 5: ON PK Fc 90 Hz Gain 0.7 dB Q 1.500
Filter 6: ON PK Fc 410 Hz Gain 0.9 dB Q 1.900
Filter 7: ON PK Fc 680 Hz Gain 1.6 dB Q 2.000
Filter 8: ON PK Fc 1000 Hz Gain -4.9 dB Q 1.700
Filter 9: ON PK Fc 3000 Hz Gain 5.1 dB Q 1.400
Filter 10: ON PK Fc 5300 Hz Gain 3.1 dB Q 2.000
Filter 11: ON PK Fc 9500 Hz Gain 6.9 dB Q 2.000
Filter 12: ON PK Fc 10000 Hz Gain -4.1 dB Q 0.500
Preamp: -6.3 dB
Filter 1: ON PK Fc 20 Hz Gain 6.1 dB Q 0.500
Filter 2: ON PK Fc 52 Hz Gain 2.3 dB Q 1.600
Filter 3: ON PK Fc 64 Hz Gain -2.3 dB Q 2.000
Filter 4: ON PK Fc 140 Hz Gain 1.2 dB Q 0.500
Filter 5: ON PK Fc 630 Hz Gain 1.3 dB Q 1.700
Filter 6: ON PK Fc 860 Hz Gain -1.0 dB Q 2.000
Filter 7: ON PK Fc 1100 Hz Gain -5.2 dB Q 2.000
Filter 8: ON PK Fc 2700 Hz Gain 6.7 dB Q 1.300
Filter 9: ON PK Fc 4400 Hz Gain -1.5 dB Q 3.200
Filter 10: ON PK Fc 5300 Hz Gain 4.1 dB Q 2.000
Filter 11: ON PK Fc 6400 Hz Gain -4.1 dB Q 1.000
Filter 12: ON PK Fc 6700 Hz Gain -3.1 dB Q 2.000
Filter 13: ON PK Fc 9200 Hz Gain 6.6 dB Q 2.000
Note that I use custom pre-amp settings in order to volume-match the two presets and the Neutral preset which has nothing but a pre-amp. The idea is to make sure that the overall volume of the music doesn't change much when switching between presets, otherwise you aren't actually evaluating the change in tonality, but the tonality and the change in volume. I generally try to volume-match the midrange, and let the bass and treble fall where they may. This results in the Harman preset being bassier than the others.
Neutral preamp: -10 dB
Tilted DF preamp: -11.5 dB
Harman preamp: -11 dB
I'm hard-pressed to call either EQ profile an improvement. All of them mess up one or more of the following: spatial layering, soundstage, openness, and general sound texture. I generally find myself going back to the stock tonality.
Tilted DF is bright as heck! I thought the same with a prior tilted DF target based on the 5128 rig. If this is with the 10 dB tilt, I don't even want to know what plain DF sounds like. I don't find the X9000 to be bright in the treble; it's pretty neutral to me in a sine sweep and sometimes verges on being dark. The DF preset injects a ton of mid-treble into the sound which makes the sound "hashy", "shimmery", or "crispy" without adding any meaningful detail. It also overboosts the presence region which makes vocals shouty and uncomfortable. It introduces a bit more sibilance compared to the stock tonality.
Harman 2018 is better in that regard. It's bassier, with some 3dB more bass boost than the DF preset and it doesn't boost the presence region or the mid-treble as much. It still makes the sound a bit hashy due to the presence region boost.
Now for the problems that these presets introduce:
1. The soundstage gets smaller. The X9000's marquee ability (with the right chain) is the ability to present spatial layering between near and far. Both presets bring the further sound elements closer in. However, it also doesn't really push the vocals further out. They do get a bit further away spatially, but not by much, and not significantly. Thus, the overall sound image simply gets more constricted.
2. The openness of the sound is lost. The stock X9000 has a "radiative" spatial quality to notes, like they dissipate into the soundstage. I highly value that aspect of sound. Both presets make sounds more "directed", like they're being shot directly at me. The spatial "edges" of notes become harder and they restrict how the note blooms into the overall soundstage. Combine that with the hash from the elevated presence region and mid-treble and the reduction in soundstage size and the sound becomes crowded, constricted, and closed in. The added treble makes it so that there's this constant "shshshsh" in the background whenever a track has a lot of treble energy and the boosted bass, particularly with the Harman preset, gives a constant "wooom" in the background from the sub-bass. I don't think either of those things add anything to the sound. They make the sound busier, and some people can interpret that as more detail, but I find it to be mere hash and noise.
3. Shouty vocals. To add additional insult, both presets make vocals shoutier. I realize I'm particularly sensitive to shoutiness in vocals, but vocals aren't much further back, and now they're shouty too due to too much ear gain. They're not comfortable to listen to particularly in combination with the spatial constriction.
I found both presets to be fatiguing to listen to. It was so relaxing to come back to the stock tonality. The sound was open, layered, and relaxed. Sure, vocals are too forward spatially, but they're rich and not shouty(!!!). No loss of detail to my ears. Maybe just a plain bass boost would work best, but I generally have no issue with the stock bass on the X9000, and I don't run the most powerful amps either. There and back again.
Edit to add: I'm increasingly convinced that the forward angling of the guard mesh has some effects that aren't easily seen in the FR graph. It seems that vocals are just spatially forward on the X9000 regardless of my attempts to EQ out that aspect.
I own both of those and now that I have an "end game" amp debating picking up the X9000 and seeing if I like it, hearing people say it is bass shy or deficient scares me off some.
https://autoeq.app is a great site to download EQ profiles and it looks like they have measured all the major Stax besides the X9000.
Does anyone have a good X9000 EQ profile for EqualizerAPO?
I could just dial in +db in the bass region on my RME ADI-2 but its not the same as an EQ profile built off of measurements.
@SolarCetacean Have you had a chance to try some of my filters in isolation? I also would avoid using preset EQ filters above 5 kHz since they may be correcting peaks or nulls of greatly different depth and location than with your own ears, possibly EQing in a peak where you already have a peak and thus incurring that overamplified treble noise, hence the need to listen to sine sweeps to manually smooth out that response. Those two presets also looked to be adding rather much ear gain where my measurements found there to be sufficient amounts; Royal Navy's EQ profile seemed more reasonable other than in the sub-bass. Does EQing down the 1 kHz to 2 kHz region affect the vocal layering? I had particularly listed a few filters for completely flattening the X9000's midrange bump at least for my ears:
Filter: ON PK Fc 626.0 Hz Gain 1.20 dB Q 10.000
Filter: ON PK Fc 763.0 Hz Gain -1.40 dB Q 5.000
Filter: ON PK Fc 1132 Hz Gain -3.50 dB Q 2.000
Filter: ON PK Fc 1667 Hz Gain -1.30 dB Q 5.000
Filter: ON PK Fc 2809 Hz Gain 2.30 dB Q 2.000
For a baseline of how my "V3.1 PEQ" compared to Harman, I did have a chance to measure the DCA Expanse, Stealth, and E3 at a shop, though the graphs have to be kept confidential. They do differ a decent bit from one another at least for my ears. See Figure 13 of https://www.head-fi.org/threads/rec...-virtualization.890719/page-121#post-18027627 (post #1,812) for how my V3.1 PEQ (the gold trace) compares to my personal neutral speaker response (the red and teal traces): the DCA E3 is probably the closest one to my personal neutral speaker response, but without the 3 kHz dip, then more 7 kHz to 8 kHz, and then a 5 dB Harman bass shelf below 150 Hz.
Otherwise, even if keeping the rest of the frequency response unEQed, I'd say it would be useful to seek out any peaks above 5 kHz (e.g. for my ears, the X9000 had a notable 6.5 kHz peak) to possibly yield an even smoother and cleaner sound, whereby I had detailed that process in the "EQing journey" parts of https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17549413 (post #4,665). I realize that I was terribly verbose then, but one basically listens to a sine sweep and looks for salient increases in the volume of the pure tone, whereby depending on the width and loudness of that peak, you would position a negative gain peaking filter there and adjust it or add more filters until the volume of the sine sweep sounds smooth through that band. Pink noise can also reveal the location of some peaks, whereby you can add a positive gain filter and move its center frequency until it sounds like it is amplifying the specific peak that you heard, your then using that filter to EQ that peak down. EQing a peak down may cause another peak to be audible, but don't overdo these adjustments; in-ear mics would be the most accurate guide and would avoid overcompensation.
@SolarCetacean Have you had a chance to try some of my filters in isolation? I also would avoid using preset EQ filters above 5 kHz since they may be correcting peaks or nulls of greatly different depth and location than with your own ears, possibly EQing in a peak where you already have a peak and thus incurring that overamplified treble noise, hence the need to listen to sine sweeps to manually smooth out that response. Those two presets also looked to be adding rather much ear gain where my measurements found there to be sufficient amounts; Royal Navy's EQ profile seemed more reasonable other than in the sub-bass. Does EQing down the 1 kHz to 2 kHz region affect the vocal layering? I had particularly listed a few filters for completely flattening the X9000's midrange bump at least for my ears:
Filter: ON PK Fc 626.0 Hz Gain 1.20 dB Q 10.000
Filter: ON PK Fc 763.0 Hz Gain -1.40 dB Q 5.000
Filter: ON PK Fc 1132 Hz Gain -3.50 dB Q 2.000
Filter: ON PK Fc 1667 Hz Gain -1.30 dB Q 5.000
Filter: ON PK Fc 2809 Hz Gain 2.30 dB Q 2.000
For a baseline of how my "V3.1 PEQ" compared to Harman, I did have a chance to measure the DCA Expanse, Stealth, and E3 at a shop, though the graphs have to be kept confidential. They do differ a decent bit from one another at least for my ears. See Figure 13 of https://www.head-fi.org/threads/rec...-virtualization.890719/page-121#post-18027627 (post #1,812) for how my V3.1 PEQ (the gold trace) compares to my personal neutral speaker response (the red and teal traces): the DCA E3 is probably the closest one to my personal neutral speaker response, but without the 3 kHz dip, then more 7 kHz to 8 kHz, and then a 5 dB Harman bass shelf below 150 Hz.
Otherwise, even if keeping the rest of the frequency response unEQed, I'd say it would be useful to seek out any peaks above 5 kHz (e.g. for my ears, the X9000 had a notable 6.5 kHz peak) to possibly yield an even smoother and cleaner sound, whereby I had detailed that process in the "EQing journey" parts of https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17549413 (post #4,665). I realize that I was terribly verbose then, but one basically listens to a sine sweep and looks for salient increases in the volume of the pure tone, whereby depending on the width and loudness of that peak, you would position a negative gain peaking filter there and adjust it or add more filters until the volume of the sine sweep sounds smooth through that band. Pink noise can also reveal the location of some peaks, whereby you can add a positive gain filter and move its center frequency until it sounds like it is amplifying the specific peak that you heard, your then using that filter to EQ that peak down. EQing a peak down may cause another peak to be audible, but don't overdo these adjustments; in-ear mics would be the most accurate guide and would avoid overcompensation.
I explored a bunch of EQ profiles from various users and for my ear, I think bnb110's simple filter of -3 dB at 1100 Hz, Q=3 was sufficient for me to get rid of the midrange bump. I ran a bunch of sine sweeps and the problem is that I can't really hear any major issues. I smoothed out the midrange with two filters: -3 dB at 1100 Hz, Q=3 and +1 dB at 2200 Hz, Q=2.5.
Small changes, but to me, those were enough to linearize the sine sweep between 500 Hz and 4 KHz. The KEMAR measurement from Den-Fi showed a colossal midrange hump, but I just can't hear it on a sine sweep. The treble was also very linear to me. A small dip around 5.5K (which might be pinna-related as I hear the same dip on some other headphones), a small 6.6K peak, then a dip back down around 7.2K and then the usual perceived rise starting around 8K and continuing into the upper treble which has some small audible wiggles in the response. Nothing major sticks out as a problem, unlike many other headphones, and so it's quite hard for me to EQ by ear since it sounds very linear already.
The two filters I use didn't really change the vocal layering, but it slightly shrank the width of the vocals, a bit like folding the edges of the sound away from me. But in many songs, the difference is nearly imperceptible. The difference is reliably audible in a sine sweep, but in music it's subtle.
Preamp: -9.3 dB
Filter 1: ON PK Fc 32 Hz Gain 9.8 dB Q 0.500
Filter 2: ON PK Fc 35 Hz Gain -1.3 dB Q 2.000
Filter 3: ON PK Fc 53 Hz Gain 1.2 dB Q 0.800
Filter 4: ON PK Fc 67 Hz Gain -2.5 dB Q 2.000
Filter 5: ON PK Fc 90 Hz Gain 0.7 dB Q 1.500
Filter 6: ON PK Fc 410 Hz Gain 0.9 dB Q 1.900
Filter 7: ON PK Fc 680 Hz Gain 1.6 dB Q 2.000
Filter 8: ON PK Fc 1000 Hz Gain -4.9 dB Q 1.700
Filter 9: ON PK Fc 3000 Hz Gain 5.1 dB Q 1.400
Filter 10: ON PK Fc 5300 Hz Gain 3.1 dB Q 2.000
Filter 11: ON PK Fc 9500 Hz Gain 6.9 dB Q 2.000
Filter 12: ON PK Fc 10000 Hz Gain -4.1 dB Q 0.500
Presets generated by the autoeq project are generally problematic, it could save some poorly tuned budget headphones, but it doesn't make sense for anything above middle-end.
Filter 1-5 are stupid, the bass response highly depends on the seal of individual head, there is no universal equalizing for bass, beside X9000 is a very hard to drive headphone, apply -10 db preamp will eaisily destory details and dynamics on high dynamic tracks like most classic.
Filter 10,11 are problematic, it tried to fix two dip measured on the GRAS rig but it may not exists on individual head, even exists it's likely at different frequency due to different ear shape, unit variance, slience revision, don't mess treble unless you can measure in ear response of the exact unit you have.
Filter 12 is beyond stupid, it tried to fix a measurement rig artifact
If someone said they can make your speakers sound better by doing room correction based on a measurement measured in a different room, people will not be fooled.
The person wrote autoeq should learn how measurements work, if he want to get a universe filter to match treble to some target the only solution is use a very wide filter(q <1.0), and it needs a better algorithm, use 5 filter to adjust bass is horrible.
The stock tuning of the X9000 is very balanced, it doesn't really need EQ.
For people want to get their hands dirty, I suggest these locations (I only did 1KHz)
Code:
around 40Hz, use a LF to get more subbass, for example +6dB Q1.0, be sure you have a powerful amp it needs corresponding preamp to avoid clipping
around 1KHz -2~4dB, traditional stax mid bump, reduce it can make the sound less forward
around 4Khz, the ear gain, reduce it a bit will make the sound less shouty, reduce too much will result compression
Presets generated by the autoeq project are generally problematic, it could save some poorly tuned budget headphones, but it doesn't make sense for anything above middle-end.
Filter 1-5 are stupid, the bass response highly depends on the seal of individual head, there is no universal equalizing for bass, beside X9000 is a very hard to drive headphone, apply -10 db preamp will eaisily destory details and dynamics on high dynamic tracks like most classic.
Filter 10,11 are problematic, it tried to fix two dip measured on the GRAS rig but it may not exists on individual head, even exists it's likely at different frequency due to different ear shape, unit variance, slience revision, don't mess treble unless you can measure in ear response of the exact unit you have.
Filter 12 is beyond stupid, it tried to fix a measurement rig artifact
If someone said they can make your speakers sound better by doing room correction based on a measurement measured in a different room, people will not be fooled.
The person wrote autoeq should learn how measurements work, if he want to get a universe filter to match treble to some target the only solution is use a very wide filter(q <1.0), and it needs a better algorithm, use 5 filter to adjust bass is horrible.
The stock tuning of the X9000 is very balanced, it doesn't really need EQ.
For people want to get their hands dirty, I suggest these locations (I only did 1KHz)
Code:
around 40Hz, use a LF to get more subbass, for example +6dB Q1.0, be sure you have a powerful amp it needs corresponding preamp to avoid clipping
around 1KHz -2~4dB, traditional stax mid bump, reduce it can make the sound less forward
around 4Khz, the ear gain, reduce it a bit will make the sound less shouty, reduce too much will result compression
Note that though in a full-scale sine sweep, having the EQ profile have areas greater than 0 dBFS could cause clipping or volume limiting, it is best to use an oscilloscope plugin like in the Reaper DAW or any other way to check the actual amplitude peaks of the EQed signal with a very busy and high amplitude test track to see if you can actually afford to have less negative gain preamp and thus not lose out too much on your DAC's dynamic range, likewise allowing your amp to operate at a lower gain; e.g. I used to have to run my FiiO K9 Pro ESS at medium to high gain after applying my EQ, but after seeing that even my most demanding tracks could use a lot more digital gain before incurring clipping (since my EQ was largely boosting the ear gain which effectively greatly reduced the bass and midrange and hence the signal's amplitude), I can now operate it on low gain. I personally use the following track as my upper benchmark for whether the most demanding music would clip when EQed (VOLUME WARNING: don't worry, the piece isn't entirely like that):
I use this track as a worst case clipping revealer to maximize the digital dynamic range at the major stages of my DSP chain.
Figure 1: Full-scale madness. It looks clipped without sounding clipped until it is actually clipped.
For Equalizer APO, to tap into its digital output, since it applies EQ at the device level, I needed to configure my MOTU M2 to receive such then use Reaper ReaRoute and Voicemeeter to send the EQed loopback signal into my Reaper DAW for analysis; the option I didn't want to use since I don't always route things through Voicemeeter into Reaper DSP is to instead configure Equalizer APO to EQ a Voicemeeter input or output before sending that signal into Reaper just for analysis, but I in my DSP application happen to want to apply Equalizer APO after applying the binaural DSP.
As one who listens to and attends lots of symphonic music, I would note that it would be good to make sure that the X9000 seals well on your own head since the circular cups are so large and are not particularly contoured. For my head, no matter the positioning, I was feeling uncomfortable gaps at the front and back of the pads incurring increased pressure felt on the top and bottom of the pads, this likewise incurring a measurable decrease in sub-bass which might be needed for visceral orchestral bass or fullness. That was the deal-breaker for me.
I've always wondered if the leather-free LCD2 pads from Audeze would fit the X9000, those are the best earpads I have ever worn, if I owned a X9000 I would for sure see if I could stick them on. They greatly increase the base on the LCD-2 Fazor.
The stock tuning of the X9000 is very balanced, it doesn't really need EQ.
For people want to get their hands dirty, I suggest these locations (I only did 1KHz)
Code:
around 40Hz, use a LF to get more subbass, for example +6dB Q1.0, be sure you have a powerful amp it needs corresponding preamp to avoid clipping
around 1KHz -2~4dB, traditional stax mid bump, reduce it can make the sound less forward
around 4Khz, the ear gain, reduce it a bit will make the sound less shouty, reduce too much will result compression
I think these are great suggestions except I wouldn't do the 4k one (X9000 actually has less at 4k than Susvara for reference) and the 1k one is personal preference if you like the forwardness or not. On mine I only do the 40Hz bass shelf but I'd like to explore adjusting around 1k more since that's really the only area where X9000 deviates.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
