No one has yet figured out how to measure it, at least no-one willing to show others.
It is still one of those things that still requires the ear/brain.
It looks like soon, if not already, we will have brain reading devices that will allow us to bypass many of the issues with subjective testing which will really help us get over the subjective.objective divide.
dave
It is still one of those things that still requires the ear/brain.
It looks like soon, if not already, we will have brain reading devices that will allow us to bypass many of the issues with subjective testing which will really help us get over the subjective.objective divide.
dave
The right speed... not too much fast not too much slow
Certainly a driver able to play the fundamental and harmonics it is playing as well as the harmonics of the driver below it is also playing with no phase issue... i.e. correct damping ? (I like the Qtc concept to explain that). And not sure a Full range are always the better at that game. It could be interesting to compare for instance an ESL63 with its delay network and a classic ESL both in relation to the op question.
.
Certainly a driver able to play the fundamental and harmonics it is playing as well as the harmonics of the driver below it is also playing with no phase issue... i.e. correct damping ? (I like the Qtc concept to explain that). And not sure a Full range are always the better at that game. It could be interesting to compare for instance an ESL63 with its delay network and a classic ESL both in relation to the op question.
.I'll start with the last statement 1st (and follow backward from there):
Because you are NOT interested in a DSP active speaker: AVOID a dipole design. Even for experienced designers, doing a passive dipole design can be very difficult - especially if you want something that isn't extremely in-efficient.
Plus: when using a driver without any cabinet (sometimes referred to as "velocity drive") most of what is referred to as "speed" is absent, with the major driver factor of "speed" being a low Qe (nearing .2).
Further - a dipole isn't a good design for avoiding room effects (except at the lowest freq.s).
IF you are having problems with the room then utilize a design you can "pull-out" into the room for critical listening. (..even SL would do that for his LX mini, which was in large measure designed to mitigate placement near walls.) IF you can't do that then look to Keyser's work here:
2-way: Waveguide + Cardioid-like
This was the design before moving onto the Dutch & Dutch loudspeaker.
You can combine a design like this with an active subwoofer(s) solution and achieve a fairly room in-sensitive result.
...
The AT driver is reasonably efficient, has low Mms for its Sd, has low Qe, and particularly goes well with the specific Ribbon tweeter manufacturers I mentioned (..because of the way those ribbon tweeters are suspended in the magnetic gap - having a fair bit of mechanical damping for diaphragm that isn't damped on its "edges".) I also know of a successful design using the RAAL and an AT driver.
The VC is itself an inductor in series: that forms a low-pass filter reducing output higher in freq.. (..driver's often counter this with a mechanical response that is more efficient at higher freq.s.) Higher Inductance at higher freq.s (around 20 kHz +) can also increase the driver's motion EMF (..back emf) magnetically damping the forward and reward pulse of the driver, BUT this largely requires the VC to leave (or almost leave) the magnetic gap, and this only happens if the VC is an over-hung design (or is over-driven beyond xmax). The Former also plays a role here, but it's very minor by comparison.
Last edited:
I would be interested in dsp, but if it came before the DAC, so I can use my a lot more expensive converter and tube preamp. I had in mind creating a passive dipole with drivers of similar sensitivity, with as few electronic parts as possible, possibly two and a half way configuration (one low, one mid and a tweeter), and then equalizing the whole speaker frequency , as it was just a one driver. I'd be happy to achive 40ish Hz and then add a sub perhaps.
Can you elaborate the part with 'when use a driver without any cabinet (sometimes referred to as "velocity drive") most of what is referred to as "speed" is absent, with the major driver factor of "speed" being a low Qe (nearing .2).' a bit further?
I've been listening to a variety of conventional speakers in several different rooms and, honestly, they don't sound half as good as they are designed to. Pull away from the walls, you get nice mids and soundstage, but inconsistent bass, with bass 'holes' everywhere around the room. Bring them closer to walls, all you hear is bass. Maybe a modular speaker, where mids and tweeter would be further from walls, and low freq.unit closer (like infinity iors), but I have no idea how to time allign them. I will look into your waveguide recommendation.
Raal seems interesting, especially since its localy produced here, it should be possible to source. I'm not sure if such vertical dispersion would go against SL idea of constant directivity. Thanks for your explanation on inductance.
Can you elaborate the part with 'when use a driver without any cabinet (sometimes referred to as "velocity drive") most of what is referred to as "speed" is absent, with the major driver factor of "speed" being a low Qe (nearing .2).' a bit further?
I've been listening to a variety of conventional speakers in several different rooms and, honestly, they don't sound half as good as they are designed to. Pull away from the walls, you get nice mids and soundstage, but inconsistent bass, with bass 'holes' everywhere around the room. Bring them closer to walls, all you hear is bass. Maybe a modular speaker, where mids and tweeter would be further from walls, and low freq.unit closer (like infinity iors), but I have no idea how to time allign them. I will look into your waveguide recommendation.
Raal seems interesting, especially since its localy produced here, it should be possible to source. I'm not sure if such vertical dispersion would go against SL idea of constant directivity. Thanks for your explanation on inductance.
Last edited:
The mid solution you are after is difficult to accomplish (200Hz upwards). It is actually not needed and possibly detrimental. The lower the crossover frequency to the bass, the greater the timing difference with the mid which might go against your search for a natural representation. However, Wilsons have their devotees but in one demo I had, if you don't sit correctly, the time alignment disintegrates.
In practice the mid needs to get down to 600Hz - 700Hz approx. This increases options. 200Hz requires more driver excursion, typically resulting in more mass. An agile driver should get to 6KHz, allowing crossover at 4KHz approx. There are not a lot of options. Ribbon or horns can manage that. Short throw 4 inch cone drivers don't seem popular, but would be a good option if you can find one.
In practice the mid needs to get down to 600Hz - 700Hz approx. This increases options. 200Hz requires more driver excursion, typically resulting in more mass. An agile driver should get to 6KHz, allowing crossover at 4KHz approx. There are not a lot of options. Ribbon or horns can manage that. Short throw 4 inch cone drivers don't seem popular, but would be a good option if you can find one.
I've done passive dipole. Since you are flattening the falling response, sensitivity depends on how low in frequency you want to go.
Some use a woofer with a peaking lower response.. However you can get the most low end if you bi-amp just so you can turn up the speaker separately. It can be made to work with any type of crossover.
Eg just one option, tilt the falling response with passive line level and the top end with passive speaker level..
Yes, you can do DSP active before the DAC (..in fact that's what I use when prototyping crossovers). You can also add to this after the DAC, as traditionally when adding a subwoofer(s) (..where you don't care as much about the quality of the signal.)
For multi-channel generally: go for a good audio interface with multiple outputs (and usually a front-end computer with software for the digital manipulation/crossover).
For something more esoteric: look to the nanoShark and split out the digital from there to DAC's of choice. (..I'd look to the Soekris DAC boards.)
https://www.minidsp.com/images/documents/Product Brief-nanoSHARC.pdf
Soekris Audio ApS, Products
...
Dipole operation of dynamic drivers usually don't sound "fast" (or "slow"). There is however some small degree of character imparting a "fast" sound when using dynamic drivers with very low Qe around .2 (..and conversely a bit "slower" sounding once you get past about .8 Qe).
...
The problem you are describing is largely a subwoofer (though really it's throughout the bass region) problem and the desire to have sound uniform throughout the room to multiple listeners where-ever they happen to be in the room. (..you can also experience substantive modal room problems higher in freq. into the midrange depending on how small the room is, plus you might find floor-bounce more audible.)
You can really only deal with that by using the multi-sub approach that markbakk mentions. Search for "Geddes multiple subwoofers" on google.
Last edited:
I have fast car speakers, clocked at 252km/h at real speed
)Due to Covid crap, I have been playing with my speakers. Managed to get 35db dip at acoustical x-over, compared to previous 23db dip. Sounds like resolution improved! Speakers have not moved, so they must be slow
But nothing can be faster than the input signal driving it, over the bandwidth it has been filtered to play. This goes for both starting and stopping. If either were to deviate from the signal (ie go faster than) then you'd have a gross source of something that sounds extremely bad.
If you've got a frequency response, tailored to your desired target, and the harmonic distortion throughout that target range is very low then the driver is as fast as it needs to be. Providing it does it over the full range of off axis angles that you require.
Another driver cannot do that range any faster than the previous. If it tried to go faster then it would probably be breaking any number of the laws of physics.
As it stands Paul at PS Audio is nothing more than a shill. Most of what he talks about is nonsense.
In my opinion the 'fast' phenomena is nothing more than effortless reproduction. That is low power compression, low harmonic distortion, and speakers that have very low colouration due to stored energy within the cabinets.
Of course lots of speakers with limited low frequency extension are often described as 'fast' sounding but this is because of the perceived effects of having very little low bass.
If you've got a frequency response, tailored to your desired target, and the harmonic distortion throughout that target range is very low then the driver is as fast as it needs to be. Providing it does it over the full range of off axis angles that you require.
Another driver cannot do that range any faster than the previous. If it tried to go faster then it would probably be breaking any number of the laws of physics.
As it stands Paul at PS Audio is nothing more than a shill. Most of what he talks about is nonsense.
In my opinion the 'fast' phenomena is nothing more than effortless reproduction. That is low power compression, low harmonic distortion, and speakers that have very low colouration due to stored energy within the cabinets.
Of course lots of speakers with limited low frequency extension are often described as 'fast' sounding but this is because of the perceived effects of having very little low bass.
I guess I took for granted that a driver that starts and stops quickly (closely follows the input signal) would make for a natural representation of sound. By doing some more reading, especially on crossovers, I see that a phase coherence might have to do more with that.
So, I'd like to rephrase the thread - would a phase coherence give an ultimate realism in reproduction and what other factors have a role in that?
I'm having trouble finding anything smaller than 5 inch. The other problem might be crossing from that mid/midtweeter to a tweeter, since all those drivers don't have a lot of frequency overlap with a ribbon or most small domes. It would need a steep slope to cross, which means more phase difference.
In fact, the closest I get to your solution is a small full range driver, that has a decent overlap to a bigger woofer and perhaps could be crossed at 400-500hz. Maybe a first order crossover to avoid phase problems and then let the full range to play the rest with no tweeter.
So, I'd like to rephrase the thread - would a phase coherence give an ultimate realism in reproduction and what other factors have a role in that?
I'm having trouble finding anything smaller than 5 inch. The other problem might be crossing from that mid/midtweeter to a tweeter, since all those drivers don't have a lot of frequency overlap with a ribbon or most small domes. It would need a steep slope to cross, which means more phase difference.
In fact, the closest I get to your solution is a small full range driver, that has a decent overlap to a bigger woofer and perhaps could be crossed at 400-500hz. Maybe a first order crossover to avoid phase problems and then let the full range to play the rest with no tweeter.
Last edited:
In simple terms: frequency response and phase are in a single driver interlinked: change the frequency response and/or stopband slopes and the phase will change accordingly. Two drivers having exactly the same transfer function, say e.g. 400 Hz-3000Hz with 12 dB roll-off on either side of the passband will have exactly the same phase response. You cannot have one without the other.
Look at so called Bode diagrams for all this, or experiment in Vituixcad.
Look at so called Bode diagrams for all this, or experiment in Vituixcad.
When i was comparing 'slow' sub to 'fast' sub, it turned out slow sub was one note samba with lots of distortion, while fast sub was quite flat and quite clean.
If i were to eq both of these for exactly the same fr response and drive them very little, so neither would distort, i assume the distinction would be gone. I assume, i had not done that. It would be low volume, so not practical.
Would something similar apply to mids? I suppose so.
Good sounding mid to me is primarily the one free of nasty breakups. Is it fast? I think this term is not applicable. Unless you exactly define what exactly 'fast' means, most of the discussion is not leading to right conclusion. You want fast midrange? Drive it in lamborghini, sorry, bad joke.
If i were to eq both of these for exactly the same fr response and drive them very little, so neither would distort, i assume the distinction would be gone. I assume, i had not done that. It would be low volume, so not practical.
Would something similar apply to mids? I suppose so.
Good sounding mid to me is primarily the one free of nasty breakups. Is it fast? I think this term is not applicable. Unless you exactly define what exactly 'fast' means, most of the discussion is not leading to right conclusion. You want fast midrange? Drive it in lamborghini, sorry, bad joke.
Have you ever thought it odd that all the letters he opens are in identical blank envelopes?
Yep since the beginning it is a filter and amp question more. People want loudspeakers than can handle fast dynamic gaps while the recordings are more and more compressed.
Making rhe group delay clean between the mid and the tweeter should be enough to answer to the question...below the Schroeder frequency so bass driver it is more trickyt but nothing impossible for a good enough quality.
Making rhe group delay clean between the mid and the tweeter should be enough to answer to the question...below the Schroeder frequency so bass driver it is more trickyt but nothing impossible for a good enough quality.
It is still a big question. This is something typically brought by a single cone FR driver, and many love theirs despite their shortcomings.
dave