You should really listen to Manger Speakers because it’s a new listening experience. If you have the opportunity, you should take it when it arises, maybe at the High End Audio Show this year. In my experience, there is no other speaker in this price range and above that is capable of delivering acoustic music so naturally, provided the source can deliver it.
For my Manger P2 with the new W06 driver I use an Ayre EX-8 2.0 Integrated Amp with the new Net2 and USB2 modules together with the excellent Ayre Minimum Phase digital filter and 16-fold oversampling. Personally, I am still waiting for Ayre to implement Native DSD in the new X-8 USB2 module for Linux based sources to try out your HQ player implemented on my Roon Server NUC.
It could be possible already, if you are comfortable enough to deal with a bit of Linux command line, I could take a look if it is relatively straightforward to add support for it. I’ve already added support for couple of other Ayre devices before. Just drop me email, since it it out of scope for this thread.
I remember reading a lot about Manger speakers and being interested.
But I lost interest when I saw some measured performance.
This kind of frequency response will have negative effect on time domain performance, which they claim is critical to great sound quality.
Just my subjective opinion of course.
I guess one could improve the performance with some measurements and convolution.
Could be. The point though is that the filter switching was broken in shipped devices, which was not caught by either QA (if there were any QA, of course) and did not stop reviewers or the few customers who had actually received what they paid for from waxing poetic about the great, audible, wife heard from the kitchen difference “switching” to their favorite filter produced.
So when people start talking about audible differences they hear, show me some real proof. People very demonstrably do believe in “hearing” differences that aren’t just arguably inaudible, but literally not there.
Again, nothing you are posting is actual science. Is a single one of these claimed audible differences observed in double blind testing with statistically significant validation? If not, any such claims are irrelevant.
Well they’re not irrelevant to this thread. I asked Jussi what his experience was, not what scientific papers existed or what double blind tests had been performed. If you aren’t interested in such experience then this is the wrong thread to be reading.
Just integrate the first amplification and impedance converter stage into the microphone, as close to the capsule as possible, to magically solve the cable problem.
The step response measurements touch on something very obvious in this thread.
Too many posts have touted about how important impulse behavior in electronics and digital filters supposedly is for so called imaging, separation, and what not - but everyone neglected to mention how the order of magnitude worse time domain behavior of the most ubiquitous multi-way-driver-with-xover-networks-speaker-designs distort this to oblivion!
For that to be remotely believable/possible, you‘d need to pony up some time coincident, time coherent designs, which are rather seldom in even high-end rigs.
Give me a break, guys … you really can‘t be hearing what you’re trying to make everyone else believe!
Not only the crossovers, but also the effect that differing driver distances from the listening position have on the time domain. I’m playing with Acourate at the moment - converting my 3-way passive PMS to fully active. Acourate generates linear phase crossovers and allows precise time delays to be applied to individual drivers to bring the time coherence error of the entire system down to a fraction of a millisecond before applying room correction filters. Using the software is challenging and the learning curve is steep, but I’m looking forward to the final results.
Lots of audiophiles sweat buckets over the minutiae and completely overlook the elephant(s) in the room - the fundamental limitations and compromises within the design of a speaker, whether active or passive, its placement within, and its interaction with the listening room.
Human hearing is amazing. In some frequency bands, we’re only a few bels from being able to hear brownian motion.
If you can, I’d suggest you listen to talks by Rob Watts, the engineer behing Chord Audio’s DACs (he usually comes to speak at Head-Fi’s CanJam london events, for instance).
At 4 kHz, Brownian motion of air molecules at room temperature is around -23 dB. However, it’s still well below the threshold of human hearing.
Er, no. Rob Watts has made some ridiculously outlandish claims about audibility of things that are measurable, but so far below the threshold of human hearing as to be absolutely inaudible.
As mentioned, there is still enough of footroom with the human audibility threshold around -5dB SPL in a very limited band.
Molecular movement is one thing, actual movement of air molecules is another one. There are some very effective (and expensive) anechoic chambers, but even in a quiet studio room the noisefloor is much higher than -23dB SPL.
You can, but solely for fullrange minimum-sized piston-like drivers as found in headphones or some coaxial loudspeakers. Most of conventional loudspeaker designs, particular inline multi-way concepts or larger fullrange drivers such as the Manger defy any correction in the time domain as you cannot correct it for larger areas but only very limited angles.
Step response is according to my experience and scientific knowledge the most misleading type of measurement. The human ear is not as sensitive to distorted group delay in the region of microseconds while the group delay problems which are actually audible, in the bass region, are usually not measured or underrepresented in a standard step response graph.
I do not deny the audibility and importance of time behavior of a reproduction chain but we should admit that we know very little about how to technically explain it or measure it.
There are quite controversial opinions on the sound quality of the Mangers but everyone who has ever listened to them would probably agree that is does sound very different from any expectations people are evolving from the measurements.
That is nothing unusual with such speaker concepts combining flexible diaphragms and different directional behavior in different frequency bands. It is to be expected.
Convolution can change neither the effect of directivity index of a speaker over frequency nor interference and narrow-band cancellation under different angles nor audible resonances/distortion. As you can see in the measurements you have linked, there is a dip at 1.7K combined with resonance phenomena seemingly as a result of parts of the diaphragm acting out of phase with each other. Textbook example of something you cannot correct with convolution filters.
To be fair, these ideas have been published by J.W.Manger decades ago and many theories in there have been verified later and largely debunked. Particularly since DSP are available to emulate or correct reproduction in the time domain.
My intention was to show an example in relation to the topic of the thread that shows exactly the discrepancy and the Manger speakers are, in my opinion, a prime example.
Especially with loudspeakers, it is very important how you measure in which environment. A speaker such as the Manger speakers with a completely different driver concept and different dispersion behavior than the usual two, three or four-way cone driver speakers may require a slightly different placement of the measurement microphone in order to reproduce a correct measurement result.
As an example: In 2019, Herb Reichert from Stereophile tested the Manger P1 and was impressed. Stereophile Manger P1 Review “From my chair in the sweet spot, the p1s appear to be an audio-engineering breakthrough that eliminates a multitude of cone-generated noise and time-domain irregularities still present in many of today’s highly regarded loudspeakers. The result of all Manger’s technological veil-lifting and detail excavation is a loudspeaker that comes preternaturally close to disappearing—while reproducing recordings in a manner that felt uniquely unmolested.” Many other reviewers from various hi-fi magazines came to similar conclusions about various current Manger loudspeakers.
John Atkinson then measured the speakers with a “mixed” verdict. Daniela Manger then very politely pointed out to him in the “Manufacturer’s Comment” that perhaps his usual measurement method for ordinary loudspeaker concepts does not always lead to correct results.
It is a good example that it is important what and how you measure and that in the end what is most important for us personally should be what we perceive acoustically with our ears and brain.
I disagree! Manger speakers sound unusual at first because they don’t have the system-related distortions of the usual cone drivers and cover the frequency range from 40 kHz to 300 Hz with one driver. Almost all manufacturers trust in this conventional concept, perhaps with an AMT for the treble range, but this leaves the mid-range frequencies that are important for human hearing untouched. Finally most of us are used to the special “sound” of this kind of speakers. But do they sound really natural? Usually not, because of their distorsions.
Many high-end manufacturers try to reduce this with AMT drivers, “diamond” tweeters or other high-tech materials, but the inherent problem of the cone drivers for the high or midrange drivers cannot be completely eliminated, even with DSP manipulation.
I had this new “experience” when, after more than 30 years with conventional loudspeakers at “high-end” level, I heard the Manger S1 for the first time at a high-end show in Munich sometime in the early 2010s. It sounded different than usual, but also somehow “right” and “natural” for someone who is familiar with the real sound of acoustic instruments and voices on a stage. Once you are familiar with the “new” sound, you don’t want to go back to a conventional loudspeaker, even if some systems, which are usually much more expensive, offer significantly more bass dynamics at extreme volumes. But who of us can really experience this in our lives and listening room reality?
A negative example of a discrepancy between human hearing and measured values for me is the Kii 3 loudspeaker. I have heard it in many environments, the frequency response may have been correctly smoothed and the measurements were perfect, but for me it was a completely soulless music reproduction, “optimized to death” with DSP. I think sometimes marketing investments and the reputation of a renowned developer work wonders when it comes to the evaluation of a product in most HiFi magazines.
That is not a legitimate conclusion. As I explained, the MST produce a different soundfield as they are partly bending-wave transducers, so a frequency response measurement by one microphone at a particular position would say nothing about how they sound. There simply is no ´correct mic position´ as the way we perceive their sound is depending on the soundfield in a larger area.
The same is true to other speakers with flexible diaphragms or bending wave transducers such as Magnepans, Goebels or NXT.
I support this claim by the manufacturer but it leaves the question open which measurement method would be doing the product justice representing what you can actually hear.
Cannot call that accurate sound reproduction