Gentlemen, pointless confrontation does not advance anyone's cause. A word of advice, before posting a contentious reply, please re-read it and consider what it says about you to a third party who is not involved.
I have experienced that even very small adjustments can determine whether a jazz singer sounds false or in tune, and with that obtaining all the important shadings 😎
I am NOT convinced that its possible only by compensating other flaws in the chain...that would only be artificial 🙂
Pinkmouse, Exactly! Nicely put! If I want to see "Who's afraid of Virginia Wolf", I'll pop in the dvd!
patrikf said:
Well I can comment that…
You are absolutely wrong Patrik - when you believe that you can falsify a new claim by just pointing at what is "common knowledge", when the common knowledge does neither contradict nor address the issue. 🙂
And please reframe from using rhetorical insinuations of "mystical claims". It is your own your lack of understanding and knowledge that disables you to tell what is mystical and what is not apart. That does not give gou the right to insult people by insinuations.
Maybe these two factors are the reason for your behaviour?
You also do not seem to know (understand/respect) that science acknowledge two roads that leads to knowledge (or at least to usable information/models) - deductive and inductive method.
What you are talking about, and doing, is unfortunately a common behaviour in the academic word, but has absolutely nothing to do with science!
If you want to falsify a new claim, you will have to come up with some studies (old or new) that proves the new claims wrong! Or at least point at something indicating the new claim is "empty" (lacks ground*).
(It is not allowed in science to launch theories that have no ground to them (then it’s only a hypothesis), neither can you launch a theory with no falsification openings. But NO theories can be proven – only built and/or disproved, and it is not my job to disprove it. It’s the job of someone else than me, with a reasonable scientific understanding.)
Saying that you lack the knowledge that'll enable you to understand the relevance of a new claim, is not falsifying the claim. It is only an indication that you need to go to school Patrik. 🙂
patrikf said:
The parallel is irrelevant. I believe that you should probably try to get some scientific schooling Patrik, and after that, you should probably learn at least the basics in physics and psychoacoustics - before you indulge in attempts to teach others.
You seem to have great difficulties to separate inductive from deductive conclusions. When doing research, being able to use both of them is essential, but it is as essential to understand their differences. (And thus avoiding to ignore or miss the inductive facts, as you do.)
patrikf said:
You seem to be close to the only one not seeing the information; can it be that it is there after all? Perhaps you believe that there is information missing, because you lack some pieces of the puzzle knowledge vice? Regarding the "common knowledge" you are talking about – maybe it would be a good idea for you to make it yours?
If the problem is that you do not want to study to get even a basic knowledge about the basics of psychoacoustics (IACC, HTRTF, HAAS/TI-T...), but instead want to have everything served at your order, then you will have to ask someone else to serve it to you!
I'm not here to educate you in the basics. There are schools for that.
I'll gladly tell anyone who's interested in my views, of them. And I always try to explain so that it reasonable easy to understand also by people without the basic training in science in general and in physics and psychoacoustics.
But since it is virtually impossible to explain anything to someone who has a habit of only criticize by introducing speculations, who asks questions he thinks he already know the answers to though he know only very little and understands even less – please Patrik: Bother someone else!
Vh, iö
*To my knowledge the cryo-claims you mention lacks all ground - they are not supported by neither common knowledge, nor by physics, new knowledge, or by outcomes from scientific experiments (double blind listening with a single variable).
They are (to my knowledge) only hypotheses, and personally, I do not see what could give such claims the ground to be a theory.
This is the basic problem for the person trying to build a theory on “sonic improvements from cryo-treatment”:
Before one could prove that a cupper conductor can be improved, there must be proven that there are some audible flaws to improve on!
The stereo system however (to the contrary) has a lot of inherent colorations. Nothing (well there may something…) is easier to demonstrate!
Use tree speakers. Put one 3 meters to your right, one 3 meters in front of you, and the third 3 meters to your right. Speaker 1 and 3 are now separated by +/- 90 degrees = 180 degrees.
Now play a mono signal on speakers 1 and 3, and compare with speaker 1. Do they sound the same? Compare timbral and geometric properties of the phantom image produced by 1 & 3, to the sound from the mono-speaker number 2. Can you even tell the direction to the phantom image?
Reduce the spacing between speaker 1 and 3 to let’s say +/- 60 degrees. Can you still tell the difference between the phantom image from 1+3 compared with 2? If you are alive, awake and of hearing, you sure can!
Reduce the spacing between 1 and 3 to let’s say +/- 30 degrees. Can you still tell the difference between the phantom image from 1+3 compared with 2? Yes, you still can, but it is now actually becoming more problematic to create a perfect phantom image at 1/3 from the side (move speaker 2 and compare).
You will find that a phantom image does in fact newer sound identical to the mono source. It does not matter if you use amplitude or time panning. There is still a flaw - the stereo system coloration! 🙂
In a good room, and with a sensible opening angle (+/- 22.5 – 23 degrees) the coloration is manageable. The reasons to the coloration can be investigated, and compensated, and that’s what I’ve done. (Well, it is one of many, many things I’ve put into the design of the speakers.)
There are no mysteries there, and anyone comparing it with cryo-treatments (of cupper, to improve sound quality in cables) has probably misunderstood science. Claims that are based on inductive studies and not only deductive reasoning (though the theories can and will be based on a combination), can not be contradicted by pointing at something that is a parallel only if inductive facts are ignored.
IngOehman said:
Did you also had the opportunity to study "colourations" of the box (type of wood, MDF, other materials...) the speakers are made of (...and the visible colours like the speakers paint, as another example, having any influence in perceived sound)?
As for some components, tests are very difficult to make but not impossible to measure or to be put to test with instruments or with people. Like for the best sounding cap and it's best subjective (or not, if measured) quality.
Regards
aah... finally a question !! Actually one I was pondering myself. Is the material the speakers are made of REALLY important (of course it is!?) and is the method it´s put togheter (screwed, glued) equally important?
Some says MDF is the best, some says it´s not very good at all -
the only reason to use it is because its easy to work with!
I built a pair of loudspeakers myself rencetly, and almost by accident I built them with different thickness of the panels - 18/20/22mm.
I also glued them together whit a "rubberglue", originally intended for glueing windows on cars. So you could say that each panel is sligthly isolated from each other, in theory at least.
Q - Is there any advantages to be had from these two "things"?
Some says MDF is the best, some says it´s not very good at all -
the only reason to use it is because its easy to work with!
I built a pair of loudspeakers myself rencetly, and almost by accident I built them with different thickness of the panels - 18/20/22mm.
I also glued them together whit a "rubberglue", originally intended for glueing windows on cars. So you could say that each panel is sligthly isolated from each other, in theory at least.
Q - Is there any advantages to be had from these two "things"?
fritiof said:
What you have there is constrained layer dampening.
It can be effective at cutting down on material resonances. The idea is to have multiple materials laminated together, each with a different resonance signature and between these is a coupler. No one resonance signature dominates and more often than not a kind of homogenised resonance is created with no particular high Q peaked nature that one typical material alone would display.
MDF generally has problems with resonances at 100-700hz. Its cheap and easy to machine is its saving grace. Used in conjunction with lamination, heavy bracing and damping it makes quite dead cabinets.
Inductor said:
Very interesting questions!
Yes, I've tried to map the audibility levels of unwanted vibrations in the cabinet, but it's a complex thing to map.
The thresholds depends not only of level of the vibration (compared to the membranes), but also of surface areas, which frequencies the problems are at, the Q of the resonant behaviour at hand, and not the least where in the box there is a vibration, the orientation of the box in the room, how the box is decoupled from the adjacent surfaces and the acoustics of the room around it!
Typically, “box sounds” seem to be a reasonable small problem when no surface exhibits a Q larger than 10, nor a vibration * area product grander than 1/100th (- 40 dB) of that of the membranes.
A lot of loudspeakers performs reasonably well in this aspect (not necessarily good enough, but enough to make something else a bigger problem). So I’d say that rather than addressing those issues more, looking at the acoustic resonances inside the box (sneaking out though the membranes later) and to an even higher degree looking at behaviours of the membranes, would be a vice investment of time for many manufacturers.
In my opinion, it is always good to look at the combination of all factors that determines the loudspeaker performance. It is easy to forget that thin walls are beneficial also. It all depends on perspective… But even if I ignore obvious factors like “how easy it is to carry the speaker around”, I can still gain qualities by doing the walls thinner (to a degree). With a given sized speaker, thinner walls means larger inner volume, which means the opportunity to go down in membrane weight, and a little bit less (but still down) in motor strength. Result: Higher sensitivity (and efficiency), larger frequency range, better linear stroke, less thermal compression, lower 2:nd harmonic due to geometric distortion in the air… All of that (and more) from having thinner walls.
I’ve said it before: I hate compromises!
There are the two main schemes I use, trying to avoid them:
1. Always map the thresholds of audibility as meticulous as possible, to avoid even going into the land of compromising!
2. Always search for the alternative synergic solution that makes it possible to work around the compromises!
-------------------
When it comes to appearance and how it influences the perceived sound from the loudspeaker*, it is no longer the psychoacoustic discipline we are in, but it is still an interesting question. And one I’ve been studying.
The conclusions are these:
1. (Short term) Most people can not ignore the appearance of the speaker as a factor for how they perceive the sound from it. Not totally. Actually, most people don’t even have to see the object to be coloured by its appearance. It’s enough to know how it looks like. So when testing prototype loudspeakers, I try to have the lights out, and if possible not indicate what equipment is running, so that people listening are “blank” inside (free from suggestions, also unintended suggestions).
2. (Long term) Most people does however not look at the speakers (or even have their eyes opened) when they are using their speakers at home, after they has bought then, only when they are “testing them” before they purchase them. So the importance of the appearance is probably going down with time. Not probably. It does.
3. People working professionally with their speakers - having them as a tool, are VERY disturbed by flashy appearances. Monitors are used in rooms that always have their lights on, by people that have their eyes open most of the time (they have to be able to manoeuvre the mixing consol). Monitors should always be matte black or dark grey, or something that melts into the environment.
Best regards, Ingvar
- - - - -
*Did you ask about that, or did you only ask about damping properties of the lacquer? 🙂
If so, it is typically of very little importance. But sometimes its action as a sealer is of some importance. MDF is not an air tight material.
Hello all, Cool, we're back to talking about speakers! I'm of the opinion that the deader the speaker cabinet the better. I want to hear what the drivers are doing, not the cabinet. In addition, I feel the less energy transmitted from the back of the cabinet through the drivers, the better. Having said that, I do like some Harbeth and Spendor speakers which go the thin panel, bituminous felt route. I don't know if they are strictly speaking,....."accurate", but they do sound very "musical",and sometimes that's good enough!
ShinOBIWAN: Tanx!
tc-60guy: Yes, a dead cabinet is a good thing, but when do you get to the limit when spending more time, money and effort in making it even deader, will just steal away that very same resources (time, money and energy) from more important engineering needs - without even being beneficial in it self seen from an audible standpoint?
To me there are three different levels that can be chosen.
1. Good enough (for the ambition) and in balance with the rest of the parameters of the design (i.e. economically optimized engineering).
2. Targeting the thresholds of hearing, at a (including time and energy) minimized cost (i.e. performance optimized engineering).
3. Taking it over the limit, to ensure that there are tolerances to better cope with unknown errors that will add, and letting it cost wht it takes (i.e. engineering with no concern to the cost).
I believe that number three is a level that both loudspeaker engineers and amateurs often choose, but hopefully for slightly different reasons.
Both professionals and the amateurs for sure do it because of the fun in doing “to much”! 😉
But often I believe that the overdoing is restricted to the parameters that are easy to understand, the visible ones. One must not ignore the rest. I believe this is a good idea:
Start by asking the questions on what is important. Then answer them! (That goes for both professionals and amateurs as well.)
Finally – building loudspeakers can start. 🙂
As I’ve said earlier, I’m an amateur myself.
Best regards, Ingvar
tc-60guy: Yes, a dead cabinet is a good thing, but when do you get to the limit when spending more time, money and effort in making it even deader, will just steal away that very same resources (time, money and energy) from more important engineering needs - without even being beneficial in it self seen from an audible standpoint?
To me there are three different levels that can be chosen.
1. Good enough (for the ambition) and in balance with the rest of the parameters of the design (i.e. economically optimized engineering).
2. Targeting the thresholds of hearing, at a (including time and energy) minimized cost (i.e. performance optimized engineering).
3. Taking it over the limit, to ensure that there are tolerances to better cope with unknown errors that will add, and letting it cost wht it takes (i.e. engineering with no concern to the cost).
I believe that number three is a level that both loudspeaker engineers and amateurs often choose, but hopefully for slightly different reasons.
Both professionals and the amateurs for sure do it because of the fun in doing “to much”! 😉
But often I believe that the overdoing is restricted to the parameters that are easy to understand, the visible ones. One must not ignore the rest. I believe this is a good idea:
Start by asking the questions on what is important. Then answer them! (That goes for both professionals and amateurs as well.)
Finally – building loudspeakers can start. 🙂
As I’ve said earlier, I’m an amateur myself.
Best regards, Ingvar
Ingvar what is your stance on correctly measuring a loudspeaker?
I've noticed many discrepancies that lead to inaccurate loudspeakers despite measurements saying otherwise.
For example designing a loudspeaker with reference to nearfield measurements (mic at 40cm or closer to baffle). Often we base our entire design on such measurements when in reality we never listen at 40cm so its easy where the conflict might be. We only do these nearfield measures because we don't all have anechoic chambers, I can understand that perfectly and I can also hear people say use gated measurements but they're frought with problems too and its really just a band aid to cut out some of the garbage.
Let me tell a quick story about my own observations.
I was doing a nearfield measurement at 40cm from the driver mounted on a finished baffle(completed loudspeaker). I noted two dips between 1 and 2Khz that must be related to diffraction since this behaviour isn't native to the driver. I then moved the mic back to 2m and redid the measurement and it disappeared. If I'd gone ahead and corrected for such dips in the crossover, I'd have unwilling induced errors at distances likely to be used for actual listening.
Here is the chart showing nearfield(green) and far field(black):
I always have a constant struggle whenever I measure a loudspeaker to find good testing methodology that accurately reflects what we hear at the listening position.
What would you suggest based on your experience?
I've noticed many discrepancies that lead to inaccurate loudspeakers despite measurements saying otherwise.
For example designing a loudspeaker with reference to nearfield measurements (mic at 40cm or closer to baffle). Often we base our entire design on such measurements when in reality we never listen at 40cm so its easy where the conflict might be. We only do these nearfield measures because we don't all have anechoic chambers, I can understand that perfectly and I can also hear people say use gated measurements but they're frought with problems too and its really just a band aid to cut out some of the garbage.
Let me tell a quick story about my own observations.
I was doing a nearfield measurement at 40cm from the driver mounted on a finished baffle(completed loudspeaker). I noted two dips between 1 and 2Khz that must be related to diffraction since this behaviour isn't native to the driver. I then moved the mic back to 2m and redid the measurement and it disappeared. If I'd gone ahead and corrected for such dips in the crossover, I'd have unwilling induced errors at distances likely to be used for actual listening.
Here is the chart showing nearfield(green) and far field(black):
I always have a constant struggle whenever I measure a loudspeaker to find good testing methodology that accurately reflects what we hear at the listening position.
What would you suggest based on your experience?
A direct comparison between a phantom image and a "real" source could be misleading.IngOehman said:
Differences doesn´t necessary translate to a coloration in practice.
The phantom source normally exist in the environment of the auditory impression created by two sound sources and should perhaps be evaluated in that context...
No Patrik.
I think you must revaluate your ideas of the dignities. The flaws are nuances that can transform the impression from fantastically alive and real, to a little bit less real and slightly artificial - to “loudspeaker sound”. But they are hardly large enough to render any attempts to “compensate by ear” during mixing.
Or have you ever heard of a recording engineer or producer who’s consciously eq:ing differently depending on the turn of the pan pot? 😉
Anyway, if someone would try to do it, it’s a lost game anyway, since it can’t be performed within the transfer chain (where the music signal lives locked in a one dimensional world). The compensation can only be added in the decoder itself, where the 3 dimensional world is accessible.
ShinOBIWAN:
Though I always optimize loudspeakers to work in real rooms, I believe that having anechoic measurement conditions to you disposal is a necessity.
It is a tool that is necessary to have, simply because even if you can see how all the different components of the radiated (and then reflected) sounds adds together in the room - without the anechoic discrimination, it¡¦s just impossible to take the components apart! And you need to do that, because the ears, as oppose to the measurement microphone, can distinguish between different directions (to a degree, heavily depending on arrival times).
You are absolutely right that the idea that ¡§40 cm is an ok distance to capture the true response of a loudspeaker¡¨ (even a small one) is simply not true, nor are gated measurement techniques free form their own set of problems.
So, basically, I agree with you!
But ¡V I would not be convinced from your measurement (from which I can not decipher amount of room, nor if there are some smoothing at hand) that the 2 meter curve is the more accurate one.
If there are some true diffraction induced interferences there, giving the dip visible at 40 cm¡¦s, it is probably an effect that is typical for that radiation direction only. Most diffraction (or rather rim induced reflections) causing interferences, are not visible in the energy curve, since they are different in different radiation directions.
Personally, I prefer to always view a loudspeakers frequency response with 20 ¡V 40 000 Hz (or 10 ¡V 40 000 Hz) range horizontally, and a 50 dB range vertically.
The former because the ¡§coarse balance¡¨ (each frequency compared to all the energy in ALL the other octaves) is always more important than the detail balance (level at 1 kHz compared to level in the range 800 - 1,2 kHz, but with the rest of the range out of site).
The latter because it is a good psychological ¡§filter¡¨ against obsessing with details ƒº, and also because it allows you to see the whole course of the curve, of relevance in a system taken apart driver by driver, crossover filter included.
Best regards, Ingvar
I think you must revaluate your ideas of the dignities. The flaws are nuances that can transform the impression from fantastically alive and real, to a little bit less real and slightly artificial - to “loudspeaker sound”. But they are hardly large enough to render any attempts to “compensate by ear” during mixing.
Or have you ever heard of a recording engineer or producer who’s consciously eq:ing differently depending on the turn of the pan pot? 😉
Anyway, if someone would try to do it, it’s a lost game anyway, since it can’t be performed within the transfer chain (where the music signal lives locked in a one dimensional world). The compensation can only be added in the decoder itself, where the 3 dimensional world is accessible.
ShinOBIWAN:
Though I always optimize loudspeakers to work in real rooms, I believe that having anechoic measurement conditions to you disposal is a necessity.
It is a tool that is necessary to have, simply because even if you can see how all the different components of the radiated (and then reflected) sounds adds together in the room - without the anechoic discrimination, it¡¦s just impossible to take the components apart! And you need to do that, because the ears, as oppose to the measurement microphone, can distinguish between different directions (to a degree, heavily depending on arrival times).
You are absolutely right that the idea that ¡§40 cm is an ok distance to capture the true response of a loudspeaker¡¨ (even a small one) is simply not true, nor are gated measurement techniques free form their own set of problems.
So, basically, I agree with you!
But ¡V I would not be convinced from your measurement (from which I can not decipher amount of room, nor if there are some smoothing at hand) that the 2 meter curve is the more accurate one.
If there are some true diffraction induced interferences there, giving the dip visible at 40 cm¡¦s, it is probably an effect that is typical for that radiation direction only. Most diffraction (or rather rim induced reflections) causing interferences, are not visible in the energy curve, since they are different in different radiation directions.
Personally, I prefer to always view a loudspeakers frequency response with 20 ¡V 40 000 Hz (or 10 ¡V 40 000 Hz) range horizontally, and a 50 dB range vertically.
The former because the ¡§coarse balance¡¨ (each frequency compared to all the energy in ALL the other octaves) is always more important than the detail balance (level at 1 kHz compared to level in the range 800 - 1,2 kHz, but with the rest of the range out of site).
The latter because it is a good psychological ¡§filter¡¨ against obsessing with details ƒº, and also because it allows you to see the whole course of the curve, of relevance in a system taken apart driver by driver, crossover filter included.
Best regards, Ingvar
IngOhman, I didn´t mean to argue against your subjective impressions.
I meant that there can be different views of what a phantom image etc represents🙂.
I meant that there can be different views of what a phantom image etc represents🙂.
IngOehman said:
Just a question or two.
1. Can a person with total hearing loss in one ear distinguish the direction of sounds?
2. Could a pair of microphones with enough processing power and the right information programmed, distinguish direction of sound?
Andy Graddon said:
Well there are degrees in hell too, but: No, not well. Close to not at all.
A person who is stone deaf in one ear, can not distinguish directions to any large degree. Not even the directions that a person with normal hearing can distinguish vertically, by using the "self experience" of the HRTF-effects (i.e. interpreting these specific timbral changes as directional information), can a one eared person work out.
The reason for the need of two ears is that working out the vertical direction is an unsolvable equation without having the horizontal direction. 🙁
It is simply a matrix that will not make sense without the two ears.
Andy Graddon said:
Yes, but you need 4 of them. And even if you can pick them apart (as you would want for listening) they are still not separable the way you need them to be for anayses of a loudspeaker. In that case you want to see each radiation direction alone. You need an infinite separation of directions.
Best regards, Ingvar
Ingvar, I seem to remember reading years ago that people who are deaf in one ear can localize sound by moving their heads back and forth and using that to triangulate.
Yes, you are right. They can use the intensity-modulation (from the head shadowing) to resolve the direction. But it is a very imprecise method. And it is more about turning the head (around differens axis) than moving it back and forth.
Actually, we all use that method when we have difficulties to make out directions to sound sources.
Best regards, Ingvar
Actually, we all use that method when we have difficulties to make out directions to sound sources.
Best regards, Ingvar
- Status
- Not open for further replies.