Hi,
"I suspected this argument would come up" is a misunderstanding really, or certainly it is in relation to my comments, as I am not arguing about anything here.
I just happen to agree with Panomaniac's earlier comment about there not being simply one road to success with such transducers as loudspeakers, and I don't believe that the perfect speaker has yet been (nor ever will be!) made.
I have been listening to, playing around with & modifying, and designing & building speakers for over 50 yrs, and, *in the main*, the best results have been when cabinet colourations have been kept to a minimum, *mainly* as they are so hard to control properly.
However, whichever way a drive unit is supported (i.e. they cannot just float in the air where we want them, unfortunately) there will always be some associated effects, mostly unwelcome, like box resonances, baffle-step effects, diffraction and so on.
Also, apart from a very few exceptions (like possibly Heil air-motion/plasma type of units), the various parts making up the drive-units themselves will not be perfect, and their basic materials will resonate and/or reflect sounds etc., like it or not.
In my opinion, the trick is to integrate any such effects (which canot be fully eliminated) into the design as a whole, so that the overall subjective performance is not compromised too much.
Without trying every possible material which could be used in speaker-construction together with every possible drive-unit on the market, I am simply not prepared to state, unequivocally, that certain clever combinations of materials/drivers cannot reproduce sounds as accurately and naturally as, or even better than, any other design.
As mentioned, I have not heard them for myself, but it is possible that well-respected makers like Bosendorpher have come up with something which is very good, and have successfully taken advantage of various effects resulting from the materials which they use in manufacture of their speakers. If this is so, it will doubtless be the result of trading-off any such cabinet effects against carefully-chosen driver characteristics, and by utilising cleverly-designed X'overs.
Regards,
"I suspected this argument would come up" is a misunderstanding really, or certainly it is in relation to my comments, as I am not arguing about anything here.
I just happen to agree with Panomaniac's earlier comment about there not being simply one road to success with such transducers as loudspeakers, and I don't believe that the perfect speaker has yet been (nor ever will be!) made.
I have been listening to, playing around with & modifying, and designing & building speakers for over 50 yrs, and, *in the main*, the best results have been when cabinet colourations have been kept to a minimum, *mainly* as they are so hard to control properly.
However, whichever way a drive unit is supported (i.e. they cannot just float in the air where we want them, unfortunately) there will always be some associated effects, mostly unwelcome, like box resonances, baffle-step effects, diffraction and so on.
Also, apart from a very few exceptions (like possibly Heil air-motion/plasma type of units), the various parts making up the drive-units themselves will not be perfect, and their basic materials will resonate and/or reflect sounds etc., like it or not.
In my opinion, the trick is to integrate any such effects (which canot be fully eliminated) into the design as a whole, so that the overall subjective performance is not compromised too much.
Without trying every possible material which could be used in speaker-construction together with every possible drive-unit on the market, I am simply not prepared to state, unequivocally, that certain clever combinations of materials/drivers cannot reproduce sounds as accurately and naturally as, or even better than, any other design.
As mentioned, I have not heard them for myself, but it is possible that well-respected makers like Bosendorpher have come up with something which is very good, and have successfully taken advantage of various effects resulting from the materials which they use in manufacture of their speakers. If this is so, it will doubtless be the result of trading-off any such cabinet effects against carefully-chosen driver characteristics, and by utilising cleverly-designed X'overs.
Regards,
Hi SD,
O.K., I now understand what happened here, and I was certainly not trying to cross swords with anyone, especially at this time of year.
As I attempted to make clear, *in the main* I agree entirely that the lowest cabinet 'reinforcement' of the sound is likely to be the best way of achieving accuracy, which as we both know is not the same as achieving a 'romantic' or maybe 'enjoyable' sound. My post-heading was tongue-in-cheek, and although wood enclosures *may* enhance reproduction of 'woody' instruments, I cannot think that they would do very much for reproducing a triangle or cymbals, for example, very accurately!
I am still a little puzzled why makers with a long-standing reputation for HQ 'real' musical sounds like Bosendorpher grand pianos, would wish to enter the loudspeaker arena, after so many years.
However, until I hear for myself the results of their efforts I will keep an open mind, as I cannot believe that they would wittingly risk ruining their repution through making and selling lousy-sounding speakers.
Seasonal greetings,
O.K., I now understand what happened here, and I was certainly not trying to cross swords with anyone, especially at this time of year.
As I attempted to make clear, *in the main* I agree entirely that the lowest cabinet 'reinforcement' of the sound is likely to be the best way of achieving accuracy, which as we both know is not the same as achieving a 'romantic' or maybe 'enjoyable' sound. My post-heading was tongue-in-cheek, and although wood enclosures *may* enhance reproduction of 'woody' instruments, I cannot think that they would do very much for reproducing a triangle or cymbals, for example, very accurately!
I am still a little puzzled why makers with a long-standing reputation for HQ 'real' musical sounds like Bosendorpher grand pianos, would wish to enter the loudspeaker arena, after so many years.
However, until I hear for myself the results of their efforts I will keep an open mind, as I cannot believe that they would wittingly risk ruining their repution through making and selling lousy-sounding speakers.
Seasonal greetings,
Thanks there SY. Talk about throwing the board back at the dart and hitting the bullseye, these papers are perfect.
Some helpful observations start on page 93 of the paper from 1992. Following the statement:
Along those lines, I am setting up an experiment to prop a jbl 125A on different places of the lower (thin spruce) frontboard removed from an upright piano. I want to get some evidence for possibly cloning the ProAc Response 2.5.
Thanks also to freddi for the Karlson coupler approach. What effects do those wings make in the sound? How big is the box overall and what's in it?
patch
Some helpful observations start on page 93 of the paper from 1992. Following the statement:
Along those lines, I am setting up an experiment to prop a jbl 125A on different places of the lower (thin spruce) frontboard removed from an upright piano. I want to get some evidence for possibly cloning the ProAc Response 2.5.
Thanks also to freddi for the Karlson coupler approach. What effects do those wings make in the sound? How big is the box overall and what's in it?
patch
One of the big issues in discussions such as these is the use of the term "natural". This can't really be defined because for live music, natural depends very much on seating distance from the instruments.
Natural for a musician sitting in the orchestra and natural for someone sitting at the back of the auditorium are very different sounds.
I read reviews of some high priced and well regarded speakers where the designer intentionally did not choose flat. A gradual roll off from bass to treble of a few dB sounds more natural to many people. It has the advantage of being more forgiving of poorly recorded music especially digital recordings.
My friend and I designed a quite expensive homemade three way speaker using very good drivers (Scan Speak and Audio Technology, all Skaaning based designs). We built in an adjustable "house curve" compensation. When we voiced them using only our ears for most "natural" sound, we ended up with a 4 dB downward shift starting at around 2 KHz and leveling out again around 10 KHz. We also liked about a 1 dB "bump" in the 80 to 300 Hz region. My friends listening room has one all glass wall and a hard wood ceiling.
If we disable these compensating networks the speaker sounds a bit edgy and lifeless. The audio equipment and cabling are all top grade audiophile and can hardly be blamed for the sound with the networks removed.
Your mileage may vary.
Natural for a musician sitting in the orchestra and natural for someone sitting at the back of the auditorium are very different sounds.
I read reviews of some high priced and well regarded speakers where the designer intentionally did not choose flat. A gradual roll off from bass to treble of a few dB sounds more natural to many people. It has the advantage of being more forgiving of poorly recorded music especially digital recordings.
My friend and I designed a quite expensive homemade three way speaker using very good drivers (Scan Speak and Audio Technology, all Skaaning based designs). We built in an adjustable "house curve" compensation. When we voiced them using only our ears for most "natural" sound, we ended up with a 4 dB downward shift starting at around 2 KHz and leveling out again around 10 KHz. We also liked about a 1 dB "bump" in the 80 to 300 Hz region. My friends listening room has one all glass wall and a hard wood ceiling.
If we disable these compensating networks the speaker sounds a bit edgy and lifeless. The audio equipment and cabling are all top grade audiophile and can hardly be blamed for the sound with the networks removed.
Your mileage may vary.
hermanv said:
I'm afraid I have to take issue with this also. If we talk natural vs. reproduced, natural is what gets to your ears from the original source, regardless of listening environment, position, etc.
Yes what the musician hears and what the audience member hears are different, but both are natural. If you place a microphone next to the musician what the musician hears is what, limited by the technology of course, will be recorded, and a microphone placed in the position of the audience member will record the sound as that person hears it. We are assuming there is some minimum perceptive difference between the two listening positions here.
When both recordings are played back on decent loudspeakers the listener will be able to tell which one was taken next to the musician and which one next to the audience member. In both cases we are hearing a reproduction of the natural sound.
And we don't need vibrating cabinets to do it!
All true, the problem is that while designing loudspeakers we know little of how a particular recording was made, little about how it's supposed to sound. Striving for natural when the source represents an unknown is an exercise in futility. We compensate by using many different recordings, that helps but doesn't solve the problem.sdclc126 said:I'm afraid I have to take issue with this also. If we talk natural vs. reproduced, natural is what gets to your ears from the original source, regardless of listening environment, position, etc.
Yes what the musician hears and what the audience member hears are different, but both are natural. If you place a microphone next to the musician what the musician hears is what, limited by the technology of course, will be recorded, and a microphone placed in the position of the audience member will record the sound as that person hears it. We are assuming there is some minimum perceptive difference between the two listening positions here.
When both recordings are played back on decent loudspeakers the listener will be able to tell which one was taken next to the musician and which one next to the audience member. In both cases we are hearing a reproduction of the natural sound.
And we don't need vibrating cabinets to do it!
I have access to some fine instrumentation, it got our design close, but most certainly not "right". We tuned our design until is sounded right, natural if you prefer.
My issue is the assumption that natural is obvious when you hear it, it is not. A small dB shift still can sound natural but the end result still sounds different, a change in flavor. Many flavors of natural exist and unless you have the opportunity to switch back and forth between a live event and the reproduction you can't really achieve this tuning. Worse. almost all modern recordings are multi-miked with separate microphone for each instrument often shoved down the instruments throat. In this case the only definition of natural is what the recording engineer decides is natural. Even large orchestral works use far more microphones than we have ears, again the end result is largely determined by the mixing console.
Even if you had the capability to switch between live and recorded, you could adjust either the recording equipment or the playback equipment to fine tune the end result and be unable to tell the difference . There's an innate assumption that flat equals natural, it can't because the processing and listening room add their own flavor to the end result. many have strived for maximum flatness, usually with a resulting design that most listeners felt lacking in musicality.
What you are talking about sounds reasonable, but the devil is in the details.
Actually I agree with much of what you're saying!
I agree that when designing loudspeakers you know little of any particular recording - an unknown - so, you don't try to design with the unknowns - there will always be an infinite number of those.
I agree that there are many different "flavors" of natural - in my post I described an example of exactly that.
But it sounds like you are describing an attempt at "averaging" all possible live recordings, so that a speaker sounds like that average of all the various flavors, and that average will sound most "natural" to the most people.
My philosophy is to dicard the unknowns and those devilish details - all the things you can't control - and boil speaker design down to one scientific principle - comparing the output signal to the input signal. The more these are alike the more accurately the speaker is reproducing the sound that entered the recording microphone(s) in the first place.
I agree that when designing loudspeakers you know little of any particular recording - an unknown - so, you don't try to design with the unknowns - there will always be an infinite number of those.
I agree that there are many different "flavors" of natural - in my post I described an example of exactly that.
But it sounds like you are describing an attempt at "averaging" all possible live recordings, so that a speaker sounds like that average of all the various flavors, and that average will sound most "natural" to the most people.
My philosophy is to dicard the unknowns and those devilish details - all the things you can't control - and boil speaker design down to one scientific principle - comparing the output signal to the input signal. The more these are alike the more accurately the speaker is reproducing the sound that entered the recording microphone(s) in the first place.
Some aspects of piano construction could be relevant to loudspeaker construction. To think of bracing a speaker cabinet not to prevent resonance but to enhance it. To allow for moveable parts. To allow for adjustable openings.
Going back to my upright, it is interesting to hear how the body of the instrument works with the sounds going on inside.
A piano's rim and cavity (maybe comparable to the walls and volume of a loudspeaker) appear designed as the basis of an instrument-within-an-instrument: a system for conveying and conserving resonant energy. Any piano builders or tuners should comment. There is a complex morphology to this. The panels, and some of the other wood (boards, blocks, strips) in there, though distal to the piano's sounding board, are connected to that source, apparently to vibrate and give richness to the sound.
The quality of the sound can be adjusted through the pedals. The sound and volume are adjusted by changing the angle of the lid. In an upright, changing the size of the gaps around the frontboards also will affect volume and sound quality.
I wonder if anyone has made a listenable loudspeaker that brings these features into play.
Going back to my upright, it is interesting to hear how the body of the instrument works with the sounds going on inside.
A piano's rim and cavity (maybe comparable to the walls and volume of a loudspeaker) appear designed as the basis of an instrument-within-an-instrument: a system for conveying and conserving resonant energy. Any piano builders or tuners should comment. There is a complex morphology to this. The panels, and some of the other wood (boards, blocks, strips) in there, though distal to the piano's sounding board, are connected to that source, apparently to vibrate and give richness to the sound.
The quality of the sound can be adjusted through the pedals. The sound and volume are adjusted by changing the angle of the lid. In an upright, changing the size of the gaps around the frontboards also will affect volume and sound quality.
I wonder if anyone has made a listenable loudspeaker that brings these features into play.
For want of consistent (coherent actually) results I stopped my short "scientistic" tests around the idea of a pleasingly resonant domestic loudspeaker enclosure or frame made of wood. No wood I used, even veneer, had adequate sensitivity and persistence. I kept getting a dull one-note effect.
For a loudspeaker to make a musical resonant field, apparently you need an enclosure or frame made of sensitive fast-responding material capable of a producing a range of middle and lower frequencies. Probably something like the stiff materials used in speaker cones and drum heads.
I'm also thinking that to enlarge any resonance-effects (without putting a stack of bass drums and tom-toms along the wall), a sensitive material could be arranged close to the core enclosure or frame. This might be a separate suspended shell. Or maybe an array of shingle-like panels, tuned to a frequency and attached to the core. All guesses.
patch
For a loudspeaker to make a musical resonant field, apparently you need an enclosure or frame made of sensitive fast-responding material capable of a producing a range of middle and lower frequencies. Probably something like the stiff materials used in speaker cones and drum heads.
I'm also thinking that to enlarge any resonance-effects (without putting a stack of bass drums and tom-toms along the wall), a sensitive material could be arranged close to the core enclosure or frame. This might be a separate suspended shell. Or maybe an array of shingle-like panels, tuned to a frequency and attached to the core. All guesses.
patch
Most of the discussions here supporting resonance in cabinets relate to sine wave effects, the problem with resonant cabinets is getting them to stop as fast as the music wants them to. This is very difficult, not impossible, but to do do it by ear alone would be unbelievably time consuming.
If you are listening to a musician playing a piano and he uses a foot pedal to damp the ringing how does your speaker now know it's time to stop resonating those notes.
If you are listening to a musician playing a piano and he uses a foot pedal to damp the ringing how does your speaker now know it's time to stop resonating those notes.
Those are good points.
My question, which I didn't answer, wasn't exactly to start and stop vibrations in near-time with the loudspeaker - I think I could do that by hooking up another driver on a delay and maybe aphexing it.
I guess it was whether or not someone might average-out and somhow sustain the drivers' vibrations in some sort of resonating action that would complement music. And do it outside the loudspeaker cavity with physical materials and not electronics. And not depend on the room to do it.
But it looks like getting to a worthwhile result fast does afterall require someone to think inside the box....the big one, the room.
My question, which I didn't answer, wasn't exactly to start and stop vibrations in near-time with the loudspeaker - I think I could do that by hooking up another driver on a delay and maybe aphexing it.
I guess it was whether or not someone might average-out and somhow sustain the drivers' vibrations in some sort of resonating action that would complement music. And do it outside the loudspeaker cavity with physical materials and not electronics. And not depend on the room to do it.
But it looks like getting to a worthwhile result fast does afterall require someone to think inside the box....the big one, the room.
"I guess it was whether or not someone might average-out and somhow sustain the drivers' vibrations in some sort of resonating action that would complement music."
I'm playing the devil's advocate here so if you don't want to hear my views let me know and I'll unsubscribe...
What sort of resonating action would compliment the music? It would still be something that was never in the music in the first place.
I'm playing the devil's advocate here so if you don't want to hear my views let me know and I'll unsubscribe...
What sort of resonating action would compliment the music? It would still be something that was never in the music in the first place.
Hi Patch;
I did cover the start/stop issue in previous posts, but I used Q to express the problem.
If a driver has a suckout, gap or dip in response, a resonant enclosure can indeed fix the problem. It can even over correct a little to add warmth or body to the sound. The property that controls how fast a resonant driver, network or cabinet stops and starts is called Q and is closely related to bandwidth.
With a modeling program it is possible to measure or model Q so it becomes possible to do exactly what you've been proposing. I was merely making it clear that this was not a trivial task, that without proper tools it becomes nearly impossible.
From your posts I thought you might be relatively new to speaker building and I suggest that to implement your idea you find an already proven design and copy it. Then modifications are built on a known starting point.
There is still considerable art to speaker design, the math has improved and will get you quite close, but that last little increment of excellence still seems to need human ears and some fiddling,
I did cover the start/stop issue in previous posts, but I used Q to express the problem.
If a driver has a suckout, gap or dip in response, a resonant enclosure can indeed fix the problem. It can even over correct a little to add warmth or body to the sound. The property that controls how fast a resonant driver, network or cabinet stops and starts is called Q and is closely related to bandwidth.
With a modeling program it is possible to measure or model Q so it becomes possible to do exactly what you've been proposing. I was merely making it clear that this was not a trivial task, that without proper tools it becomes nearly impossible.
From your posts I thought you might be relatively new to speaker building and I suggest that to implement your idea you find an already proven design and copy it. Then modifications are built on a known starting point.
There is still considerable art to speaker design, the math has improved and will get you quite close, but that last little increment of excellence still seems to need human ears and some fiddling,
I personally believe all views are welcome, the reproduction of music by electronic means is far from perfect. In other words we may not know all there is to know on the subject, your views are as likely to add information as anyone else's.sdclc126 said:
When my friend and I finished our three way design it measured flat to +/-2dB from 30Hz to 18KHz and -3dB to 22Hz and 20KHz. This was pretty good and reflected our use of quite expensive drivers and crossover components. We used two state of the art programs; MLSSA for measuring and LEAP for both enclosure and crossover design. We spent thousands and about 7 years (part time).
The resulting speaker was very detailed, revealing and coherent with excellent extension at the frequency extremes and sounded sort of mechanical, not particularly musical.
With extensive listening tests we "voiced" the speaker to sound good to us. This is a subjective rather than objective approach. In order to achieve what we felt was a musical speaker we added roll off at the high end, a gentle slope starting at about 2Khz and ending up about 5 db down at 20KHz. We also added about 1.5 dB of boost starting at about 90Hz and dropping off at about 500 Hz.
This is pretty consistent with what I've read, truly flat speakers tend towards a lifeless sound. MLSSA allows one to make a speaker where the acoustic output is a close match to the input waveform, as you have advocated. Having tried exactly that, we personally were dissatisfied with the end result. The voicing process "fixed" the problem. Our speaker sounds as good to me as the very best I've ever heard at any price.
"This is pretty consistent with what I've read, truly flat speakers tend towards a lifeless sound. MLSSA allows one to make a speaker where the acoustic output is a close match to the input waveform, as you have advocated. Having tried exactly that, we personally were dissatisfied with the end result. The voicing process "fixed" the problem. Our speaker sounds as good to me as the very best I've ever heard at any price."
Voicing a speaker to personal taste is fine - but that is a far cry from intentionally making a resonant speaker cabinet - it seems impossible to me, conceptually, to have the cabinet resonate when and how you want and to never do so when you don't.
Please post your loudspeaker design - I'm very intrigued!
Voicing a speaker to personal taste is fine - but that is a far cry from intentionally making a resonant speaker cabinet - it seems impossible to me, conceptually, to have the cabinet resonate when and how you want and to never do so when you don't.
Please post your loudspeaker design - I'm very intrigued!
sdclc126: Many details of our design (the other team member is Curmudgeon) have been posted on other threads, here are some of the threads:
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Opinions Needed on Final Driver Selection
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Accuton vs Audio Technology
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Need impressions of capacitors
diyAudio Forums > Top >Other Stuff >Electronics and Parts >High Quality 100 uF midrange cap.
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Best X-over parts
There are more bits and pieces here and there, we have never created a thread specifically to showcase our design. We have shared details with some DIY members. The design is expensive and complex, not recommended for newbies.
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Opinions Needed on Final Driver Selection
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Accuton vs Audio Technology
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Need impressions of capacitors
diyAudio Forums > Top >Other Stuff >Electronics and Parts >High Quality 100 uF midrange cap.
diyAudio Forums > Top >Loudspeakers >Loudspeakers >Best X-over parts
There are more bits and pieces here and there, we have never created a thread specifically to showcase our design. We have shared details with some DIY members. The design is expensive and complex, not recommended for newbies.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.