If It Ain't Broke, Don't Fix It

I’ve found both ASR and GR to be useful resources.  
 

I find the cat fights pretty pointless.  Just take it for what it’s worth to you or not and move on.   Or if one really feels they got the answer keep publishing your own content to make the world a better place. ✌️

@hjdca Thanks for the clarification and sharing your opinion!

I have found that usually when somebody was banned from ASR, they were banned for a pretty good reason :-)

 

that said:

 

@xmbw4 

 

 

— “I challenge anyone to look at all of ASR’s reviews and find a single instance where he used something that would be considered real science in anything he’s ever done.”—

 

I find these critical callouts from  anonymous folks on audio forums to be amusing.

 

Amir was a senior executive overseeing codec and media technology development at Microsoft.  He  oversaw the development of successful audio and media codecs.  They employed standard scientific protocol for developing codecs - eg double blind testing.

The technologies made it into billions of devices and even received Emmy awards for their development.

 

He uses top shelf, well-known lab equipment to do many standard electronic measurements for distortion, etc.   And he evaluates the relevance of the distortion measurements with respect to known audible thresholds, well established in psychoacoustics research.

 

He evaluates loudspeakers using currently among the world‘s best measuring devices - the Klippel - one reason being is the type of measurements you need to render data relevant to all the SCIENTIFIC research cited by Floyd Toole and other such experts.  In other words, Amir is using sophisticated measurements that allow him to correlate those measurements to the best science we have in terms of predicting subjective impressions of loudspeaker behaviour.

 

So he is basically following the best current science we have.

Amir’s approach and data has got Floyd Toole’s approval (he posts

on ASR fairly often).

But for some anonymous people on the Internet…. Of course it’s not enough. ;-)

 

 

 

Hello All. My experirnce with the GR upgrade kits for Magnepan LRS and LRS+ speakers resulted in MUCH improved sound. I've been emeshed in this hobby for over 65 years, have built many sets of speakers and modified many manufacturer's products. I find low priced parts in the crossovers of everybody's speakers. There are fifty cent capacitors, five dollar capacitors, fifty dollar capacitors, and more. If you are manufacturing 1000 pair of speakers, that price difference really makes a difference. Most of the money in manufactured speakars is in the finish, packaging, shipping, and profit. Maybe 10-20% of a manufacturer's speaker is in parts. That's where they can save money. The fifty dollar capacitor does sound better than the five dollar capacitor, but not ten times better. But we, as customers and buying only one pair of speakers, can afford the difference in parts price and enjoy the improvement in transparency  and "air" that improved parts provide. Is the difference between a ferrite core, small diameter wire inductor and the same rated flat foil fancy copper hockey puck $100+ inductor audible? Yes, instantly. Can you afford $200 to upgrade your inductors? Probably. Let the drivers show you what they can do by getting the signal to them without being degraded by cheap parts. GR Research has to make a profit on what they sell. But they buy large numbers of parts at wholesale prices. The "upgrade" parts they supply are likely to be cheaper than you will pay (single units, retail) for parts. They will send you the circuit diagrams for free, you can build them yourself using whatever parts you wish, but if you don't buy the best parts, you won't get the best results. Isn't that reasonable? In my experience, upgraded crossover parts always improve performance. Of course, you could use an electronic crossover and one power amp for each driver. That's what give me the best results. Enjoy the music.

Thanks to everyone who weighed in. Lots of different views and I wanted to try to summarize.

Several people questioned ASR’s general methodology or credibility (@wsrrsw, @audioman58, @xmbw4). I understand that skepticism — ASR has a strong measurements-first ideology that rubs some people the wrong way. But I’d ask those folks to notice that the video isn’t making ASR’s usual ’measurements are everything’ argument. It’s making a narrower, more specific claim: that the measurements Danny used to justify this particular mod were methodologically insufficient for the precision work he claimed to be doing, and that the resulting changes are demonstrably worse by almost any metric you care to use.

@mapman and others took a sensible both-sides view — ASR useful, GR useful, let’s not fight. I understand that position. But I’d push back gently: the both-sides framing treats this as a dispute about values (measurements vs. listening) when it’s actually a dispute about the quality of specific measurements used to justify specific claims. Those are different arguments.

@dz13 made a fair point that Danny responds to owners who are already dissatisfied. True, and worth noting. But the consent of the owner doesn’t validate the technical methodology or the public claims made about the speaker and its designer.

@nlitworld raised the most technically substantive challenge from the GR-sympathetic side, noting that gated measurements have been standard practice for decades. This is correct — and actually the video addresses it directly. Amir’s point isn’t that gated measurements are useless, but that they’re unsuitable for the degree of precision Danny’s modifications required. Using a coarse tool for fine work and then claiming improvement is the problem. If the defects are gross and obvious (as in the Klipsch RP600M case, which Amir concedes Danny handled reasonably), gated measurements are adequate. If the speaker is already extremely well-optimized — as the Ascend Sierra 2EX demonstrably is — they aren’t.

@prof made a useful contribution by pushing back on xmbw4’s sweeping dismissal of ASR’s scientific credentials. Amir’s background, his use of the Klippel NFS, and Floyd Toole’s endorsement of his methodology are not nothing. Toole is probably the most important researcher in the scientific study of loudspeaker perception, and his approval matters. I’d add to prof’s point: the Klippel Near Field Scanner that Amir uses — and that Dave at Ascend also purchased after early critical reviews motivated him to upgrade his measurement capabilities — costs $150,000. Dave invested in that equipment, improved his designs based on what it told him, and built a better speaker. That’s the right response to criticism.

@decooney, @simonmoon, and @boomerbillone all offered personal experience with GR upgrades. Interestingly, these accounts aren’t uniform. Simonmoon and boomerbillone report genuine improvements on the speakers they worked with, and I don’t doubt that. But decooney’s account is worth looking at — his friend’s upgraded speaker measured flat, but sounded like "some of the life was taken out of it," and the owner’s verdict was that it sounds "okay." That’s a telling result: technically flatter, but not obviously better to listen to. In any case, Amir himself credits Danny with a legitimate fix on the Klipsch RP600M, so the question was never whether GR upgrades can ever be worthwhile. It’s whether this upgrade, on this speaker, was justified — and the evidence says no.

The core point I keep wanting to return to is that Dave at Ascend built the Sierra 2EX using state-of-the-art measurement tools, years of careful listening and iteration, and a deep understanding of how drivers, crossovers, and room acoustics interact. When Amir gave an early Ascend speaker a poor review, Dave didn’t argue or dismiss it — he studied the methodology, bought the same measurement equipment, and improved his designs.

What the video documents is that Danny, using inadequate measurements, made confident public claims that this carefully optimized speaker needed fixing, sold a $500 kit that introduced measurable distortion, degraded the tweeter’s performance, and worsened the off-axis response — and then doubled down when challenged. That should matter to anyone in this hobby who cares about accountability.