My sentiments exactly.
BTW, I agree completely with your first post too. Very refreshing to have around somebody like MaVo who is competent in a sea of self-aggrandizing technical-sounding nonsense posts.
But back to "fast." I guess I am concluding there is no such physical parameter to a driver (besides the usual and obvious relevant parameters). Rather, "fast" is a word applied to a perception... as MaVo suggested above, probably when you have lumpy, resonant bass, it sounds "sluggish." But then I am allergic to any tuned box whether bass-reflex or tapped horn.
I'll second that- no resonant peaking please!
Q should always be equal to or less than 0.577 of the driver in the box.
I'm not a big fan of EQ but maybe this is what the "shouty" upper mids of the lowthers need after all...can anyone suggest a half-decent unit that won't break the bank or a schematic they have tried? I would like to avoid any DSP - I want my signal to remain analogue from cartridge to coil.
Thanks
Thanks bentoronto, it truly felt like a lost struggle 
HM, firstly, you wont find the answers to physics questions in magazines like stereoplay. Those are industry paid add vehicles with a bare minimum of usefull content. For real knowledge, one has to read real books (or at least wikipedia). Secondly, while the membrane can have different speeds, that does not apply to music in the sense you want to give it. As bentoronto has said, speed is a term to describe our perception of music, which is not the same as the music itself. Music can have different speeds, but this is measured in beats per minute and this isnt affected by the speaker you have.
I do agree with you that horns have benefits, but those can be showed in quantitative and clear ways, such as directivity, distortion, diffraction and frequency response. Those quantifiable entities determine how we perceive the music. The physics of speakers are the link between music and perception. Dont neglect this.
HM, firstly, you wont find the answers to physics questions in magazines like stereoplay. Those are industry paid add vehicles with a bare minimum of usefull content. For real knowledge, one has to read real books (or at least wikipedia). Secondly, while the membrane can have different speeds, that does not apply to music in the sense you want to give it. As bentoronto has said, speed is a term to describe our perception of music, which is not the same as the music itself. Music can have different speeds, but this is measured in beats per minute and this isnt affected by the speaker you have.
I do agree with you that horns have benefits, but those can be showed in quantitative and clear ways, such as directivity, distortion, diffraction and frequency response. Those quantifiable entities determine how we perceive the music. The physics of speakers are the link between music and perception. Dont neglect this.
Last edited:
A good, clean second hand KT DN410 goes for around £400 ($575) around here. In fact there is a nice looking one on Gumtree at the mo'.
Also the Rane is a digital unit, something crOwl3y wanted to avoid.
Last edited:
Hard to argue with some parametric equalization. I find it pretty helpful dealing with a bump or two in my Klipschorn bass. I have an old kit with a bunch of channels to work with. But pretty simple to cobble one from an IC cookbook or maybe to revise a crossover into a parametric equalizer (nearly identical circuits but with pots instead of fixed resistors).
Hard to think of any value to horns in an audiophile setting, possibly bass excepted. From Klipsch on, they have always sounded awful to me and with all kinds of shortcomings. But some eq can me 'em better in respect to bumps and dips, I suppose.
Hard to think of any value to horns in an audiophile setting, possibly bass excepted. From Klipsch on, they have always sounded awful to me and with all kinds of shortcomings. But some eq can me 'em better in respect to bumps and dips, I suppose.
Check this, not to expensive: http://www.marchandelec.com/wm8.html
Never used his stuff but has a good reputation.
Never used his stuff but has a good reputation.
Very clever device.
I modified my little parametric equalizer kit to extend the range of some channels quite low. So practice, not a lot of difference from the Marchand although that is configured a bit more to the purpose.
On the other hand, with parametric-type controls you have a Q (bandwidth) control which the Marchand lacks (and maybe doesn't add much). With a mixed bass and patience, I guess the parametric device gives you more control.
Post #28 (Equalizer) and general thoughts.
Hi Charles,
You are definitely right, my oversight, the Rane still looks like a nice unit. Once you start adding equalizers to the signal chain the difference between analog and digital may not be all that important (?). Maybe a simple passive RCL filter/correction circuit could help? That would drastically reduce the added electronic complexity, and keep the signal in the analog domain.
Hi cr0wl3y:
As to matching the bass response of a back-loaded horn Lowther enclosure, a tapped horn would seem to make a lot of sense, the Acousta's horn output is delayed by definition, and it's impulse response (character) cannot possibly be close to a sealed enclosure. Finally, a tapped horn build around a 15" professional driver (e.g.: Eminence 4015LF) will get at least close to matching the sensitivity of the DX4s. Modelling the Acoustas, and a tapped horn in Hornresp might be a useful starting point.
Regards,
Hi Charles,
You are definitely right, my oversight, the Rane still looks like a nice unit. Once you start adding equalizers to the signal chain the difference between analog and digital may not be all that important (?). Maybe a simple passive RCL filter/correction circuit could help? That would drastically reduce the added electronic complexity, and keep the signal in the analog domain.
Hi cr0wl3y:
As to matching the bass response of a back-loaded horn Lowther enclosure, a tapped horn would seem to make a lot of sense, the Acousta's horn output is delayed by definition, and it's impulse response (character) cannot possibly be close to a sealed enclosure. Finally, a tapped horn build around a 15" professional driver (e.g.: Eminence 4015LF) will get at least close to matching the sensitivity of the DX4s. Modelling the Acoustas, and a tapped horn in Hornresp might be a useful starting point.
Regards,
Hi Oliver,
I have no (or very little anyway) that the Rane is a perfectly usable unit but it seems cr0wl3y has a completely analogue system and understandably wants to keep it that way.
KT are famous for their analogue eq and justly so, if one wants a transparent (ie non-euphonic) KT is as good as it gets.
That said I think the Ranes weak point is the lack of a digital in- and out put as I like to keep conversion steps (I include cartridges, rec/replay heads and speakers as well as digital ADDA here) to a minimum in the chain.
I think a tapped horn might be the best solution for a sub to match up with horns though.
I have no (or very little anyway) that the Rane is a perfectly usable unit but it seems cr0wl3y has a completely analogue system and understandably wants to keep it that way.
KT are famous for their analogue eq and justly so, if one wants a transparent (ie non-euphonic) KT is as good as it gets.
That said I think the Ranes weak point is the lack of a digital in- and out put as I like to keep conversion steps (I include cartridges, rec/replay heads and speakers as well as digital ADDA here) to a minimum in the chain.
I think a tapped horn might be the best solution for a sub to match up with horns though.
What I would build is a couple of dipole subs using high Qts, low Fs drivers in the 12" to 15" variety. Most likely, you'll have to look at pro drivers for this. These can be put into H-frames, W-frames or the more compact N-frame.
The best drivers that I can find for this application are still the Pyle Pro PPA12's and PPA15's. I used four of the PPA15's for my huge H-frame dipole subs (two per channel) years ago and to this day are still the best sounding subs I ever heard.
I still have those drivers sitting around. Maybe one day I'll build a pair of W or N-frames with these drivers and see if I can get similar results to the old H-frames, though needing a bit more EQ on the bottom end due to their smaller design.
The best drivers that I can find for this application are still the Pyle Pro PPA12's and PPA15's. I used four of the PPA15's for my huge H-frame dipole subs (two per channel) years ago and to this day are still the best sounding subs I ever heard.
I still have those drivers sitting around. Maybe one day I'll build a pair of W or N-frames with these drivers and see if I can get similar results to the old H-frames, though needing a bit more EQ on the bottom end due to their smaller design.
I think the common "a fast subwoofer is a mid driver" counterargument misunderstands the problem. The perception of 'fast' bass might be better characterized as 'immediate' bass.
Suppose a kick drum attack is recorded. The signal consists of a very large initial pressure wave created by the beater, and then the ringing of the drum. In order to reproduce this kick drum impact with a sense of immediacy, the system needs to proprortionately reproduce both the initial pressure wave and the ring exactly as they were recorded by the microphone. This becomes a problem when a loudspeaker relies heavily on resonant means to reinforce its low-frequency output, because this initial pressure wave contains information almost down to DC, but the loudspeaker's resonant means (port, passive radiator, transmission line, horn, daline, acoustic lever, etc.) may require several cycles to reach its steady-state condition. Every resonant system stores and releases energy over time, and the ratio of this stored energy to the released energy describes the resonant character of that system. We deal with this all the time in a loudspeaker; this is expressed in the "Q" parameter. When the ratio is greater than 1, more energy is stored over time than is released. The kick drum attack consists of a single cycle plus harmonics and ring-out, and the low frequency system has to be able to respond immediately rather than waiting several cycles to 'ramp up' to its steady state output level.
A friend of mine and I saw the correlation between the rise time of a resonant system and the subjective perception of 'fast' bass when we tried to apply mass-loading to a tapped horn design (much like the mass-loaded horn design sold by Gemme Audio as their "VFlex" bass loading scheme, discussed here some time ago, which mass-loads a rearloaded horn). By adding a mass element to the terminus of the tapped horn, we introduced an abrupt change in acoustic impedance at the mouth of the horn (and typically, you want smooth flare transitions in a horn to avoid exactly the phenomenon we hoped to exploit). The sudden increase in acoustic impedance causes part of the traveling wave to be reflected back toward the throat end of the horn, where it combines with the next wave being induced in the horn by the driver, with the magnitude of the combined wavefront being dependent on the relative phase between waves. This combined wave travels down the horn to the mass element, where again part of it is reflected and combines with the next wave from the driver, and so on. This creates a resonant system with a rise time of several cycles, but very high efficiency at very low frequencies (and the design was useful for loading drivers below Fs.) We built and tested a prototype that used a single 10" driver, and found that by introducing the mass element and increasing the mass, we increased the rise time of the system (indicated in the hornresp simulation by a rise in group delay, though high group delay is just one symptom of a resonant system with a long rise time), but also the subjective 'woofy' or 'slow' character, establishing a direct experimental correlation between rise time and subjective 'speed'.
A loudspeaker and its enclosure may not be entirely to blame for 'slow' bass in a room. The room is also part of the acoustic system, and has its own rise time and decay characteristics, based on the amount of low frequency damping provided by the room's furnishings (or purpose-built acoustic treatments for low frequency waves), and if a room has poorly-damped modal behavior or closely-spaced modes, this can make even a sub that might sound 'fast' in another room sound 'slow' in that particular room due to the rise time of the room itself independent of the acoustic source exciting it.
The way to achieve bass with a short 'rise time' is to first design a low-frequency system that itself has a very short rise time (a critically-damped sealed alignment with Qtc of 0.5 or less would work well but have low sensitivity, so you'll need multiple drivers and plenty of power), and then apply something like the Geddes multi-sub approach together with room EQ to confound the room's own modal behavior and compensate for whatever you can't eliminate with the multi-sub approach by applying the EQ, so that the combined output of the multiple subbass sources is roughly in-phase at the listening position and the average output of the multiple sources is relatively free of modal reinforcement and the summation of the room pressurization is even throughout the room.
That's the way I see it. My experience with the physics of loudspeaker enclosures and rooms confirms the existence of 'fast' bass, but this is not to say that the driver cone needs to necessarily be lightweight or short-excursion to achieve it - the system just needs to be well-damped and able to capture the initial pressure wave in proportion with the ring-out, instead of delaying the resonant system's steady-state contribution or that of the room by several cycles. This is bass without shortcuts; it will take some serious radiating area and excursion to make up for what is typically gained from resonant behavior, even in a sealed enclosure. It would be advantageous to design a low-frequency system with excellent extension up into the midbass, just because the subwoofer does make some contribution up there, even when the crossover frequency is set in the subbass range, because most subwoofer crossovers aren't 'brick walls'.
Cheers,
Rory
Suppose a kick drum attack is recorded. The signal consists of a very large initial pressure wave created by the beater, and then the ringing of the drum. In order to reproduce this kick drum impact with a sense of immediacy, the system needs to proprortionately reproduce both the initial pressure wave and the ring exactly as they were recorded by the microphone. This becomes a problem when a loudspeaker relies heavily on resonant means to reinforce its low-frequency output, because this initial pressure wave contains information almost down to DC, but the loudspeaker's resonant means (port, passive radiator, transmission line, horn, daline, acoustic lever, etc.) may require several cycles to reach its steady-state condition. Every resonant system stores and releases energy over time, and the ratio of this stored energy to the released energy describes the resonant character of that system. We deal with this all the time in a loudspeaker; this is expressed in the "Q" parameter. When the ratio is greater than 1, more energy is stored over time than is released. The kick drum attack consists of a single cycle plus harmonics and ring-out, and the low frequency system has to be able to respond immediately rather than waiting several cycles to 'ramp up' to its steady state output level.
A friend of mine and I saw the correlation between the rise time of a resonant system and the subjective perception of 'fast' bass when we tried to apply mass-loading to a tapped horn design (much like the mass-loaded horn design sold by Gemme Audio as their "VFlex" bass loading scheme, discussed here some time ago, which mass-loads a rearloaded horn). By adding a mass element to the terminus of the tapped horn, we introduced an abrupt change in acoustic impedance at the mouth of the horn (and typically, you want smooth flare transitions in a horn to avoid exactly the phenomenon we hoped to exploit). The sudden increase in acoustic impedance causes part of the traveling wave to be reflected back toward the throat end of the horn, where it combines with the next wave being induced in the horn by the driver, with the magnitude of the combined wavefront being dependent on the relative phase between waves. This combined wave travels down the horn to the mass element, where again part of it is reflected and combines with the next wave from the driver, and so on. This creates a resonant system with a rise time of several cycles, but very high efficiency at very low frequencies (and the design was useful for loading drivers below Fs.) We built and tested a prototype that used a single 10" driver, and found that by introducing the mass element and increasing the mass, we increased the rise time of the system (indicated in the hornresp simulation by a rise in group delay, though high group delay is just one symptom of a resonant system with a long rise time), but also the subjective 'woofy' or 'slow' character, establishing a direct experimental correlation between rise time and subjective 'speed'.
A loudspeaker and its enclosure may not be entirely to blame for 'slow' bass in a room. The room is also part of the acoustic system, and has its own rise time and decay characteristics, based on the amount of low frequency damping provided by the room's furnishings (or purpose-built acoustic treatments for low frequency waves), and if a room has poorly-damped modal behavior or closely-spaced modes, this can make even a sub that might sound 'fast' in another room sound 'slow' in that particular room due to the rise time of the room itself independent of the acoustic source exciting it.
The way to achieve bass with a short 'rise time' is to first design a low-frequency system that itself has a very short rise time (a critically-damped sealed alignment with Qtc of 0.5 or less would work well but have low sensitivity, so you'll need multiple drivers and plenty of power), and then apply something like the Geddes multi-sub approach together with room EQ to confound the room's own modal behavior and compensate for whatever you can't eliminate with the multi-sub approach by applying the EQ, so that the combined output of the multiple subbass sources is roughly in-phase at the listening position and the average output of the multiple sources is relatively free of modal reinforcement and the summation of the room pressurization is even throughout the room.
That's the way I see it. My experience with the physics of loudspeaker enclosures and rooms confirms the existence of 'fast' bass, but this is not to say that the driver cone needs to necessarily be lightweight or short-excursion to achieve it - the system just needs to be well-damped and able to capture the initial pressure wave in proportion with the ring-out, instead of delaying the resonant system's steady-state contribution or that of the room by several cycles. This is bass without shortcuts; it will take some serious radiating area and excursion to make up for what is typically gained from resonant behavior, even in a sealed enclosure. It would be advantageous to design a low-frequency system with excellent extension up into the midbass, just because the subwoofer does make some contribution up there, even when the crossover frequency is set in the subbass range, because most subwoofer crossovers aren't 'brick walls'.
Cheers,
Rory
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.