Dips for example are a lot less audible.
Directivity above 10kHz is hard to detect.
Edge diffraction on itself can be really hard to detect as problematic.
Worst I'm afraid is the audioXpress driver testing articles with Vance in the lead... terrible. Everything is fantastic engineering and performance.. despite awful measurement results presented.... so bought.... e.g.: https://audioxpress.com/article/test-bench-w4-2315-home-audio-4-5-coaxial-driver-from-tb-speaker
Page content otherwise fine.
//
Page content otherwise fine.
//
A deep, narrow-band notch would not be audible. But it'd look horrific in a frequency response.
Most of us don't notice the comb filtering that naturally occurs in a room unless it's either really bad or something else is going on as well. But comb filtering looks bad in a frequency response.
And, no. I'm not suggesting that we throw frequency response out with the bathwater. I would never accept an amplifier with a frequency response that looked like a comb filter.
Tom
I agree. And their application is inconsistent at best. But I honestly don't think Amir cares about the panther as a review summary. He just thought they'd make the product pictures look unique and give a sense of scale. I can see where he's coming from with that, but in my view he should recognize that the community he serves views it differently and adapt his ways. But unless you can somehow plant the idea for change in his mind and make him come up with it I doubt you'll see any change.
Tom
Well, since when are
Impedance blips like:
WF like Island after a volcano erruption
3-4% distortion in tweeters
not to mention 15% in the most critical band for the woofer (changing the scale an attempt to hide the result?? - good advertiser?)
and 4 dB unit difference over q quite wide region
"a very well designed and configured coincident coax transducer. TB Speaker’s products are consistently well engineered with a combination of innovation and creativity,"
"W4-2315 representing another good example of the company’s craft." - I hope it isn't for their sake...
What do you think Jan?
Also, in these test there are almost always a few editorial errors like "Figure 10: TB Speaker W4-2315 on-axis frequency response (black curve=woofer, blue curve=compression driver)." - ni compression drivers what I can see... this is sloppy reuse...
//
Impedance blips like:
WF like Island after a volcano erruption
3-4% distortion in tweeters
not to mention 15% in the most critical band for the woofer (changing the scale an attempt to hide the result?? - good advertiser?)
and 4 dB unit difference over q quite wide region
"a very well designed and configured coincident coax transducer. TB Speaker’s products are consistently well engineered with a combination of innovation and creativity,"
"W4-2315 representing another good example of the company’s craft." - I hope it isn't for their sake...
What do you think Jan?
Also, in these test there are almost always a few editorial errors like "Figure 10: TB Speaker W4-2315 on-axis frequency response (black curve=woofer, blue curve=compression driver)." - ni compression drivers what I can see... this is sloppy reuse...
//
@TNT
It's not totally fair, because that distortion is at a whopping 94dB.
Poor little 4 inch coax (effective Sd less than 4 inch). 🙁
It's a mistake from the tester to measure these small drivers at such high SPL.
They have done it before with like 2 or 3 inch drivers as well.
Nuts.
Although the tweeter should still be fine.
I agree with you.
You missed something else that is important, the BL and Kms graphs that are both highly asymmetrical.
These asymmetries happen with poor production and assembly.
Which shows bad engineering as well.
I completely agree with you that these conclusions shouldn't be in that test.
The data is not in line with the conclusions at all.
So yes. There's a lot wrong with this article.
I have been reading these testbench measurements for over a decade now.
Very often conclusions don't line up with the data.
So I stopped reading the verbal text all together.
It's not totally fair, because that distortion is at a whopping 94dB.
Poor little 4 inch coax (effective Sd less than 4 inch). 🙁
It's a mistake from the tester to measure these small drivers at such high SPL.
They have done it before with like 2 or 3 inch drivers as well.
Nuts.
Although the tweeter should still be fine.
I agree with you.
You missed something else that is important, the BL and Kms graphs that are both highly asymmetrical.
These asymmetries happen with poor production and assembly.
Which shows bad engineering as well.
I completely agree with you that these conclusions shouldn't be in that test.
The data is not in line with the conclusions at all.
So yes. There's a lot wrong with this article.
I have been reading these testbench measurements for over a decade now.
Very often conclusions don't line up with the data.
So I stopped reading the verbal text all together.
That coax is very NOT good.
I'm with @TNT and @b_force
Many of us have been skipping the verbal text, and just looking at the graphs. But after awhile noticed that there is lot of distortion that isn't consistent with other measurements. Then we go back and read the text...
well the electret condenser mic own slipping point is 126dB
But then the mic is exposed (measures) excess of 120dB..
@jan.didden
Do you have a direct line to Vance?
Vance is using the 1/4" constant current powered GRAS 46BE microphone for frequency response measurements.
Why not use it for distortion measurements? It's 3% distortion point is 160dB (and thus 150dB for 1%).
Conversely, at given given SPL, it's own self distortion is at least 24dB lower (an order of magnitude) than the electret condenser mic...
I'm with @TNT and @b_force
Many of us have been skipping the verbal text, and just looking at the graphs. But after awhile noticed that there is lot of distortion that isn't consistent with other measurements. Then we go back and read the text...
well the electret condenser mic own slipping point is 126dB
But then the mic is exposed (measures) excess of 120dB..
@jan.didden
Do you have a direct line to Vance?
Vance is using the 1/4" constant current powered GRAS 46BE microphone for frequency response measurements.
Why not use it for distortion measurements? It's 3% distortion point is 160dB (and thus 150dB for 1%).
Conversely, at given given SPL, it's own self distortion is at least 24dB lower (an order of magnitude) than the electret condenser mic...
Last edited:
I start to see a systematic error in the used measurement system - it seem to be at 1kHz. I think that it needs a review - yes.
//
//
The Audio Express website seems to be drunk/broken lately in general.
I almost check it daily (work related), and very often newer articles seem to all of a sudden pop up within the older articles.
Yet having an older date on it.
Sometimes you have to go back a page even to read the missing articles.
Extreme confusing and not a very pleasant experience in general this way.
I almost check it daily (work related), and very often newer articles seem to all of a sudden pop up within the older articles.
Yet having an older date on it.
Sometimes you have to go back a page even to read the missing articles.
Extreme confusing and not a very pleasant experience in general this way.
Amir at Audio Science Review is pretty good. And the guy at Benchmark. And Blaine at headphones.com
A pic from Benchmark of their listening room speakers:
Pic from https://benchmarkmedia.com/blogs/application_notes ...about half way down the page.
"Benchmark's listening room is equipped with a variety of signal sources, amplifiers and loudspeakers, including the selection of nearfield monitors shown in the photo. It is also equipped with ABX switch boxes that can be used to switch sources while the music is playing."
Does anyone else besides me see what is terribly wrong here?
Maybe that's why DAC-3 sound stage is narrow and forward. IME forward is not good; its suggestive of some type of random-noise jitter issue (a type of jitter which doesn't show up well on typical J-Test measurements). Not surprising if they couldn't hear the effect if all those passive dampers are interacting with the each other and with the actively driven speakers.
Yes. The speakers are packed up against one another and can therefore not provide in-room performance as "designed" (which they probably aren't anyway). Also it's Benchmark - a brand pushing $150 DACs at $3000 via the bogus talkers of audiofool magazines.
Any audio magazine needs to be viewed with more than an ounce of sceptisim. Capital gains and honesty are opposing forces in this world. U tube is were its at these days.look what happened to golden sound commenting on that fancey english eletist audio.
Most reviews need to be read carefully. If there is any warning at all about shortcomings, it will probably right around the second to the last paragraph of a review. Those who know what to look for will not be fooled; others may not be so fortunate. Sorry, but its the way it is. Review magazines can only exist around here if they are not openly critical of products they accept for review. Maybe it was different in another time, and or maybe its still different in another part of the world. However, knowing about where the warnings are likely to be makes reviews still of some use to the initiated. Better than nothing so long as you know.
Last edited:
To be fair, Benchmark dacs cost more to make than Topping dacs at $150. Don't assume SINAD is the right metric by which to judge. That's already a big mistake. You can blame that other website for promoting the unproven theory while failing to disclose it is only a theory of Amir, and not established science.