I'm just offering the altenative viewpoint from most others. People are still singing the praises of high-efficiency speakers, which have their virtues but not for your application. I propose you'd be happier in the enigneering trade-offs if you just decided up-front to say the heck with efficiency and figure out how to reap benefits in every other category, like transient response, bandwidth, box size, frequency response (with EQ) etc. For one thing, you can have incredible bass without the big boxes; isobaric is perfect for your requirements IMHO.
I don't have the time to tweak a crossover, nor the inclination. Nor do I want speakers that are all in 1 box. It's a down-sizing from my PA days, but I expect my sound system to be of modular components and a flexible commercial active crossover and EQ.
Good luck, have fun!
Last edited:
We're not talking about horns here so the spread between inefficient and efficient drivers is about 10dB. Reducing the power amp's voltage gain to unity from the typical 25 to 30dB increases the signal by, well, 25 to 30dB without imposing any restrictions on driver solution or having to deal with a reduced signal level point in the amp frontend and the corresponding SnR decrease (it also lowers the Johnson noise in the feedback network by around 12dB if the resistor values are well chosen). If you spend an extra two cents on resistors to make the amp's feedback loop differential and route it with Kelvin taps there's an extra 40dB CMRR to be had (60dB if you spend the extra $1.50 for 0.1% feedback resistors).
As an example, consider a low SPL speaker design with typical 90dB/W efficiency triamped with LME49600s to get 1W peak per speaker. Levels---both input and output since it's unity gain---will be typically few hundred mV and the output noise floor is -140dB. So the SNR is on the order of 120dB. If that's not good enough, you can switch to 85dB efficient drivers and the somewhat quieter LMH6321 (with the CL pin left open to get the necessary increase in drive current), but keep in mind 120dB down from 90dB is -30dB SPL and the quietest audible sound is defined to be 0dB SPL. Exceptional individuals can hear 10 or maybe even 15dB below this but even they would need to have their ears within four inches of the speaker in an environment with negligible ambient noise to pick up on the difference.
Last edited:
I believe the frequency response is the critical factor in this matter (boring, I hear some saying to this but this is no simple parameter). The closer it is to correct for your system, the easier it is to hear the performers as you would naturally. When there is an anomaly in the response it limits our ability to see through the reproducing system. (We may well see what's on the other side of a smudged window.. there is no experience like looking through a spotless window, like a quality pair of shades.)
I'm not referring to loudness contours. I'm not discounting them either, there may possibly be more than one 'right', but I don't believe you can choose an arbitrary curve just because it suits the music at hand, and still expect transparency.
Though recordings are different, your speakers are mostly a constant. In choosing the correct curve, you are primarily limited by your speakers and room. The major factors as I see them are as follows:
1. The total of all sound at all angles put out by the speakers.
2. Your on-axis response.
3. Whether the room is involved (boundary cancellations shouldn't be blanket equalised)
4. The time before off-axis sound reaches you vs. direct sound
5. The frequency in question (re: the need to control reflections and the sensitivity to issues vs. frequency)
6. The reverberation time at a given frequency.
7. The nature of the anomalies - wide/shallow, narrow/deep, how much, where. etc.
All of the above points are related to the amount of exposure we receive at a given frequency, considering our sensitivity to high vs. low frequencies, and direct sound vs. early and late reflections. This means that the required curve may or may not appear logical at first. I am talking about fitting this in a window of the necessary curve, whatever that may be, to within 1dB over a broad range.
Speakers therefore either require comprehensive measurement and theoretical analysis to reach the point where they can uncover and reveal themselves enough to be listened to critically at any level. Controversially though, I believe that some of us can get pretty close just using our ears...but it takes practice
. I use both and consider both to be important ... BTW not just any speaker/room can be coaxed into sounding excellent. Furthermore, if your speakers sound worse when you turn them up, you'll have to deal with that...
In addition I feel that the power response (the overall sound produced by the speakers) into the room is more important than the on-axis response in achieving very good sound reproduction, more even than contributing factors such as the on-axis phase response. I believe this needs to be spot on though if you want to go from very good to excellent, but there's little point in obsessing with it if you don't have the other factors under control. (This is system dependent).
A final word on the quality of recordings. They differ considerably. I have many times equalised (on the fly) a vocalist to relieve sibilants (a process that can involve more than just the top end)...only to find that the rest of the recording is thrown out (as different tracks within a recording can have their own 'emphasis' added by the engineer). I suspect some studio monitors, some studios, and some sound engineers are biased in one form or another., sometimes for reasons out of the normal realm of control, and sometimes of their own choice....
So, I've heard it said that speakers need to roll off a little (or other things) to compensate for poor recordings, but I would debate this. Sometimes it is just a compromise to make a bad recording tolerable. In the end, don't base your assessment of your system on a single recording, good or bad, and don't compensate for a bad recording unless you're listening to nothing else, because it may possibly not be correct for any recording.
I'm not referring to loudness contours. I'm not discounting them either, there may possibly be more than one 'right', but I don't believe you can choose an arbitrary curve just because it suits the music at hand, and still expect transparency.
Though recordings are different, your speakers are mostly a constant. In choosing the correct curve, you are primarily limited by your speakers and room. The major factors as I see them are as follows:
1. The total of all sound at all angles put out by the speakers.
2. Your on-axis response.
3. Whether the room is involved (boundary cancellations shouldn't be blanket equalised)
4. The time before off-axis sound reaches you vs. direct sound
5. The frequency in question (re: the need to control reflections and the sensitivity to issues vs. frequency)
6. The reverberation time at a given frequency.
7. The nature of the anomalies - wide/shallow, narrow/deep, how much, where. etc.
All of the above points are related to the amount of exposure we receive at a given frequency, considering our sensitivity to high vs. low frequencies, and direct sound vs. early and late reflections. This means that the required curve may or may not appear logical at first. I am talking about fitting this in a window of the necessary curve, whatever that may be, to within 1dB over a broad range.
Speakers therefore either require comprehensive measurement and theoretical analysis to reach the point where they can uncover and reveal themselves enough to be listened to critically at any level. Controversially though, I believe that some of us can get pretty close just using our ears...but it takes practice
. I use both and consider both to be important ... BTW not just any speaker/room can be coaxed into sounding excellent. Furthermore, if your speakers sound worse when you turn them up, you'll have to deal with that...In addition I feel that the power response (the overall sound produced by the speakers) into the room is more important than the on-axis response in achieving very good sound reproduction, more even than contributing factors such as the on-axis phase response. I believe this needs to be spot on though if you want to go from very good to excellent, but there's little point in obsessing with it if you don't have the other factors under control. (This is system dependent).
A final word on the quality of recordings. They differ considerably. I have many times equalised (on the fly) a vocalist to relieve sibilants (a process that can involve more than just the top end)...only to find that the rest of the recording is thrown out (as different tracks within a recording can have their own 'emphasis' added by the engineer). I suspect some studio monitors, some studios, and some sound engineers are biased in one form or another., sometimes for reasons out of the normal realm of control, and sometimes of their own choice....
So, I've heard it said that speakers need to roll off a little (or other things) to compensate for poor recordings, but I would debate this. Sometimes it is just a compromise to make a bad recording tolerable. In the end, don't base your assessment of your system on a single recording, good or bad, and don't compensate for a bad recording unless you're listening to nothing else, because it may possibly not be correct for any recording.
Last edited:
A lot of good ideas, again
If I choose to go FR, Lowthers are surely on my shortlist. Never heard them, though, cabinets are rather big and I'd need a new amp... Seems a bit risky and too much compromise for the price. But hearing Omega Super Alnico 8's is what got me into hifi in the first place, so quality FR's are never off the table.
DEQX is 2k$ at cheapest, plus customs, shipping and 24% VAT; not much left for the speakers there. Are there any cheaper options for an active XO+EQ+room correction?? One that would work with an integrated amp as well?
Maybe this: DSPeaker Anti-Mode 2.0 Dual Core. They say it can be used as an active XO (plus it's a room corrector, dac and eq to begin with). But in the signal chain, can you place XO before amp?? The plan would be: transport/pc - Anti-mode - amp - speakers, can it be done?? About 800€, I'd still have enough for drivers etc, plus I'd get the eq too. I emailed them, we'll see what they have to say.
About open baffles: they need to be big, right: large drivers on a large baffle. Is there a way around this, all OB's I've seen are far too big for our livingroom? A modular design (or small OB) plus an active XO (+eq and/or room correction) is certainly an option. Though it leaves me with choosing the right drivers and everything... Plus potentially expensive changes to signal path (ie. separate pre and power amps).
A lot of fine options to choose from.
[FONT=Verdana, Arial, Helvetica, sans-serif][/FONT]
If I choose to go FR, Lowthers are surely on my shortlist. Never heard them, though, cabinets are rather big and I'd need a new amp... Seems a bit risky and too much compromise for the price. But hearing Omega Super Alnico 8's is what got me into hifi in the first place, so quality FR's are never off the table.
DEQX is 2k$ at cheapest, plus customs, shipping and 24% VAT; not much left for the speakers there. Are there any cheaper options for an active XO+EQ+room correction?? One that would work with an integrated amp as well?
Maybe this: DSPeaker Anti-Mode 2.0 Dual Core. They say it can be used as an active XO (plus it's a room corrector, dac and eq to begin with). But in the signal chain, can you place XO before amp?? The plan would be: transport/pc - Anti-mode - amp - speakers, can it be done?? About 800€, I'd still have enough for drivers etc, plus I'd get the eq too. I emailed them, we'll see what they have to say.
About open baffles: they need to be big, right: large drivers on a large baffle. Is there a way around this, all OB's I've seen are far too big for our livingroom? A modular design (or small OB) plus an active XO (+eq and/or room correction) is certainly an option. Though it leaves me with choosing the right drivers and everything...
Plus potentially expensive changes to signal path (ie. separate pre and power amps).A lot of fine options to choose from.
[FONT=Verdana, Arial, Helvetica, sans-serif][/FONT]
The miniDSP should do what you need at a reasonable cost. There is even a miniDSP forum here.
OB is great, but does take space for the baffles and the placement. Not the most practical apartment arrangement. Unless you've heard the Lowthers and loved them, I 'd not go there. Some folks love them, some hate them.
I still think that small, fullrange satellites and a small sub would be the ticket. A DIY version of the John Blue JB3.
OB is great, but does take space for the baffles and the placement. Not the most practical apartment arrangement. Unless you've heard the Lowthers and loved them, I 'd not go there. Some folks love them, some hate them.
I still think that small, fullrange satellites and a small sub would be the ticket. A DIY version of the John Blue JB3.
Full range satellites and a sub, I just looked into that. You mentioned it earlier, thanks I like FR's, so this might be a way around their shortcomings.
So you mean an about 4"-5" FR driver, no XO, in a closed or BR enclosure? Sounds good to me.
And the separate subwoofer then, what needs to be considered about the integration? My amp has 4 pairs of speaker outputs but no subwoofer output. So I'd need one (or two) active-subwoofer, with it's own amp and low-pass filter, then I could connect it to my amp, right?
If I'd use, say, 5" Fostex FR's, what should I use for sub?
And then: SnR. My current speakers are 84db and I can hear a "hiss" from about 1 foot away. Not a problem, but what about with 98db sensitivity?? Amp specs say SnR is 90db, which is good but not great. Would I need to get a new amp, too, or is there another way around this?
Also, I currently use volume settings from 6 o'clock to 8 o'clock position. With 98db sensitivity I'm left with what, 6 o'clock to 6:15? Is there something I can do about this? I've heard of attennuator cables, would that be an option?
I like FR's, so this might be a way around their shortcomings.So you mean an about 4"-5" FR driver, no XO, in a closed or BR enclosure? Sounds good to me.
And the separate subwoofer then, what needs to be considered about the integration? My amp has 4 pairs of speaker outputs but no subwoofer output. So I'd need one (or two) active-subwoofer, with it's own amp and low-pass filter, then I could connect it to my amp, right?
If I'd use, say, 5" Fostex FR's, what should I use for sub?
And then: SnR. My current speakers are 84db and I can hear a "hiss" from about 1 foot away. Not a problem, but what about with 98db sensitivity?? Amp specs say SnR is 90db, which is good but not great. Would I need to get a new amp, too, or is there another way around this?
Also, I currently use volume settings from 6 o'clock to 8 o'clock position. With 98db sensitivity I'm left with what, 6 o'clock to 6:15? Is there something I can do about this? I've heard of attennuator cables, would that be an option?
Yes; look at the thread linked in post 34 for one end of the design space---there are a variety of ways to reduce the size from there, though reducing the width below that of a 15" driver has a way of involving painful compromises in a living room speaker (gainphile's done a couple desktop OBs with 8" drivers).
@twest: I'm new to this whole swinging dipole thing, so I can't really comment. Building thread is long and rather technical for me... Looks interesting and my cats would definitely love them.
It's "just" drivers hanging from the frame, external crossover? Is there something special about the drivers, or would just any drivers do? What if I hung that Altec 604 from ceiling? Minimalist and bizarre, I love it.
It's "just" drivers hanging from the frame, external crossover? Is there something special about the drivers, or would just any drivers do? What if I hung that Altec 604 from ceiling? Minimalist and bizarre, I love it.
That's about 900€ plus I don't need to pay VAT when buying from a private person (well I would but I bet he/she could send it as gift). Not bad, given it's not outdated... But it only goes between pre and power, and I have an integrated, one that I really like, too. So miniDSP would probably be a lot easier way of doing this, right, no need to buy new amp? It's still an option, thanks for the tip, I'll look into it tomorrow
The vibration management benefits of isolation mounting apply to any pistonic driver. A decoupling suspension does not really help with box resonances or crosstalk between drivers, though it can mitigate some of the modes. Personally I'd look at putting the 604s on inverted pendulums (CLS's approach), swings, or bicycle tubes (Stig Erik) rather than hanging to create a simple pendulum. The difference between a swing and a regular pendulum's rather blurry though. From a sonic performance perspective it doesn't really matter how you get the decoupling. So it's more about personal preferences.
Basic crossover and EQ configuration is in the third post of the thread with some updates here. I've described the general case for magnetostats in this thread and the swinging build thread covers dipole specific considerations such as front to rear symmetry and magnet shadowing. If you have specific questions about the driver selection for that speaker and such just ask 'em in the thread and I'll pick it up.
Basic crossover and EQ configuration is in the third post of the thread with some updates here. I've described the general case for magnetostats in this thread and the swinging build thread covers dipole specific considerations such as front to rear symmetry and magnet shadowing. If you have specific questions about the driver selection for that speaker and such just ask 'em in the thread and I'll pick it up.
Sort of what i'm listening to right now. uFonkenSET with a pr of stealthWoofers borrowed from the Frugal-Horn Mk3.
It is pretty easy these days to do up a diy that outperforms the JB3.
dave
Yeah, lot's of nice small fullrangers to choose from. Put them in the right box and be happy. One nice thing about the JB3 is that it has very good tonal balance - which makes it sound much bigger. Superb imaging, too. Strict limitations on volume, tho.
If the <90Hz range can be handed off to a sub, it makes the job of the little FR a lot easier.
If the <90Hz range can be handed off to a sub, it makes the job of the little FR a lot easier.
Pano: I'm thinking of building one cabinet, with two separate compartments: approx 10 liter front ported BR for a 5" fullrange; approx 30 liters closed cabinet for a side firing 8" sub. Like, say, Penaudio Serenade. You see, we have these cats, and even small satellites would need a good footprint for stability. Might as well put them all in one box and save the extra sub, right?
Please tell me if this is a bad idea, I'm really not good at electronics... Just tell me if this could work at all, and how? Here goes:
My amp (2x70W) has 4 sets of speaker posts (bi-wire?), so I could run 4 passive speakers, right? In this idea I would have four speakers: two 5" FR's and two 8" woofers. Does it work like this??
Between the amp and woofer-driver, I'd put an adjustable low-pass filter, say 60hz-100hz. Between the amp and 5" FR I'd put an adjustable attennuator, to match it's level with the woofer.
Does my amp have enough juice to drive this all at civilised levels, what with the subs being passive (I'm thinking 8" Fostex rated at 90db)? Does the attennuator ruin the whole FR thing? Does the attennuator attennuate noise as well: if signal is attennuated 50%, do I get 50% noise too, or same 100% noise as unattennuated? Am I asking too many questions? Where/how can I get such low-pass filter? Is this at all how it works?
Are those side firing subs supposed to face inside or out? I also have a screen between the speakers, so I'm limited to max 50cm/1,5ft away from the side walls... Could be a problem if subs face the walls.
Please tell me if this is a bad idea, I'm really not good at electronics...
Just tell me if this could work at all, and how? Here goes:My amp (2x70W) has 4 sets of speaker posts (bi-wire?), so I could run 4 passive speakers, right? In this idea I would have four speakers: two 5" FR's and two 8" woofers. Does it work like this??
Between the amp and woofer-driver, I'd put an adjustable low-pass filter, say 60hz-100hz. Between the amp and 5" FR I'd put an adjustable attennuator, to match it's level with the woofer.
Does my amp have enough juice to drive this all at civilised levels, what with the subs being passive (I'm thinking 8" Fostex rated at 90db)? Does the attennuator ruin the whole FR thing? Does the attennuator attennuate noise as well: if signal is attennuated 50%, do I get 50% noise too, or same 100% noise as unattennuated? Am I asking too many questions? Where/how can I get such low-pass filter? Is this at all how it works?
Are those side firing subs supposed to face inside or out? I also have a screen between the speakers, so I'm limited to max 50cm/1,5ft away from the side walls... Could be a problem if subs face the walls.
If you are thinking of side-firing woofers you could use two per speaker in a force-cancelling arrangement a la Barefoot Sound.
Barefoot Sound
Barefoot Sound
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.