The REAL Problem of March Audio's Sointuva WG (Review, Measurements and Reinforcements with Klippel device)

[digicidal]

As someone who began their membership here significantly "under fire" by many, for questioning the veracity - or at least ubiquity of application - of single-sample reviews... I have mixed emotions. The irony that one of the most condescending and defensive voices in that exchange was Alan March is particularly notable. Admittedly, I made two mistakes, which I've hopefully made significant improvements to in the years that followed: 1) I didn't spend enough time considering the wording and specifics of my question/concern, and 2) I made the posts while my account was very new.

However, the response I got from Amir and others - but most vehemently from Alan himself - was that the frequency of manufacturing flaws and/or sample-to-sample deviations was so small as to be safely disregarded. At the time, I was mostly concerned with the potential for "golden samples" or "brown samples" (my term) invalidating good or bad reviews, respectively. While that may be a largley defensible position regarding the likes of individual ICs and/or products from very large manufacturers like DM, Harman Group, etc. - it's clearly not the case with more boutique offerings (even from some long-established brands with mature production lines).

As it pertains to this review's subject device and manufacturer - this appears to be the completely opposite position. That any flaws, inconsistencies, etc. are the fault of the mass manufacturer (Purifi in this case) - which I know has significantly larger volumes and revenues and I would therefore presume also significantly more precise (or at least more automated) QC practices. At the same time, the small volume manufacturer "knows all of these flaws, and can fix them", but then still ships at least two units with significant variations of assembly and performance. So Purifi can easily release products with glaring, obvious problems... but the company that just purchases goods off-the-shelf and assembles in-house cannot possibly have any?!?

I have nothing but praise for the OP review (and reviewer), and while everything has a solution available to a determined, DIY-enthusiast consumer - outside of that limited demographic, precious few consumers are willing to even consider disassembly of a ~$4K purchase to improve performance. Not my business, so not my call, but personally I would rather lose a whole month's revenue to preserve/establish a pristine customer service reputation - than risk losing all of it to retain the profits on a few problematic samples. Even if I was 100% sure it was shipping damage, component failure, etc. Personalities and egos aside... assuming the customer and/or OEM is always wrong, but your own operation cannot ever be - is definitely not confidence inspiring to say the very least.

I should say in defense of March Audio (and against my personal feelings on the matter): Once all of the remediation actions were taken - it seemed to perform quite well overall, and I would expect it to sound good if you got a pair that had none of the issues the review samples had. Whether that is a net win or loss would be up to the (hopefully informed) consumer to decide.

 

[abdo123]

I'm convinced with March Audio's response, they recognized the mistake of not sealing the binding posts. These things happen.

For the distortion issue I'm 100% convinced too. while it sounds stupid it could be that if the magnet is a bit physically cushioned by the extra absorption material and the tension on the baffle mount is reduced as a result causing the resonance to go away. Either way March Audio is not in the wrong here.

the impedance plots after the absorption material added show a completely different tuning, if sample B doesn't have the issue but sample A does and when sample A was 'fixed' by making it something that is electrically different than Sample B then that was not a fix. lack of absorption material is in no way shape or form the root cause here.

As for the tweeter leak issue, well everyone is free to draw their own conclusions.

Personally I think Alan should have provided the reviewer with instructions from the get go on how to deal and diagnose the binding post issue and this should have never evolved to what it is right now.

 

[Thomas_A]

This seems to be a mixed bag. The driver itself apparently has a resonance, which will be affected by the inertia of mounting and probably also if this could excite other cabinet resonances. So one could expect some individual unit variation and since mounting inertia will change over time. A motor strut/brace is used in some speakers, which could be a solution, and also use sandwich construction with damping of the cabinet.

That said, faulty units happen sometimes, and there was one example on ASR of a Revel f208 with a bad cap. No support from Revel, replacement from the OP fixed it, however opening it caused cracks in speaker finish.

Speaker faults are rare, but they can happen. This Purify woofer seems to have some issues though.

 

[Willem]

I have no views on the technical issues, but I do think some of the posts here smell of a vendetta. Alan March has been civil in his correspondence, and that includes the money back offer.

 

[Nuyes]

Maybe, and I understand that the OP has difficulty in english so maybe he can answer clearly, but in the support email (not the later justification), Alan March ask a simple question: "However, can I ask if the fixing bolts where tighten?" we don't see a clear response to that and again, beside the OP post said that well he noticed that screw torque was uneven "On his perspective, at disassembly" but goes on with "I decided to ignore Alan March response. Sounds to me as a way to partly avoid the question. Maybe @Nuyes can clarify, but again there is truth and great value to the OP's investigation, I do think that Ignoring support attempt and then expressing disappointment is not fully fair. At least you answer, especially if you are going to use the email communications in your review to voice your disappointment.

I want you to know that it has been over a month since Sointuva WG's first review was published before this ASR report was created.

Who wants their precious $4,000 speaker to be ruined by an anonymous reviewer?

The owner of this speaker and I haven't even seen each other's faces yet.

We reported a problem that A/B testing of this speaker revealed (before disassembling it) to Alan, who was suspicious of reviewers from the start.

The speaker owner, having a gut feeling that this difficult tuning process would not come to a conclusion, entrusted me with an analysis of all these issues.

 

[digicidal]

I'm convinced with March Audio's response, they recognized the mistake of not sealing the binding posts. These things happen.

For the distortion issue I'm 100% convinced too. while it sounds stupid it could be that if the magnet is a bit physically cushioned by the extra absorption material and the tension on the baffle mount is reduced as a result causing the resonance to go away. Either way March Audio is not in the wrong here.

the impedance plots after the absorption material added show a completely different tuning, if sample B doesn't have the issue but sample A does and when sample A was 'fixed' by making it something that is electrically different than Sample B then that was not a fix. lack of absorption material is in no way shape or form the root cause here.

As for the tweeter leak issue, well everyone is free to draw their own conclusions.

Personally I think Alan should have provided the reviewer with instructions from the get go on how to deal and diagnose the binding post issue and this should have never evolved to what it is right now.

I cannot dispute the (potential at least) truth in that... however, I would definitely dispute the "either way March Audio is not in the wrong here" part. While it may indeed be the case that design problems and/or manufacturing defects on the Purifi side are the cause... it does not change the obligation of the final manufacturer to stand behind their product completely. That goes double if they are, in fact, fully aware of those defects.

While it may be Takada's airbag that was the problem... it is the manufacturer that made the car that has to fix the problem, initially at their own cost. If the problem is significantly damaging to their business... then they can pursue remuneration from the OEM. To Alan's credit, it seems a full refund was offered and/or a return & fix repair - which is appropriate. However, that does not answer the question of how apparently known defects, with apparently known solutions - escaped an apparently flawless QC process in the first place.

I completely agree with your last sentence, and I admit to probably doing too much "reading between the lines" in the tone of his responses. That doesn't change the fact that refusing to even consider any possibility of negligence or oversight on the final assembly and production side is a big issue. It's definitely not new, nor unique to March Audio however - there's an epidemic of blaming everyone else inherent in almost every product and/or service offered today. Although maybe that's just my unfortunate experience and others have much better luck?

 

[Sir Sanders Zingmore]

I haven't read every post to see if this has been asked ... How is that the purifi driver can be hailed as the bestest lowest distortion driver in the universe if it has a fundamental issue like a nasty fundamental resonance that means it actually has a region of very high distortion?
(*it's quite possible my question is borne of ignorance, apologies if it is)

 

[Shadrach]

March Audio has a few quality control issues. Most companies have them. I've seen similar and worse in other products.
The responses from March Audio I've read here seem reasonable while perhaps not the last word in customer care.
I didn't follow the events that led to Erin and Alan getting banned. It's amir's forum. He can ban who he wants.

I have no idea what would Nuyes motivation is for firstly bothering to make such measurements and then publish them here and frankly I'm not interested in reading any explanation.

I think it would have been better politics and better forum management to post the original post from Nuyes and then lock the thread.
All the comments, constructive, or otherwise are not going to make the slightest difference to the existing problems.

 

[PeteL]

I want you to know that it has been over a month since Sointuva WG's first review was published before this ASR report was created.

Who wants their precious $4,000 speaker to be ruined by an anonymous reviewer?

The owner of this speaker and I haven't even seen each other's faces yet.

We reported a problem that A/B testing of this speaker revealed (before disassembling it) to Alan, who was suspicious of reviewers from the start.

The speaker owner, having a gut feeling that this difficult tuning process would not come to a conclusion, entrusted me with an analysis of all these issues.

Thank you, but the question is simple, did the bolts where tighten before measuring?

 

[beren777]

Can someone send a sample to Amir for measurement?

 

[SDC]

My all 4 Purifies are getting ready for measurement.

Hope 15T al6061 and Irwin HD is rigid enough for everyone.

Will be measured with @Nuyes 's distortion analyzer 2.

edit: 3T al changed to al5052 if anybody cares.

 

[digicidal]

Thank you, but the question is simple, did the bolts where tighten before measuring?

I may incorrect in this, but I believe no in the first measurement but yes in the second. I remember reading that some were loose enough to be turned with fingers alone while others were tight. Regardless, I think it's safe to assume that they were not uniformly tight in the first test? Since I don't remember any mention of using specific torque values in the final reassembly it's possible neither was uniformly tight... just tighter tolerances than initially discovered.

 

[Nuyes]

Thank you, but the question is simple, did the bolts where tighten before measuring?

Again, I've never tampered with the speakers before taking measurements.
Including bolts.


Alan goes on to say.
"I changed the speaker arbitrarily and then proceeded with the measurement," he said.

But I swear I never touched anything.
Even when passing a bunch of data to Alan through a sample A/B test.

 

[Nuyes]

I may incorrect in this, but I believe no in the first measurement but yes in the second. I remember reading that some were loose enough to be turned with fingers alone while others were tight. Regardless, I think it's safe to assume that they were not uniformly tight in the first test? Since I don't remember any mention of using specific torque values in the final reassembly it's possible neither was uniformly tight... just tighter tolerances than initially discovered.

It happened after the speaker's owner had confirmed the progress of the repair.

 

[hmt]

His story is not consistent with what is said in OP. He had no motivation to dig into the speaker if the distortion was not there to start.

Maybe the problem was the binding posts in the first place. Tbh the explaination given by march audio does not seem worng.

 

[Nuyes]

I may incorrect in this, but I believe no in the first measurement but yes in the second. I remember reading that some were loose enough to be turned with fingers alone while others were tight. Regardless, I think it's safe to assume that they were not uniformly tight in the first test? Since I don't remember any mention of using specific torque values in the final reassembly it's possible neither was uniformly tight... just tighter tolerances than initially discovered.

I have been very honest and I am not rude.
I never arbitrarily touched anyone's precious $4,000 speaker.
Even if it's $40.

 

[jae]

A little bit off topic, but do you have a database/site of your other reviews @Nuyes ? Do you produce Spinorama/CEA2034 datasets?

 

[hmt]

I haven't read every post to see if this has been asked ... How is that the purifi driver can be hailed as the bestest lowest distortion driver in the universe if it has a fundamental issue like a nasty fundamental resonance that means it actually has a region of very high distortion?
(*it's quite possible my question is borne of ignorance, apologies if it is)

Plus it has also been overseen by amir in his review (equipment stack?). No problem, happens to everyone. But what's really unfair is that some accuse another reviewer of cheating for some oversight while it is just ignored here. Double standards.

 

[digicidal]

I have been very honest and I am not rude.
I never arbitrarily touched anyone's precious $4,000 speaker.
Even if it's $40.

I understand, and I was not suggesting otherwise.
As far as I am concerned, no reviewer should have to take any steps beyond calibration and configuration of the testing equipment... if the DUT has assembly flaws or component defects... that's on the manufacturer and not the reviewer.

 

[beren777]

I understand, and I was not suggesting otherwise.
As far as I am concerned, no reviewer should have to take any steps beyond calibration and configuration of the testing equipment... if the DUT has assembly flaws or component defects... that's on the manufacturer and not the reviewer.

Not to put words in his mouth, but the reason why should be very clear: as soon as a reviewer makes changes to a DUT, the manufacturer can blame the reviewer for issues.