DYNAUDIO C4 VS B&W 800 D3

[Art Vandelay]

I agree. And still I'm saying, these are not the only good sounding speakers, that's not the point, but it's an effort worthwhile.

I had an enjoyable time recently listening to a pair of ML Summit x electrostatic speakers. They probably defy the Harmon design / performance criteria in many respects, but these are a speaker I could live with quite easily. There's definitely something to be said for time and phase coherency, and with ML's it's not significantly at the expense of other important parameters. Obviously you need plenty of space around them, and as with all panel speakers the sweet spot is pretty limited.

In the future, decades from now, I can see most speakers being multi-way active dsp designs with dsp correction, pre-distortion, and high order crossovers etc. It's all possible now of course but the evolution will take time. There's still quite a bit of resistance, and much of that from within the industry itself. My feeling is that many companies are frightened to launch a technically too-perfect, too-integrated product for fear that it will reduce future sales and profits.

 

[Art Vandelay]

All I can say is that the Stereophile measurements would indicate that the speaker would have a degree of sibilance, and yet I detected not a trace of it. In addition the speaker would be expected to sound somewhat forward. The speaker infact sounds slightly recessed which is much more in line with the slight dip in response I measured at 5 KHz.

The slightly recessed sound is more due to the power response dip around 2kHz. The peak at 4kHz is the crossover frequency, so it's not a power response peak. It's only a peak on-axis. Similarly, the 5kHz dip is not present off-axis in the H plane due to the fst mid driver response roll-off off-axis.

And lastly, the peak at 10kHz looks far worse than it is - again due to the off-axis roll-off. The power response, which is the sum of all radiated energy, is quite flat above 3kHz, which explains why the speaker sounds balanced in a normal furnished listening room.

The on-axis peak at 10kHz is more pronounced with the B&W diamond dome than many others, but it's still due to a combination of dome geometry and relative size proportional to the wavelength. The latter is what causes 'beaming' and the former creates the roll-off above 15kHz. It's because at that high frequency the physical distance between the outer dome and the dome's centre, viewed from the front - on T axis, starts to approach 180 degrees. The way to negate it is to use a smaller or flatter dome, but listening 20-30 degrees off-axis also works.

And for those who are not familiar with the B&W crossover design, the dips either side of the peak at 4kHz are the result of b&w's 'mechanical' phase alignment. The use of 2nd order filters usually necessitates inversion of the tweeter polarity, but B&W decided to move the tweeter forward by a half wavelength - to maintain absolute polarity (they claim). The downside is an out-of-phase situation between the mid and tweeter - one octave below and half an octave above the crossover point. Personally I think they would have done better to invert the tweeter, but it does have the added benefit of eliminating the potential for diffraction effects.

 

[shadyJ]

I had an enjoyable time recently listening to a pair of ML Summit x electrostatic speakers. They probably defy the Harmon design / performance criteria in many respects, but these are a speaker I could live with quite easily. There's definitely something to be said for time and phase coherency, and with ML's it's not significantly at the expense of other important parameters.

ML summits probably don't have a significant advantage in phase than conventional designs, perhaps it would have even less. The very broad radiating surface is too large to maintain phase coherence at any single listening position. If you want something totally phase coherent, you need to look at a full-range driver speaker, and not one with a huge surface area.

 

[killdozzer]

as with all panel speakers the sweet spot is pretty limited.

I never knew this. I thought it was quite the opposite.

 

[Dennis Murphy]

I never knew this. I thought it was quite the opposite.

This is just a matter of the width of the radiating element. The wider it is, the more restricted the dispersion. ML mitigates this a bit by curving the element, but that only helps a little.

 

[Swerd]

Does that mean one can't just decide to participate in a DBT? @Swerd, is that correct?

Anyone who wants can participate in a single or double blind listening test. There are no requirements that I know of.

Of course a philosopher could argue that anyone who actually does want to participate in such a test probably has preconceived notions that might or might not introduce a testing bias. By this thinking, the best participants would have to be unaware of the true nature of the question involved.

It's my impression, based on reading Toole's book a few years ago, that blind listening tests of loudspeakers using experienced or inexperienced listeners produced little difference. If anything, experienced listeners performed slightly less reliably than inexperienced listeners.
Sorry for the late answer.

 

[Art Vandelay]

ML summits probably don't have a significant advantage in phase than conventional designs, perhaps it would have even less. The very broad radiating surface is too large to maintain phase coherence at any single listening position. If you want something totally phase coherent, you need to look at a full-range driver speaker, and not one with a huge surface area.

El panels still measure far better for phase than conventional speakers. Even the Stereophile measurements at close proximity bear that out.

It's because when seated afar from the panel it's mostly coherent - apart from the treble. For a 4ft tall panel, when seated at 10 ft, 3/4 of the panel is effectively phase coherent up to about 3kHz.

 

[Bucknekked]

Anyone who wants can participate in a single or double blind listening test. There are no requirements that I know of.

Of course a philosopher could argue that anyone who actually does want to participate in such a test probably has preconceived notions that might or might not introduce a testing bias. By this thinking, the best participants would have to be unaware of the true nature of the question involved.

It's my impression, based on reading Toole's book a few years ago, that blind listening tests of loudspeakers using experienced or inexperienced listeners produced little difference. If anything, experienced listeners performed slightly less reliably than inexperienced listeners.
Sorry for the late answer.

Swerd
I am involved in reading Toole's 3rd edition book right now and thankfully I have covered the testing scenarios section. What he said about experienced vs inexperienced users is that inexperienced users can be trained to listen critically in very short order. In other words, listening aint rocket science and with a few helpful hints and tips a relative newb can perform as consistently as someone who is an experienced listener in relatively few sessions.

Can you "want" to participate in a DBT without fouling the test. Of course you can. It happens all the time. Without people who want to listen critically to audio the Harmon testing center would be screwed. The genius of the DBT is neither the testor (the guy or gal running the test sequences) nor the listeners know what is being measured and tested. You can be gung ho as hell about being part of the test, but, you will not know what you are testing or listening for or what variables are being changed behind the blind curtain.

In a true DBT, you don't know if it is source music that's being tested, speakers, amplifiers, cables, connectors, or any of 100 or so things we can do audible tests for. You can't see what's being tested and you don't know the order or even if things are being changed behind the curtain.

If more people who post on this forum about their golden ear ability to discern between A-B testing understood how flawed a method that is, they wouldn't be crowing so loud about how they can hear the difference between a 16bit DAC and a 24 bit DAC when they can see both and know which one is playing and that's what is being tested.

anyway, that's my two cents. I am loving Floyd Toole's 3rd edition book. It should be required reading material. For whom, I'm not sure. But it should be required reading.