I would say the former, so at output level of 120dB distortion componenets are <0.08% or 62dB lower than the signal, so THD<58dB at 120dB output.
Does that make "dynamic range" 62dB?
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Speakers and IM distortion
- Thread starter Mtbf
- Start date
I would say the former, so at output level of 120dB distortion componenets are <0.08% or 62dB lower than the signal, so THD<58dB at 120dB output.
Does that make "dynamic range" 62dB?
- Joined
- Feb 23, 2016
- Messages
- 20,696
- Likes
- 37,431
I would say the former, so at output level of 120dB distortion componenets are <0.08% or 62dB lower than the signal, so THD<58dB at 120dB output.
Does that make "dynamic range" 62dB?
That isn't the conventional definition of dynamic range. Usually it is max clean input level minus self noise. Microphone SNR is by convention 94 dbSPL minus self noise. 94 dbSPL is one Pascal of pressure.
By the way, Earthworks simply give an SPL level for which there is no distortion. So how low is distortion to be called no distortion. Typically on their measurement mikes it is 140 dbSPL at least for the sound level.
So minus self noise but not minus (THD + self noise)?
Does it mean that their mike can accept 140dBSPL at it's membrane without measureable distortion?
Probably the best theoretical treatment of the microphone distortions I've seen so far: http://www.gedlee.com/downloads/AT/Chapter_11.pdf. The math isn't easy to understand from just this chapter - the fundamentals are introduced earlier in the book. Please don't despair - the gist can be discerned without understanding the whole mathematical derivation.
Qualitatively, it is easier to make a highly linear condenser microphone than a highly linear dynamic transducer. The microphone effectively measures a very small diaphragm displacement caused by sound pressure - a rather direct conversion in quasi-linear displacement range. The transducer has to generate sound pressure through acceleration within a comparatively wide displacement range, which is affected by multiple factors, including non-linear ones, in addition to the voltage applied to its coil.
On the other hand, the issue of self-generated noise is virtually non-existent in the dynamic transducers, whereas in the condenser microphones noise considerations tend to dominate practical designs.
As Dr. Geddes noted about a condenser microphone: "The first thing to notice is that it is not difficult to get an almost ideal response curve (except for the high end, which, as we will see, is easily controllable)". I guess this general result may have led to a too liberal approach to microphones design. Proper geometry of microphone's chamber and sufficient mechanical damping of its diaphragm are crucial for approaching the theoretically achievable linearity. It appears that some budget microphones are skimping on those.
- Joined
- Feb 23, 2016
- Messages
- 20,696
- Likes
- 37,431
Correct.
I don't know, it has to have some level of distortion, I just don't know what they are calling without distortion. It could be 1% or .1% or something else they consider inconsequential.
I like this however, distortion of air at 194 dbSPL.
That seems like avoiding the truth to me..
Arent't that kind of pressure a nuclear blast thing? Admit that you posted this only for Sergei!
I can imagine his reaction..
- Joined
- Feb 23, 2016
- Messages
- 20,696
- Likes
- 37,431
That seems like avoiding the truth to me..
Arent't that kind of pressure a nuclear blast thing? Admit that you posted this only for Sergei!
I can imagine his reaction..
View attachment 29499
On the DPA site, they have recordings of the Saturn 5 liftoff of the Apollo 13 mission. They say the sound of distortion you hear isn't the microphones, but the air because sound levels exceeded 194 db SPL.
You can listen here if you wish:
https://www.dpamicrophones.com/mic-university/the-basics-about-distortion-in-mics
Maybe a different perspective on distortion:
"Starting on the page labeled 100 in this Mechanix Illustrated guide (or PDF page 107), there is a long interview with various loudspeaker questions posed to multiple top engineers of the time, including Paul W. Klipsch. With too much text to reproduce here, the questions from the moderator and PWK's answers are shown below. Check out the link directly for the complete article."
https://www.americanradiohistory.co...als/Mechanix-Illustrated-Hi-Fi-Guide-1956.pdf
-----------------------
MR. HOEFLER: Much has been said about the number of elements in a loudspeaker system. One school holds that the theoretical ideal would be a single radiating element which operates with equal efficiency over the ten or more oc- taves audible range. At the other extreme are those who feel that a large number of specialized speakers, each operating over a restricted range, is the better approach. What is your opinion about the number of speaker elements and their configuration?
COL. KLIPSCH: It is a matter of fact, not opinion, that a single direct radiator of small size will give a more nearly flat
response over a wider range than some larger and more complicated systems. But likewise it is a fact that a small direct - radiator speaker, say six inches in diameter, would have to perform an excursion of several inches to produce a barely audible fundamental output at a few feet distance at some low frequency like 60 cycles, which, even if possible, would result in distortion many times the fundamental output. It is also a fact that when two or more frequencies are mixed in a single piston radiator, the lower frequency amplitude will modulate the upper one, which causes the oft -referred -to inter - modulation distortion. The result of recognition of these facts is that the best loudspeaker systems now employ at least
two loudspeaker elements. The design problems entail minimizing the frequency anomolies and interferences between a pair of speakers radiating the same frequency
near a crossover point. At present, only multiple speakers have been capable of duplicating the sound to a high degree of accuracy.
-----------------------
MR. HOEFLER: Other than assuring the proper impedance connection between the output transformer of an amplifier and
the input circuit of a speaker system, what other precautions may the home hi-fi listener observe to be certain of optimum
damping and minimum transient distortion?
COL. KLIPSCH: The subject of optimum damping is one which could fill a book. In my opinion the safest procedure
beyond proper impedance matching is to use an amplifier with an internal impedance between the limits of about ,r-s and 1
the nominal impedance, without any application of variable damping. This is expressed as an opinion rather than a universal fact, as there are exceptions.
-----------------------
MR. HOEFLER: The number of enclosures now on the market, each with its own claims of excellence in design and
performance, has many prospective hi-fi hobbyists thoroughly confused. Is there any one type which you feel is outstanding? Are there any basic principles which you can suggest to guide the prospective purchaser?
COL. KLIPSCH: Being a designer of horns, I feel justified in a strong prejudice for them in preference to enclosures. Experience with enclosures since 1938 indicates the closed back box as providing the best enclosure from the standpoint of lowest distortion. The enhanced bass of various ported boxes seems to be in the form of peaks which are in the wrong places. Note the distinction here between a "box" or "enclosure" and a "horn." The former is a means of isolating the front from the back of a cone radiator. A horn, on the other hand, provides a transformer action, like the gear box on a car (or, if you can't remember that far back, think of it functionally as a torque converter). With it, the diaphragm excursion and consequently the distortion may be reduced for a given output power.
-----------------------
MR. HOEFLER: The dynamic speaker has been in common use for so long that we sometimes forget that development in this area has included many other types such as balanced armature, inductor driver, iron diaphragm, electrostatic and crystal units. Now the electrostatic speaker is enjoying a resurgence and there is much discussion about the ionic
speaker. Do you feel that there is any speaker type, either now in production or in someone's ivory tower, which will
supplant the dynamic system?
COL. KLIPSCH: What form of driving mechanism for the loudspeaker diaphragm will emerge as best would require a crystal ball. Currently there are several new products advertised as "new in principle." They sound different, all right, including a strong difference from the original sound. One speaker of low excursion capability exhibits such high distortion at very low levels as to attribute the "brilliance" to
this cause. For now, these devices are for sound effects rather than sound reproduction. One electrostatic tweeter showed
over 400 per cent distortion at 5,000 c.p.s. at an output corresponding to that of a dynamic unit with an input of a small fraction of a watt.
-----------------------
MR. HOEFLER: There has been so much
dramatic improvement in amplifier design in recent years that the loudspeaker system is now frequently criticized as
being the weakest link in the hi-fi chain. Do you feel that such criticism is justified? Do you feel, as a practical matter, there
is much room for improvement in loudspeakers? Just what are the toughest design problems faced by the loudspeaker engineer?
COL. KLIPSCH: The dramatic improve- ment in amplifiers is largely in the minds of the advertising managers. There is, currently, a plethora of mediocre ampli- fiers on the market, and the good ones capable of 30% of their rated output at
10 c.p.s. can be counted on the fingers of a couple of cow's hoofs. Even so, the electromechanical transducers (speakers, pickups, microphones) remain subject to more forms of distortion than even some of the "just fair" amplifiers. The tough prob- lem in a speaker remains that of producing
an undistorted pressure in an imponder- able medium like air, with a piston made of a material with a specific gravity over
1,000 times that of air.
-----------------------
MR. HOEFLER: Are there any other
comments you would care to make about
loudspeakers: design, manufacturing, merchandising, application?
COL. KLIPSCH: There are some speakers, boxes, and speaker box combinations for which their makers advertise the
most fantastic claims, such as 16 c.p.s. and lower response from radiators of only one foot diameter. It could be granted that any small speaker could radiate any wavelength, but the fundamental radiated by a 10 -inch piston at a distance of ten feet in free space at 30 c.p.s. is demonstrably below the threshold of audibility. The "reports of independent laboratories" on such devices, as quoted, could stand a lot of
revision in the interest of practical accuracy. Many of these speakers are good within the size and price bracket, but the
advertising claims are so exaggerated as to be ludicrous to the experienced worker, but still capable of sucking in that part of the public which is too inexperienced to know the limitations of fundamental physics, or which actually likes to be fooled. The high fidelity art could stand CONSTRUCTION COMPLETED, the author and his some illumination of the quackery which pervades every new art. Even some of the reputable manufacturers have been beguiled or trapped into
exaggerations of both performance and claims. Some very fine speakers are deliberately designed with peaks (usually
about 50 and 8,000 c.p.s.) to exaggerate the "high fidelity" effects; some less than fine systems make up the loss in advertising specifications. The little single 12 -inch speaker with "response to 20 cycles" may be demonstrably responsive to this frequency in laboratory test, but the fact remains that at tolerable excursions (from distortion standpoint) the power response of a 12 -inch speaker is only 0.0005 watt at 30 c.p.s. This would give a 60 decibel intensity at about nine feet, which is some 4 db below the 64 -decibel threshold of hearing. Considering the 80 to 110 db levels encountered in recording, it is readily seen how far short of any useful response the small radiator must necessarily fall."
This article was from the late 40s, then, as now, there were tradeoffs between flat frequency response from a single driver and distortion.
Klipsch was at the forefront of testing, providing accurate specifications for products, as well as being involved establishing meaningful and reproducible testings standards for AES.
The laws of "fundamental physics" haven't changed. Paul Klipsch would have loved a forum like this.
Travis
"Starting on the page labeled 100 in this Mechanix Illustrated guide (or PDF page 107), there is a long interview with various loudspeaker questions posed to multiple top engineers of the time, including Paul W. Klipsch. With too much text to reproduce here, the questions from the moderator and PWK's answers are shown below. Check out the link directly for the complete article."
https://www.americanradiohistory.co...als/Mechanix-Illustrated-Hi-Fi-Guide-1956.pdf
-----------------------
MR. HOEFLER: Much has been said about the number of elements in a loudspeaker system. One school holds that the theoretical ideal would be a single radiating element which operates with equal efficiency over the ten or more oc- taves audible range. At the other extreme are those who feel that a large number of specialized speakers, each operating over a restricted range, is the better approach. What is your opinion about the number of speaker elements and their configuration?
COL. KLIPSCH: It is a matter of fact, not opinion, that a single direct radiator of small size will give a more nearly flat
response over a wider range than some larger and more complicated systems. But likewise it is a fact that a small direct - radiator speaker, say six inches in diameter, would have to perform an excursion of several inches to produce a barely audible fundamental output at a few feet distance at some low frequency like 60 cycles, which, even if possible, would result in distortion many times the fundamental output. It is also a fact that when two or more frequencies are mixed in a single piston radiator, the lower frequency amplitude will modulate the upper one, which causes the oft -referred -to inter - modulation distortion. The result of recognition of these facts is that the best loudspeaker systems now employ at least
two loudspeaker elements. The design problems entail minimizing the frequency anomolies and interferences between a pair of speakers radiating the same frequency
near a crossover point. At present, only multiple speakers have been capable of duplicating the sound to a high degree of accuracy.
-----------------------
MR. HOEFLER: Other than assuring the proper impedance connection between the output transformer of an amplifier and
the input circuit of a speaker system, what other precautions may the home hi-fi listener observe to be certain of optimum
damping and minimum transient distortion?
COL. KLIPSCH: The subject of optimum damping is one which could fill a book. In my opinion the safest procedure
beyond proper impedance matching is to use an amplifier with an internal impedance between the limits of about ,r-s and 1
the nominal impedance, without any application of variable damping. This is expressed as an opinion rather than a universal fact, as there are exceptions.
-----------------------
MR. HOEFLER: The number of enclosures now on the market, each with its own claims of excellence in design and
performance, has many prospective hi-fi hobbyists thoroughly confused. Is there any one type which you feel is outstanding? Are there any basic principles which you can suggest to guide the prospective purchaser?
COL. KLIPSCH: Being a designer of horns, I feel justified in a strong prejudice for them in preference to enclosures. Experience with enclosures since 1938 indicates the closed back box as providing the best enclosure from the standpoint of lowest distortion. The enhanced bass of various ported boxes seems to be in the form of peaks which are in the wrong places. Note the distinction here between a "box" or "enclosure" and a "horn." The former is a means of isolating the front from the back of a cone radiator. A horn, on the other hand, provides a transformer action, like the gear box on a car (or, if you can't remember that far back, think of it functionally as a torque converter). With it, the diaphragm excursion and consequently the distortion may be reduced for a given output power.
-----------------------
MR. HOEFLER: The dynamic speaker has been in common use for so long that we sometimes forget that development in this area has included many other types such as balanced armature, inductor driver, iron diaphragm, electrostatic and crystal units. Now the electrostatic speaker is enjoying a resurgence and there is much discussion about the ionic
speaker. Do you feel that there is any speaker type, either now in production or in someone's ivory tower, which will
supplant the dynamic system?
COL. KLIPSCH: What form of driving mechanism for the loudspeaker diaphragm will emerge as best would require a crystal ball. Currently there are several new products advertised as "new in principle." They sound different, all right, including a strong difference from the original sound. One speaker of low excursion capability exhibits such high distortion at very low levels as to attribute the "brilliance" to
this cause. For now, these devices are for sound effects rather than sound reproduction. One electrostatic tweeter showed
over 400 per cent distortion at 5,000 c.p.s. at an output corresponding to that of a dynamic unit with an input of a small fraction of a watt.
-----------------------
MR. HOEFLER: There has been so much
dramatic improvement in amplifier design in recent years that the loudspeaker system is now frequently criticized as
being the weakest link in the hi-fi chain. Do you feel that such criticism is justified? Do you feel, as a practical matter, there
is much room for improvement in loudspeakers? Just what are the toughest design problems faced by the loudspeaker engineer?
COL. KLIPSCH: The dramatic improve- ment in amplifiers is largely in the minds of the advertising managers. There is, currently, a plethora of mediocre ampli- fiers on the market, and the good ones capable of 30% of their rated output at
10 c.p.s. can be counted on the fingers of a couple of cow's hoofs. Even so, the electromechanical transducers (speakers, pickups, microphones) remain subject to more forms of distortion than even some of the "just fair" amplifiers. The tough prob- lem in a speaker remains that of producing
an undistorted pressure in an imponder- able medium like air, with a piston made of a material with a specific gravity over
1,000 times that of air.
-----------------------
MR. HOEFLER: Are there any other
comments you would care to make about
loudspeakers: design, manufacturing, merchandising, application?
COL. KLIPSCH: There are some speakers, boxes, and speaker box combinations for which their makers advertise the
most fantastic claims, such as 16 c.p.s. and lower response from radiators of only one foot diameter. It could be granted that any small speaker could radiate any wavelength, but the fundamental radiated by a 10 -inch piston at a distance of ten feet in free space at 30 c.p.s. is demonstrably below the threshold of audibility. The "reports of independent laboratories" on such devices, as quoted, could stand a lot of
revision in the interest of practical accuracy. Many of these speakers are good within the size and price bracket, but the
advertising claims are so exaggerated as to be ludicrous to the experienced worker, but still capable of sucking in that part of the public which is too inexperienced to know the limitations of fundamental physics, or which actually likes to be fooled. The high fidelity art could stand CONSTRUCTION COMPLETED, the author and his some illumination of the quackery which pervades every new art. Even some of the reputable manufacturers have been beguiled or trapped into
exaggerations of both performance and claims. Some very fine speakers are deliberately designed with peaks (usually
about 50 and 8,000 c.p.s.) to exaggerate the "high fidelity" effects; some less than fine systems make up the loss in advertising specifications. The little single 12 -inch speaker with "response to 20 cycles" may be demonstrably responsive to this frequency in laboratory test, but the fact remains that at tolerable excursions (from distortion standpoint) the power response of a 12 -inch speaker is only 0.0005 watt at 30 c.p.s. This would give a 60 decibel intensity at about nine feet, which is some 4 db below the 64 -decibel threshold of hearing. Considering the 80 to 110 db levels encountered in recording, it is readily seen how far short of any useful response the small radiator must necessarily fall."
This article was from the late 40s, then, as now, there were tradeoffs between flat frequency response from a single driver and distortion.
Klipsch was at the forefront of testing, providing accurate specifications for products, as well as being involved establishing meaningful and reproducible testings standards for AES.
The laws of "fundamental physics" haven't changed. Paul Klipsch would have loved a forum like this.
Travis
A lot of things Paul Klipsch said are still relevant. Some of them less so:
Invention of sub-woofer servo control by Philips researchers, and consequent polishing and popularization of the concept by Velodyne, led to the current state of art: servo subwoofers can maintain THD of <2% at huge excursions.
Advent of advanced materials used in tweeter domes reduced the attractiveness of horns for home and studio use. Klipsch statements are still relevant for PA systems.
watchnerd
Grand Contributor
Interesting....
I currently have a Dynaudio Contour 20 pair.
I was considering the TAD Compact Evolution One or Micro Evolution One as a possible upgrade (along with possibly the Dynaudio Confidence 20, but it doesn't come in my preferred color).
I might have to reconsider the TADs......
watchnerd
Grand Contributor
watchnerd
Grand Contributor
I'll admit, I find the psychoacoustics of distortion perplexing and the somewhat foggy attitude speaker reviewers (and maybe makers) have towards it bordering on schizophrenic.
Subjective reviewers will refer, at times, to be bad behaviors by speakers at higher SPLs, often commenting how they become dynamically compressed, or imaging gets worse, etc, which would seem to be manifestations of distortion. And yet they never call it such, and Stereophile, at least, doesn't seem to ever measure distortion.
On the flip side, speaker manufacturer marketing departments regularly talk about some feature that reduces distortion....whether a particular motor structure, a high sensitivity design, or a particular material. And yet few publish THD specs.
And I think we've all heard little mono smart speakers, or videos/music played on smart phones too loudly, and heard obvious distortion.
What's going on here?
It sometimes feels like speaker distortion is the famous Mark Twain quip about the weather -- everybody talks about it, but nobody does anything.
Is IMD created "in the air" when multiple tones are present?
Or is it strictly a product of transducers and electronics emitting multiple tones?
Or is it strictly a product of transducers and electronics emitting multiple tones?
KSTR
Major Contributor
You need insane sound pressure levels to make air distort.
Therefore, using two tones from two sound sources is a way to verify mic distortion, working back from the IMD components in the output, which contains of course all the harmonics from both sources but no IMD products
Therefore, using two tones from two sound sources is a way to verify mic distortion, working back from the IMD components in the output, which contains of course all the harmonics from both sources but no IMD products
KSTR
Major Contributor
See for example B&K's Microphone Handbook Part 1, Fig. 2.24. With a good microphone capsule and buffer the distortion seems to rise 20dB/decade for the 2nd harmonic and 40dB/decade for the 3rd, at low (stiffness controlled diaphragm region) frequencies. So, from that the order of magnitude of distortions to expect at a given level can be extraploated when you know the max level distortion.
For distortion measurement noise doesn't matter, so I'd pick a 1/4" capsule (has a high stiffness corner) and a good preamp. Any proper model from B&K, G.R.A.S, DPA, Gefell, .... should do.
I own a Gefell MK301 (1/4" capsule spec'd 100kHz and 168dBSPL) and stopped loosing sleep about mic distortion, the mic will never be a limiting factor at the levels it will ever see in reasonable speaker tests.
Similar threads
