On the subject "Where to next in the speaker design"? and as an open challenge to the expert DIYers:
It is rather obvious by now that there are common threads between even subjective impressions of what makes for a good audio reproduction by speakers.
Ok, "No Box" approach is already in (SL, JohnK, Elliot, TroelsG OBs, Pass, et al.); so is the uniform polar response and CD (now even a commercial JBL LSR6332 albeight with a limited +-30deg dispersion). But how about these comments:
SiegfriedL (Issues in speaker design): "I think a recent experience while attending a San Francisco Symphony performance gave me further insight into why we are able to tell a loudspeaker from a live instrument……Live instruments had a space between tones, like a black background, and even en mass always remained articulate. ………It seems to me that the ongoing-ness of sound is one of the major problems with speakers.". OR
TroelsG (TJL-3W): "But once these peaks are gone, there's a calm, "dark" and uncoloured sound left. "Dark"? The magnesium drivers have a quietness contrary to most other cone materials and for lack of better words, I call it "darkness". "
What made me notice these comments is that I also for a long time have had the same impression of B&W's clarity where the voice tones seemingly jump out of "darkness" and go right back into it, (and those are kevlar so it is not a Mg property per se).
So two questions:
Q1. for everybody: do you share the same subjective impressions? and
Q2. for the experts: can this be the next design challenge/criteria to consider in speaker design? (e.g. SL considered before quantifying "stored energy" in making driver selections but later stopped bringing it up and even changed away from a Mg driver in LX521). But he at least pointed out the new measurement techniques to try and capture some of this. maybe something like that could be applied to testing finished speakers as well?
It is rather obvious by now that there are common threads between even subjective impressions of what makes for a good audio reproduction by speakers.
Ok, "No Box" approach is already in (SL, JohnK, Elliot, TroelsG OBs, Pass, et al.); so is the uniform polar response and CD (now even a commercial JBL LSR6332 albeight with a limited +-30deg dispersion). But how about these comments:
SiegfriedL (Issues in speaker design): "I think a recent experience while attending a San Francisco Symphony performance gave me further insight into why we are able to tell a loudspeaker from a live instrument……Live instruments had a space between tones, like a black background, and even en mass always remained articulate. ………It seems to me that the ongoing-ness of sound is one of the major problems with speakers.". OR
TroelsG (TJL-3W): "But once these peaks are gone, there's a calm, "dark" and uncoloured sound left. "Dark"? The magnesium drivers have a quietness contrary to most other cone materials and for lack of better words, I call it "darkness". "
What made me notice these comments is that I also for a long time have had the same impression of B&W's clarity where the voice tones seemingly jump out of "darkness" and go right back into it, (and those are kevlar so it is not a Mg property per se).
So two questions:
Q1. for everybody: do you share the same subjective impressions? and
Q2. for the experts: can this be the next design challenge/criteria to consider in speaker design? (e.g. SL considered before quantifying "stored energy" in making driver selections but later stopped bringing it up and even changed away from a Mg driver in LX521). But he at least pointed out the new measurement techniques to try and capture some of this. maybe something like that could be applied to testing finished speakers as well?
Last edited:
Have you considered, that much like conspiracy theories, these things feed off of one another? Somebody hears something described in fashion X, likes it, uses it and adds their own spin. And so on down the line in your classic game of telephone.
I'd be a lot more inclined to take such subjective descriptions seriously, if they could at least demonstrate some correlation, with some sort of measurable characteristics.
Otherwise it's just exercising your vocabulary and imagination. Especially given that your perception of things and the way you think about them, has been well proven to strongly affect highly subjective assessments of things.
I'd be a lot more inclined to take such subjective descriptions seriously, if they could at least demonstrate some correlation, with some sort of measurable characteristics.
Otherwise it's just exercising your vocabulary and imagination. Especially given that your perception of things and the way you think about them, has been well proven to strongly affect highly subjective assessments of things.
Last edited:
if it was only that easy
that IS just the thing about it. there are no consented metrics or standard tests to quantify this. on his site Siegfrid L. was engineering his own excitation technique and a metric to look into stored energy, intermodulation, etc.
this is probably the next frontier. that is why I wanted to spearhead this discussion rather than to add to the already known topics of enclosure resonancies, speaker directivity etc. those have already been baked.
Anyone else? opinions? suggestions? too difficult a topic?
that IS just the thing about it. there are no consented metrics or standard tests to quantify this. on his site Siegfrid L. was engineering his own excitation technique and a metric to look into stored energy, intermodulation, etc.
this is probably the next frontier. that is why I wanted to spearhead this discussion rather than to add to the already known topics of enclosure resonancies, speaker directivity etc. those have already been baked.
Anyone else? opinions? suggestions? too difficult a topic?
I've heard good speakers described as "quiet" - which may seem odd at first. But I think it's similar to what you quoted above. In other words, a speaker that doesn't impose much of its own sound on the music. Low distortion, lack of ringing, box talk or other artifacts. A term for a subjective impression that may well have origins in objective qualities.
Yep. you got it Pano man 🙂. and I think we (DIY community) should challenge the front runners to start grinding the topic of coming up with objective metrics which can then (hopefully) be correlated to subjectives. excitation techniques for all sorts of testing with transients, modulated signals etc.
it is a tall order but so does everything new appear to be at first, and a few years later everybody starts taking it for granted, and eventually even a manufacturers' specs start showing the data, right?
it is a tall order but so does everything new appear to be at first, and a few years later everybody starts taking it for granted, and eventually even a manufacturers' specs start showing the data, right?
Last edited:
adding some objective discussion
Ok maybe we should spice the discussion up by introducing more objective data already provided by Siegfried L.
He clearly showed that just a few tones played simultaneously will introduce a broadband noisy background for all the other tones to come. The idea behind the "darkness" is to keep that background noise as low as possible i.e. "quiet", so any tone jumping out of that background and going back into it has a large dynamic range. In other words prevent the individual tone "spillover" into the other tone frequencies and their intermodulation. So he called for introducing multitone distortion measurements.
check the following links:
Frontiers
on the example of the tweeter excited with just two tones, a carrier tone and a modulating tone at 1/10th the carrier frequency where only the three tallest spectral lines are expected, one gets a whole spread of new ones running at a lower level. so there you are looking at the content inside the "darkness". well if that content is not kept low the background will not appear as "dark" and the speaker clarity will suffer.
there is no reason why these tests could not be done on both the raw drivers and the finished speakers.
Ok maybe we should spice the discussion up by introducing more objective data already provided by Siegfried L.
He clearly showed that just a few tones played simultaneously will introduce a broadband noisy background for all the other tones to come. The idea behind the "darkness" is to keep that background noise as low as possible i.e. "quiet", so any tone jumping out of that background and going back into it has a large dynamic range. In other words prevent the individual tone "spillover" into the other tone frequencies and their intermodulation. So he called for introducing multitone distortion measurements.
check the following links:
Frontiers
on the example of the tweeter excited with just two tones, a carrier tone and a modulating tone at 1/10th the carrier frequency where only the three tallest spectral lines are expected, one gets a whole spread of new ones running at a lower level. so there you are looking at the content inside the "darkness". well if that content is not kept low the background will not appear as "dark" and the speaker clarity will suffer.
there is no reason why these tests could not be done on both the raw drivers and the finished speakers.
Attachments
Last edited:
I tend to agree with George here. I dislike discussions that do not have a concrete foundation in quantifiable metrics.
That said, I have long believed that a compact impulse response is a highly desirable characteristic. It should decay rapidly with no significant tail. For the most part this implies a fairly flat and smooth frequency response, but not always. It is possible to have a flat frequency response and a longish tail if the system is not minimum phase. I have long argued that minimum phase aberrations in horns were a significant factor in their poor perception. I showed how the foam that I use is significant at "compacting" the impulse response by reducing the non minimum phase aspects of the impulse response.
Those are all quantifiable aspects of the sound and I have pursued them for a fairly long time. Is this "darkness"? I have no idea, but it is clear that a tail on the impulse response will fill in gaps between transients in a musical signal. I continue to believe that the impulse response tells us all that we need to know if we know what to look for.
By the way, very early reflections in the room will also fill in the gaps, but many argue that this is a "good thing". That is one argument that I don't buy. To me very early reflections cause a multitude of problems and this is very much a speaker design issue as well as a room one.
PS. Try as I may I cannot find any correlation between nonlinear aspects of a loudspeaker and sound quality. At this point I have to exclude it from the list of important criteria. Please note that it is not that nonlinearity is never an issue, only that it can easily be made insignificant and as such if it is an issue then the speaker is "broken" or at least badly designed for its application.
Last edited:
I heard the "darkness" thing myself, when upgrading amplifiers. Strangely, it took me a while to get used to it; my ears were used to the old "bright" sound. My guess is the "darkness" is simply the brain's reaction to a lack of intermod products, as koja says. I have noticed that along with the darkness came more low-level detail.
Glad to hear from you gents talking right on the topic.
Believe me I now it is a challenging one. SL also put it under "frontiers" heading for a reason. (p.s. I can also appreciate the complexity since I process spectral data daily for microphone measurements, only for non-audio aplications.).
I am just looking for more gurus (like Earl) to start considering multitone measurements and/or other means to start capturing the residual background through which the real music has to come through.
-----------
sideline: not to digress too much but to give Earl a credit for CD work: I listened to his system last winter and it was my first experience with a system where early reflections (& bass distribution) has been all taken care of. I was taken aback since the sound was like a curtain staying at a distance in the front and not coming closer and the speakers were not really locatable. It felt like a stage outdoors, no walls. I missed at first the semi-locatable feel of my speakers which i always associated with some imaging, but thought by the end: "I could really get used to this". I sort of recognised the same in the comments from people who moved from SL's Orion speakers to LX521.
But back closer to the topic:
Earl, since you mentiond foam as an aid to keeping the after-transient distortion low in horns: that foam you use looked to me much like what is used for windscreens on the microphones in the automotive measurements. Please answer only if not proprietary to your designs. Thx!
Believe me I now it is a challenging one. SL also put it under "frontiers" heading for a reason. (p.s. I can also appreciate the complexity since I process spectral data daily for microphone measurements, only for non-audio aplications.).
I am just looking for more gurus (like Earl) to start considering multitone measurements and/or other means to start capturing the residual background through which the real music has to come through.
-----------
sideline: not to digress too much but to give Earl a credit for CD work: I listened to his system last winter and it was my first experience with a system where early reflections (& bass distribution) has been all taken care of. I was taken aback since the sound was like a curtain staying at a distance in the front and not coming closer and the speakers were not really locatable. It felt like a stage outdoors, no walls. I missed at first the semi-locatable feel of my speakers which i always associated with some imaging, but thought by the end: "I could really get used to this". I sort of recognised the same in the comments from people who moved from SL's Orion speakers to LX521.
But back closer to the topic:
Earl, since you mentiond foam as an aid to keeping the after-transient distortion low in horns: that foam you use looked to me much like what is used for windscreens on the microphones in the automotive measurements. Please answer only if not proprietary to your designs. Thx!
Last edited:
I was using a "pro" amp for the horn section, and went to a First Watt J2. I suspect the pro amp had lots of crossover distortion at the 5mW output level where I was running it.
Ok George but that was not really constructive. a lot of people here spend a great deal of effort making amplifiers in attempt to improve sound with at least high efficiency speakers. now if you do not really appreciate the topic here, there are a lot more out there that may fit your interests better. Thx.
Hard coned drivers and low distortion that's what I did to get such clarity.
For a good description of the Accuton sound signature read Troels Jenzen-Accu: The sound
For a good description of the Accuton sound signature read Troels Jenzen-Accu: The sound
I believe you. Unfortunately I never heard Accutons. Since they are often referred to as detailed as almost clinical, it would be very interesting to see some measurements on those drivers as well (unfortunately SL also did not have them).
I however noticed that in his measurements Vifa P13WH which was well regarded in Olsons's Ariel speaker did very well. The ranking does not quite reflect it since SL, in his test, stuck to a constant volume displacement and this driver was in a smaller size than the rest, so the comparison for this driver was not apple-to-apple.
I also noticed a nice bouncy shape of the rolloff curve, which may be of importance in addition to considering the mere area under the curve.
I however noticed that in his measurements Vifa P13WH which was well regarded in Olsons's Ariel speaker did very well. The ranking does not quite reflect it since SL, in his test, stuck to a constant volume displacement and this driver was in a smaller size than the rest, so the comparison for this driver was not apple-to-apple.
I also noticed a nice bouncy shape of the rolloff curve, which may be of importance in addition to considering the mere area under the curve.
Attachments
Hi Koja
Thanks for the review. Some years ago at a show I realized that my speakers take some getting used to because they do sound "so different". On short listening experiences many put this down to a poor quality, but you have it exactly correct when you say "I could really get used to this". (Speakers at shows tend to jump out at you, a feature that in the long run will get very annoying.) Now that I have lived with this sound for so long, I can't listen to anything else and take it seriously.
As to multi-tone, it does not show those things that I do consider as important in the impulse response because it is a steady state test.
I would like to make a plea here about terminology. The term 'distortion" gets confused and we need to resolve that. When someone says, for example, that a horn sounds "distorted" that may be quite true. But that could be two things - linear distortion or nonlinear distortion. The two things are quite different and using the single term "distortion" does not adequately define the meaning. For the most part I exclude nonlinear distortion in a loudspeaker as a major effect (as does Toole and Olive). But I certainly believe that linear distortion is a major effect.
Most people will think of "distortion" as being nonlinear, as was implied when you mention multi-tone because this test will only show nonlinear distortion. All linear distortion is evident in the impulse response, but nonlinear distortion is not.
So I just hope that we can all be clear on the terms and meanings so that things don't go off on a tangent due to bad terminology.
I would consider the cross-hatched area to be indicative of a bad response, although ETC is too smoothed for my taste.
Well high efficiency speakers are the one situation where I will admit amps can make a difference. When the speaker's efficiency is near the S/N ratio of the amp, that can amplify noise to audibility. So a higher S/N ratio is necessary.
Tube amps do sound different as well, but that basically because they are lousy amps. Yes this is off topic though and I apologize for that.
However outside of those specific situations, the whole "amps sound different" belief is relevant to what you are talking about. Since what you are dealing with there, is people's belief's affecting their perceptions of things, when there is no actual difference.
In all probability, the problem with trying to define "darkness" as a distinct set of detectable sonic characteristics, is that it is different for different people. That it is largely a construct of their beliefs and how that shapes their perceptions. Like the claims about "airyness" or "width of soundstage" with amps, far more than it is any actual characteristics.
If you do want to try and do this with some basis in facts, I would suggest starting with your metaphor and try to figure out what it is the lack of that distinguishes it. Much as darkness is the absence of light.
THD and FR measurements can be found on the Accuton website
I wouldn't call them clinical but rather correct, they don't hide or masquerade anything.
This means that they need a good amplifier, source and recordings (no mp3s) otherwise you're listening to the noise/distortion of your amp/source or the mixedup recording.
But with good material they sound fabulous to me 🙂
I wouldn't call them clinical but rather correct, they don't hide or masquerade anything.
This means that they need a good amplifier, source and recordings (no mp3s) otherwise you're listening to the noise/distortion of your amp/source or the mixedup recording.
But with good material they sound fabulous to me 🙂
I'm not sure why people would use metaphorical terms like "darkness" when there are more literal terms that seem more appropriate, e.g. quiet, silence, and so on? Is the term 'darkness' supposed to be indicating or evoking something that terms like 'silence' or 'quiet' don't adequately capture?
The term 'darkness' does have Romantic overtones (with an emphasis on the primacy of feeling), so there could be something there, I suppose, but I can't think of any other good reason for using it in this case (unless I'm simply misunderstanding how the term is being used).
ON second thought, viewed from a more musical point of view (rather than that of mere 'sound') one could view the silence of some live music as akin to the negative space of a drawing, sculpture, or piece of architecture (which is often deliberately emphasized as a fundamental feature or ingredient of the work in question). I suppose if viewed in this sense there could be something to be said for using terms other than the more literal ones (as a way of indicating something that is essential to the music in question).
The term 'darkness' does have Romantic overtones (with an emphasis on the primacy of feeling), so there could be something there, I suppose, but I can't think of any other good reason for using it in this case (unless I'm simply misunderstanding how the term is being used).
ON second thought, viewed from a more musical point of view (rather than that of mere 'sound') one could view the silence of some live music as akin to the negative space of a drawing, sculpture, or piece of architecture (which is often deliberately emphasized as a fundamental feature or ingredient of the work in question). I suppose if viewed in this sense there could be something to be said for using terms other than the more literal ones (as a way of indicating something that is essential to the music in question).
Last edited:
Thanks Earl. Yes, the "non-linear" distortion was implied in that post. this is only because i used SL's examples to get the discussion off the ground.
I really meant not to limit the thread to the non-linear effects, rather aim it at whatever makes the speakers "dark" i.e. appear as they have a dark background from which the music comes and goes back to, so the fine details are clearly audible.
the problem is that there is not much in the way of measurements here (well maybe the OEMs benchmark competitors with some in-house metrics; I am almost certain B&W has something that works for them).
btw Earl, I credit you with the CD work, really!, made me consider it for a future design (when i can get a dedicated media room in my house). however, you do not use drivers in true mid size/range, so even for you this topic provides for some room to grow into.
I'd be interested in white papers where Toole considered some of this. please feel free to post. Thx.
I really meant not to limit the thread to the non-linear effects, rather aim it at whatever makes the speakers "dark" i.e. appear as they have a dark background from which the music comes and goes back to, so the fine details are clearly audible.
the problem is that there is not much in the way of measurements here (well maybe the OEMs benchmark competitors with some in-house metrics; I am almost certain B&W has something that works for them).
btw Earl, I credit you with the CD work, really!, made me consider it for a future design (when i can get a dedicated media room in my house). however, you do not use drivers in true mid size/range, so even for you this topic provides for some room to grow into.
I'd be interested in white papers where Toole considered some of this. please feel free to post. Thx.
- Status
- Not open for further replies.