ETA Genesis Graphene (Gen-G) Measurement and Analyses

Disclaimer: On top of usual attention that must be paid to any measurement, I recommend readers be more cautious in translating my results this time around. Take any statements with a pinch of salt. Cause.. I admit I am very enthusiastic and biased toward ETA Genesis Graphene (aka Gen-G or GG) for their performances both subjectively and objectively.

Aside: Whenever I upload my measurement, I'd always emphasize not to take results too seriously. There may always be a lot of error-inducing traps in the 'objectivist' realm. For example, I may measure/present wrongly (e.g., I incorrectly over-report LCD-X's failure which later turned out to be an amplifier problem). Right interpretation may be often dragged by our intuition. Minidsp EARS may have its own habit. Even when everything is done right, subjective perception may not be associated with measured patterns. Etc. Etc. There are really many BS scenarios we may be fooled.​

Here’s the plan to post my analyses. In the next series of posts, a set of measurement results with my separate notes will be dropped. Then several comparative FR plots will be added to give you some idea how GG can sit in the product space, possibly with my comparative evaluations. Finally once I feel confident, I will provide my thoughts on subjective and unmeasurable performance.

One thing I'd like to make clear is the revision status. GG analyzed in this thread is version 1.1 (as per ETA). But it seems they first hand-made tuning parts then later changed to machine-cut. Mine originally had parts manufactured in a very early stage and pads kinda deformed asymmetrically across channels. Evan kindly sent me the newest material and brand new pads. I took measurements before and after applying new material. Throughout the series of my posts, I will call 'prototype' to indicate the former while 'production' refers to the latter.

As all GG owners already know, GG can be tunable on users' end via foams, plugs, and fabric rolls. In my preliminary analyses, I consistently used the stock setting as recommended (white foam ring + no plug inserted + two large holes and one small hole open + pads attached with thicker place on top). If I measure a different configuration, exact deviation from the stock one will be specified in the post explicitly.

Ok, enough introduction. Let's start.

Links to Relevant Threads

• ETA Gen-G v1.1 impressions: https://www.superbestaudiofriends.org/index.php?threads/eta-genesis-1-1-impressions.11249/
• ETA Gen-G old impressions: https://www.superbestaudiofriends.org/index.php?threads/eta-genesis-g-p-impressions-thread.10892/
• ETA Genesis and its mod general discussion: https://www.superbestaudiofriends.o...ne-intro-discussion-thread-3dp-version.10517/

 

1. Frequency Response



Slight channel imbalances found but nothing too serious. It was not that easy to get symmetric contacts on EARS for both channels. Mild left-right differences on lower treble is still hearable but not very bothering. The results indicate they managed to hold a more or less similar sound pressure level. Bass differences are very likely caused by imperfect seals (aka air gap behavior). And from what I observed and heard, and assuming no driver-level imbalance issue, I am suspecting the resonance frequency is located around 50 to 60hz (free air resonance point maybe closer to the former).



As I declared in the OP, I replaced pads, grills and tuning parts with fresh new ones. But there didn’t exist any measurable differences between the two ( v1.1 prototype vs production), which isn’t quite consistent with what I heard myself. I tested both versions back and forth multiple times. All objectively not too different (well within run to run variation) but subjectively discernible.

Major subjective differences not shown in the graph include (1) reduced LR channel imbalances (I did hear more imbalances with proto); (2) more airs and top octave extension with the production version; (3) tad thicker sounding with the production version.

There is a peak around 2khz followed by 2.5khz dip. This pattern repeated up to 5khz. None of the peaks or dips were shitty. Not at all. Never. They rather excite “fun” components in modern recording. Note that I have a firm belief headphones DO NOT HAVE TO have flat compensated response on trebles unless compensation curve is very accurate and personalized. GG’s treble is voiced very tastefully and feels like a good sweet spot between neutrality and engagement. I really like how GG’s treble response gets along with female vocal tracks (I mostly hear modern or latest tracks these days). A little laid-back and romantic reproduction. But it maintains a proper level of energy, clarity, and sharpness. Depending on the recording style and preference, mileage may vary though.

Honestly speaking, I am more impressed with GG’s upper mid to lower treble than with its (dominantly well-received) bass performance.

I usually prefer a narrow dip around 8-9khz in SBAF-compensated results. A shy dip is observed at 8.3khz with GG. The amplitude of the dip is a little less than I like. But no subjective dissatisfaction. Perhaps that’s just a proper amount for GG’s overall dark and downward tonal balance. But I digress for now.

 

2. Cumulative Spectral Density

Above is what I said in my initial impression. And I think the waterfall plot confirms my subjective discrepancy. Sustained decay observed a little below 2khz -- likely 1.6-ish hz. Not too high level (-30db to the signal start) but meaningfully higher than neighboring frequencies. These seem to make 2k-peak less protrude and mix well with the upper midrange.

3. Distortions




Considering my measurements done at 95db SPL@300hz, the drivers seem to be under good control all over the spectrum. Bass distortion looks good down to 40hz (most dynamic headphones can be a lot worse). I am pretty sure sub-bass harmonics would reduce well below 1% once I decrease the sine sweep level a bit.

It’s interesting to see a fairly consistent level of the 2nd harmonics. Majority of good dynamic headphones present rather extreme variation on both ends: very high on lows but beyond a certain point extremely low (e.g., HD650). I’m guessing consistency like GG may have something to do with better and more cohesive subjective experience. Needs more investigation and thoughts though.

I will post comparative plots and evaluation in other days.

 

Adding a couple of comparative plots that I believe indicate where GG's tonal balance sits in the product space.



Not surprisingly, ESX900 and GG are pretty similar in many aspects. They resemble each other in timbre and texture regardless of materialistic difference. If anyone likes one of them, then it's very possible to like the other, too.

But fortunately (to me), they seem to serve for different tastes. As the graph suggests, GG has significantly higher energy around 2khz and treble level is a lot closer to what I consider neutral. Considering both have already boosted bass level, GG works for my use cases much better in that it does not sacrifice clarity, snappiness, and airiness. Overall I found GG way more engaging to my liking.




Now, let's compare to HD650 (to be specific, the exact comparison done with 6xx and KISS-mod applied). This can be another interesting comparison. If we can ignore bass below 200hz, the rest of frequency response is rather similar than different between the two... at least by the amount of energy.

One of the reasons I'm sticking to SBAF EARS compensation v1.0 is this compensation curve exactly illustrates where I am uncomfortable with HD650/6xx. There show a wider peak around 3khz and another narrower peak at 5khz. Both never too substantial (<+5db) but I personally feel annoyances with those parts. Time to time I hear tracks too bright, too edgy, or timbre not quite right. These issues were not alleviated with Kiss-mod. JAR 650 addressed them to some extent though.

In this comparative context, GG's highs feel like 'fully fixed' 650. And 2k peak as bonus (as you may guess, I like this peak with music I hear). 650's (occasional) annoyance and offensiveness are completely gone in GG, which may explain why I am super enthusiastic this time.

To someone else who have different tonal tastes, HD650 can be more spot-on, and GG may feel energy-less or muted. So, YMMV.

-----

If my statements look like GG takes the best parts from both ESX-900 and HD650, you're not wrong. That's what I exactly think GG is in tonality.

 

Subjective Evaluation

Pros

• Enjoying LCD or HEDD, I thought I was rather insensitive to headphones weight. But later I realized that’s simply because lightweight headphones sacrifice too many things (bass, pads comfort, dynamics, etc) to tolerate. GG was the first pair of headphones to me proving lightweight (to be specific <300g) cans sound this good. They show a much higher level of perfection than something like Ether 2.
• I really like how GG contrasts loud vs quiet passages. They also tell small nuances/details apart from other chunks of sonic elements. My DAC (Soekris 2541) has a great ability to drop all unnecessary microstuffs to feel like a super great contrast ratio. GG driven by Jotunheim 2 seems to take advantage of full potential from signals from the converter. Extremely synergistic.
• Graphene material (+strong magnet) behaves like a balanced sweet spot between uber metal and biocellulose drivers. Transient 90-95% feels like Be. And yet it mimics the romantic timbre of BC or paper drivers to some extent. Not to mention the ETA team could make use of the full potential of graphene drivers. Overall sound is technical AND attractive, which don’t often come together.
  

Cons

• This can be a matter of preference. But the semi-closed nature of GG’s construction makes it sound not fully open. For a whim, I tried out blue foams and sushi removals which open up a bit (to much less extent than I thought) but come with harsher highs and too much energy on trebles. I immediately had to restore every change I made. What I want is something like Clear Mg (dark, downtilt, yet extensively open sounding) , which I don’t seem to achieve with GG’s airflow restriction. Gen-G Supra seems to have a much better structure than GG Over-Ear for that respect though.
• When listening to very low notes, GG’s subbass (particularly below 35hz) sometimes sounds blurry and loses articulation. Unsure if that’s GG’s inherent trait or amplification failure. Both Clear mg and Verites are less problematic to me on this front. Not a showbreaker at all. I’m just splitting hairs. This might be a hidden cost for lightweightness. When Mjo1 arrives here, I will test this again to see if there’s any improvement.

Please note that I tried to make this post as concise as possible. You can assume anything I didn’t list NO LESS THAN “very good to me”. I also decided not to list things already reported well.