GR Research Speaker Upgrade Review (Sierra-2EX V2)

[totti1965]

Here is my take on the EQ.
Please report your findings, positive or negative!

For the score rational your journey starts here
Explanation for the sub score
The following EQs are “anechoic” EQs to get the speaker right before room integration.
If you able to implement these EQs you must add EQ at LF for room integration, that is usually not optional… see hints there.

The raw data with corrected ER and PIR:

Score no EQ: 5.4
With Sub: 7.5

Spinorama with no EQ:

• A bit rough
• Decent design within the idiosyncrasies of ribbon tweeters

View attachment 530008
Directivity:

Better stay at tweeter height, precelit at the tweeter middle point.
Horizontally, better toe-in the speakers by 10/15deg and have the axis crossing in front of the listening location, might help dosing the upper range. Explanation here.
Typical ribbon vertical directivity issue

View attachment 530015
View attachment 530022
EQ design:
I have generated one EQ. The APO config file is attached.

• The first one, labelled, LW is targeted at making the LW flat
• The second, labelled Score, starts with the first one and adds the score as an optimization variable (here).
• The EQs are designed in the context of regular stereo use i.e. domestic environment, no warranty is provided for a near field use in a studio environment although the LW might be better suited for this purpose.
• One can model the EQ with Vituixcad by using the DSP "Generic" setting with 96000Hz sampling rate.
• LW and Score are too close to separate
• One can experiment with changing the second biquad gain:
  Preamp: -1.60 dB
  Filter 2: ON PK Fc 87.2 Hz Gain -4.25 dB Q 1.06
  to:
  Preamp: -2.30 dB
  Filter 2: ON PK Fc 87.2 Hz Gain -2.50 dB Q 1.06 for a similar tonal balance vs. no EQ: the penalty is just 0.1 point across the scores

Score no EQ: 5.4
With Sub: 7.5

Score EQ Score: 6.7
with sub: 8.7

Ascend Sierra-2EX V2 APO Score EQ 96000Hz
May052026-114755

Preamp: -1.60 dB

Filter 1: ON HPQ Fc 47.2 Hz Gain 0.00 dB Q 1.39
Filter 2: ON PK Fc 87.2 Hz Gain -4.25 dB Q 1.06
Filter 3: ON PK Fc 419.2 Hz Gain -1.95 dB Q 3.19
Filter 4: ON PK Fc 747.2 Hz Gain -2.00 dB Q 4.31
Filter 5: ON PK Fc 916.6 Hz Gain -1.13 dB Q 5.99
Filter 6: ON PK Fc 992.5 Hz Gain 1.46 dB Q 1.38
Filter 7: ON PK Fc 1293.7 Hz Gain -3.21 dB Q 5.95
Filter 8: ON PK Fc 3935.2 Hz Gain -1.57 dB Q 1.56
Filter 9: ON PK Fc 8540.9 Hz Gain -1.80 dB Q 3.63
Filter 10: ON PK Fc 14105.9 Hz Gain 0.38 dB Q 1.29

View attachment 530013

Spinorama EQ Score
View attachment 530009

Zoom PIR-LW-ON
View attachment 530012

Regression - Tonal
View attachment 530011

Radar no EQ vs EQ score
Some improvements?
View attachment 530010
Comparison vs GR Research mods
@totti1965

Stock Score no EQ: 5.4
Stock with Sub: 7.5

GR mods Score no EQ: 5.1
GR mods with Sub: 7.2

View attachment 530035
Spinorama with mods

View attachment 530038
Regression Comparison

View attachment 530034
Zoom Comparison

View attachment 530037
Can't see why anyone would do that to this speaker
BTW some flerfs laws may be applicable to this situation:

Toon’s Laws of Flerf

Flerf citations always contradict the flerf’s claim. No exceptions. Flerf challenges are fake. 2 Exceptions. Flerfs are pseudoscientists when evaluating FE and science deniers when evaluating…

mctoon.net

A huge Hug and a big THANK YOU to @Maiky76!

My take: The Original Version of the Ascend Acoustics Sierra-2EX V2 tested by @amirm is quite close to the measurements taken by Ascend Acoustics (David Fabrikant) themselves

David Fabrikant´s Klippel measurements (Original Version - copyright https://spinorama.org):

• Tonality (Preference) Score is 5.9 and would be 7.97 with a perfect subwoofer.
  • Details: NBD: ON 0.359, LW 0.321, SP 0.297, PIR 0.282; SM: SP0.943, PIR0.876; LFQ 0.829, LFX 43Hz
• Tonality (Preference) Score is 6.91 with an EQ and would be 8.64 with a perfect subwoofer and the same EQ
  
  
  
  @amirm ´s Klippel measurements (also Original Version) according to @Maiky76:
  
  Stock Score no EQ: 5.4 With a perfect Subwoofer: 7.5
  
  Score EQ Score: 6.7 Same EQ and with a perfect Subwoofer: 8.7
  
  So firstly with a bit of Equalization the speaker seems to be very very nice sounding and the Klippel results are consistent.
  
  
  
  
  The differences Erin found out with the Sierra LX (David Fabrikants measurements vs Eric´s measurements) are unfortunately a bit larger, so that we must take the
  Sierra LX fantastic Tonality (Preference) Score of 7.21 with an EQ measured by Ascend Acoustics with a grain of Salt.
  Erin´s Score was a Tonality (Preference) Score of only 6.43 with an EQ
  

 

[voodooless]

Yesterday Danny posted a new video in his YouTube channel of his modifications to some JBL L100 classic which also Erin had measured and reviewed in the past. It seems he produced the video after the ASR analysis of his "improvement" as he compares his crude measurements for few seconds only with Erin NFS' ones and claims they match very well only his are more precise as he measures just a single angle while the NFS averages over many!

It's the Danny Kruger effect

 

[NTTY]

View attachment 530024

Thanks for sharing. Incredibly childish reply. That alone is enough to ignore anything this guy has to say/sell.

 

[mj30250]

Like in all hifi voodoo/snake oil or bad managment and politics I always wonder how much of such is due to incompetence and how much acting in bad faith.

It's likely more of the latter in this case. It's difficult to amass years of experience in commercial loudspeaker design and still completely flub basic fundamentals. He has shown competence with certain designs on occasion, so he is capable of following best practices when it suits him. Unfortunately, his operational strategy seems to be willfully misguiding if not outright deceiving his customers, who do not possess the motivation nor the knowledge to challenge him. Everyone else gets censored.

 

[Chaos01]

Yesterday Danny posted a new video in his YouTube channel of his modifications to some JBL L100 classic which also Erin had measured and reviewed in the past. It seems he produced the video after the ASR analysis of his "improvement" as he compares his crude measurements for few seconds only with Erin NFS' ones and claims they match very well only his are more precise as he measures just a single angle while the NFS averages over many!

Like in all hifi voodoo/snake oil or bad managment and politics I always wonder how much of such is due to incompetence and how much acting in bad faith. He has become a good contender to replace Michael Fremer in the top position of persons from audio I dislike the most.

He fundamentally misunderstands how the Klippel NFS works.

It does not “average multiple mic positions.” It reconstructs the full 3D sound field and derives a mathematically correct anechoic on-axis response. So the on-axis data from Erin's Audio Corner is pure 0°, not mixed with off-axis.

He’s likely confusing this with estimated in-room response, which does combine multiple angles—but that’s a completely different metric.

His claim that Klippel “averages in off-axis” (causing extra HF peaks) is incorrect. Those features are real and often captured more accurately by NFS. If they don’t show up in his measurement, it’s due to typical limitations (gating with 1/3 octave, mic placement, old mic, old calibration, resolution), not because Klippel is “adding” anything or simple because Erin measured the MKII !?

Also problematic:

• A single-point measurement is not more “real” than NFS — usually less accurate
• Differences aren’t just “scaling/zoom” — that’s misleading
• Flat on-axis alone ≠ better speaker (directivity matters)
• Claims about crossover parts “sucking life out of the signal” lack solid evidence

Biggest misstakes:
Klippel NFS = “average of many measurements” → wrong
On-axis from Klippel includes off-axis → wrong
Confusion with estimated in-room response → likely
Peaks in Erin’s data come from averaging → wrong
His single-point measurement is “more real” → opposite is true

Bottom line:
He mixes up on-axis, off-axis, and in-room metrics, and misrepresents what Klippel data actually shows.
Also he claims to be a better engineer than the JBL engineers, my guess is that they voiced this model to a "old school" fr-response and Not because JBL engineers are incompetent...

 

[GM3]

Rusty Jefferson​

Quote from: amirm on Today at 01:55 am
It is not my job to do that. It is Danny's to prove that his mods make an improvement.....

....P.S. ABX tests are improper for this kind of testing. Speakers can likely be recognized to be different. The real test would be one of preference, not whether a difference existed.....

I would respectfully disagree, it is your job to double blind test (and I did say for preference). If you have a problem with what GR Research is doing and only rely on the measurements, your criticism of them is incomplete science. In a fraction of the time you took to measure and make the video you could have done double blind testing with a small group of both trained and untrained listeners. Combined WITH your measurements, if a significant number of listeners preferred the original speaker you would be on more solid footing to criticize what GRR is doing. Right now, you're just making an assumption. And still, as I asked previously, which does the customer prefer? It's a bespoke modification to his/her speakers, that's really the only thing that matters here.

Jesus... Has audiocircle always been that retarded? Well, the people who posted were likely mostly GR Research fanboys from their dedicated forum section... Don't get me wrong, nothing against DBT, but in context asking for a DBT is downright idiotic, when detailed measurements show an increase of distortion, worse off axis or room response, etc., and he's asking for a DBT for listener preferences... Managed to read a little the thread before it was removed, but most members of audiocircle who posted... Sigh... Depressing...

Danny, is still is still up to his slimy vomit inducing BS, trying to make it look like the tested crossover didn't come from him, ex:

They have the one you sold to a customer you lying piece of shit.

There's absolutely zero speculation. And why not explain why it's not the same version you're selling? It's because Dave of Ascend measured and exposed how bad your crossover was, to the point where it would damage the RAAL tweeter if used, and you had to completely scrap your original crossover. You then deleted the original video about the crossover, removed any trace from the website, and made a new video and crossover.

I see that thread is now "quarantined".

Censorship = tool of the weak, the one who cannot defend his own position. What a fucking sleazebag this Danny... Disgusting...

If both speakers of the pair were wired out of phase, you shouldn’t have been able to tell a difference. However, if one of the speakers in the pair was wired in correct phase and the other was wired out of phase, you’d be able a heck of a difference!

(EDIT: btw, only the tweeters were wired out of phase, rereading, you were talking about 2 speakers wired out of phase sounding normal, which is correct! I somehow misread and misinterpreted as in "if both tweeters were wired out of phase", not both speakers! So below refers to my misinterpretation; 2 tweeters wired out of phase vs woofers.)
Btw, not correct. Logically, just like two woofers would cancel each other if one was out of phase, if the tweeter and woofer are out of phase, there will be cancellations at the crossover point; rather than a summation, both will cancel each other's frequencies so you'd get a big suckout / dip at the crossover region; so you'd probably be able to hear some funky stuff.

 

[Maxim-ICOM]

... are more precise as he measures just a single angle while the NFS averages over many!

Indeed, yes.

Latest trend is that we do not see real off-axis graphs for speakers measured with Klippel, most are published without.
Standart 2034 report is smoothed/averaged graphs and it does not tell full story about speaker.

Take a look on Perlisten S7T: all look very very good
S7T spinorama

Here is another measurements of S7T from this topic: https://www.audiosciencereview.com/forum/index.php?threads/perlisten-speakers.19745/
(probably it is from published datasheet)
Orange line (listening window) look perfect, but it is averaging of few measurements under certain angles.
It is not a question to Klippel, but to 2034 what say it should be averaged.

And here is S7T measurements under certain vertical angles - clearly seen that in a tweeter array drivers not in phase under certain angles.

 

[wwenze]

Has audiocircle always been that retarded?

Yes

 

[René - Acculution.com]

Posting this before it gets deleted:

Is GR using outdated measurements for kits/mods? Klippel Wrong measurements?

Is GR using outdated measurements for kits/mods? Klippel Wrong measurements?

www.audiocircle.com

View attachment 529978

One thing I don't understand is a sort of flip-flopping going on. If Danny truly does not listen to the loudspeakers after modifying them (does he state that?), then why talk about people not being able to hear 'audible differences in parts, wire, connectors, etc'? Are the loudspeaker modications based on measurements, while his cables/posts are based on listening tests? If the measurements don't give him the full picture generally, then why not listen to the loudspeakers, and just measure connectivity in the cables and be done with it...

 

[dfuller]

Standart 2034 report is smoothed/averaged graphs and it does not tell full story about speaker.

The CEA2034 listening window/early reflections/sound power are not unique to the Klippel NFS. That is part of the CEA2034 standard measurements and it is all defined.

The NFS can and will break out vertical and horizontal off axis by whatever degree you want it to.

Those "jelly blobs" *are* those data, just represented in a different way than you're used to.

 

[Steve Dallas]

One thing I don't understand is a sort of flip-flopping going on. If Danny truly does not listen to the loudspeakers after modifying them (does he state that?), then why talk about people not being able to hear 'audible differences in parts, wire, connectors, etc'? Are the loudspeaker modications based on measurements, while his cables/posts are based on listening tests? If the measurements don't give him the full picture generally, then why not listen to the loudspeakers, and just measure connectivity in the cables and be done with it...

"Better" parts obviously sound better than "cheesy" parts. Obviously. No listening tests required. And stop asking questions.

 

[DatSpeakerTho]

Jesus... Has audiocircle always been that retarded?

I dont support the language used here, but pretty much every other audio forum is garbage except for ASR ime. Audio related sub reddits are the worst as the vote system puts misinformation right at the top.

 

[dfuller]

For those here who don't know. The "G" in GR Research is Jeff Glowacki, the man behind sonicraft, which is the company that imports Sonicaps. Danny and Jeff were business partners, Jeff was the engineer, Danny was the business/marketing guy. Danny tries hard to hide this info, but don't get caught up in this very obvious scam. There are only 2 places to buy sonicaps, at sonic craft's own website and at GR. Of course Danny raves about these capacitors, he is probably paying $1-$2 for these, massive markups and massive profit margins.

The plot thickens.

And wouldn't you know it: here's Sonic Craft trashing the same exact stuff that Danny trashes.

Also the whole "sand cast resistor bad" thing drives me up the wall. What exactly does Danny think that means? They're all wirewounds, just in a ceramic package. Speaker companies aren't stupid enough to use ones with stray inductance high enough to actually have any effect in the audible range, even in the cheap boxes.

 

[Multicore]

What a disaster. It reminds of people 'upgrading' otherwise fine cars with ugly spoilers and other stuff.

yeah, like "upgrading" the tire/wheel/suspension subsystems.

 

[Jim Taylor]

It's the Danny Kruger effect

I'm stealing that!

 

[fragzone]

Yesterday Danny posted a new video in his YouTube channel of his modifications to some JBL L100 classic which also Erin had measured and reviewed in the past. It seems he produced the video after the ASR analysis of his "improvement" as he compares his crude measurements for few seconds only with Erin NFS' ones and claims they match very well only his are more precise as he measures just a single angle while the NFS averages over many!

Like in all hifi voodoo/snake oil or bad managment and politics I always wonder how much of such is due to incompetence and how much acting in bad faith. He has become a good contender to replace Michael Fremer in the top position of persons from audio I dislike the most.

He is such an asshat that it's tempting to write him off as a fool, but I don't think it's that. Upon seeing the way he subtly utilizes language to imprint an impression on his audience, such as essentially calling the designer of the Sierra 'crazy' (in his revised crossover video) with an elaborate web of words and facial expressions, makes me believe he is quite frankly, acting almost entirely in bad faith.

 

[Multicore]

Amir, as someone who enjoys confrontation and takes pleasure in fighting, I would encourage you to give this guy few punches for this level of arrogant smack talk. . . but I know you're way to professional to fight fire with fire.

Maybe challenge him to send you his v2 Ascend EX "upgrade" and let the data do the punching/talking?

View attachment 530024

I don't believe one can competently apply Fourier analysis without a decent grasp of the bandwidth theorem. Narrow functions in time have broad frequency spectra and broad functions in time have narrow spectra. In the extreme, Dirac's delta function has all frequencies and a sinusoid must extend to both ends of time to have only one frequency. Gating any function broadens its spectrum. It's math. Ok, it's not beginner's math but it's well within what an engineer is expected to handle. Whatever. It is an inescapable theorem. That Mr. Richie accuses Amir of leading a space of flat earth followers when he doesn't know the fundamentals of signal analysis is ridiculous.

 

[tw 2022]

It's the Danny Kruger effect

he has certainly become the industry standard of the dunning -kruger effect...

 

[DanielT]

Also the whole "sand cast resistor bad" thing drives me up the wall. What exactly does Danny think that means? They're all wirewounds, just in a ceramic package. Speaker companies aren't stupid enough to use ones with stray inductance high enough to actually have any effect in the audible range, even in the cheap boxes.

Check out the thread linked below. Thanks for testing, measuring and creating that thread @ctrl

Now we see the effect of the higher inductive part of the sand-cast resistor, because at high frequencies the frequency response drops slightly compared to the low-inductance resistor - up to 20kHz by as much as 0.02 dB!

Even if the effects should increase at higher resistance values, even 5 times the value is inaudible.

The effects of replacing a sand-cast resistor with a low-inductance resistor are practically zero.

Replace resistor by low-inductance resistor - Is it audible?

Since it is often claimed here in the forum that replacing cheap sand-cast resistors with low-inductance resistors would lead to clearly audible improvements, and since it's no use pointing out that the sound change (unless it's completely imagined) is very likely due to slight deviations in the...

www.audiosciencereview.com

 

[thewas]

Indeed, yes.

Latest trend is that we do not see real off-axis graphs for speakers measured with Klippel, most are published without.

Nope, this is not an issue of the NFS but how some users present their results.

Standart 2034 report is smoothed/averaged graphs and it does not tell full story about speaker.

Again this is not an issue of the NFS and even like this the 2034 tells much more and relevant to characterise and find audible issues of a loudspeaker than the couple of arbitrary horizontal and vertical angle FR plots Danny shows.