Considering you are a dealer for them, I am not surprised by your comment. If they are in fact the best measuring speaker, that in no way makes them the best sounding speaker for a given room. It's all about the sound, not the measurements.
Harshness – What causes it? What relieves it?
- Thread starter caesar
- Start date
Considering you are a dealer for them, I am not surprised by your comment. If they are in fact the best measuring speaker, that in no way makes them the best sounding speaker for a given room. It's all about the sound, not the measurements.
Maybe "shown" is too strong a word. Suggested maybe: http://www.stereophile.com/rmaf2010/nordost_and_vertex_measurements/index.html
It would be a fun exercise to actually use that software to analyze system changes...
That is very true. However, I have heard both rooms and speakers that measure "flat" and the sound would make the hair on my two cats stand up
Christian,
We are still waiting to be shown the measurements that support such absolute claim - as far as I have seen we have several speakers that show better measurements than the Grimm LS1's.
I will take your word for it micro. Hunting down measurements for various speakers for comparison reasons is about as fun to me as going to the dentist.
The LS1 step response looks textbook; Almost like a quad. The measurements I've seen show very limited off axis. I think Stereophile goes all he way out to 90 degrees. IIRC, bruno and company only measure the LS-1 to 30. Too me, that's not enough information since I believe Power Response is really important.
Page 8 in the following link contains the measurements I think Keith used to support his claim:
http://www.grimmaudio.com/site/assets/files/2737/audio_mai_2013_grimm_ls1_testbericht_english.pdf
The distortion graph is incomplete. It needs to be resized to read it and off axis only to 30 degrees.
Last edited:
Its a interesting thread , although i have problems reading it , if i want to go to a page in the middle of the thread i go automatically back to page no 1 ???
Harshness has certainly been one of the key driving forces of my own speakerdesign , i am with jack on this , many times i thought it was the recording , but nowadays there is virtually none anymore i am talking digital , on analogue it wasnt so much of a problem in the first place , another strange thing i discovered is that on a real basspotent , non resonant system, digitalsound becomes much better / more acceptable .
Having a reasonable FR on a speaker is a must on a high $$ system, its a basic requirement,preferably with a SOFTdome tweeter with SOFT written with thick letters, best thing is they are not even expensive , all the more left over money that can be put in the housing
Harshness has certainly been one of the key driving forces of my own speakerdesign , i am with jack on this , many times i thought it was the recording , but nowadays there is virtually none anymore i am talking digital , on analogue it wasnt so much of a problem in the first place , another strange thing i discovered is that on a real basspotent , non resonant system, digitalsound becomes much better / more acceptable .
Having a reasonable FR on a speaker is a must on a high $$ system, its a basic requirement,preferably with a SOFTdome tweeter
with SOFT written with thick letters, best thing is they are not even expensive , all the more left over money that can be put in the housing
Last edited:
Agreed.
I am somewhat surprised that no one picked-up on Gary's post about PA in many concert venues. Yes, the MET uses these sometimes (they don't always mention it) and it is hard to know when they (and other venues) do. it is however an annoying concept for the audiophiles .. I must say that I have enjoyed more than a few Classical Music concerts not knowing there were PA involved. upon knowing i started to construct "things" in my mind. Which deterred to the enjoyment of the music. The lesson? We audiophiles have to learn not to fret so much . Just enjoy the damn Music!!!
Frantz,
It is widely known that when we have visual and ambiance clues the audiophile shield is down. It needs a good power conditioner to be energized again...
BTW, IMHO, this type argument (audiophiles constructing "things" in their minds) does not help finding a cure for audiophile anxiety. Only proper understanding of the "things" can cure it. Unfortunately our hobby is not a simple one.
In the 11th post of this thread, Diapason wonders why truly flat sounds bright. I believe the reason may have to do with high frequency absorption by air. Out where the audience sits, the highs have been attenuated somewhat by passing through a lot of air. Up where the microphones (usually) are, those highs are still in full force. So when what the microphones picked up is reproduced through a "flat" system, it sounds bright to most of us. It is less likely to sound bright to a musician who is used to listening at the source.
Okay, what causes harshness? Improperly used dome tweeters. But explaining how and why will take a few paragraphs. And this won't be an all-encompassing answer of course, but hopefully it will describe one possible source of harshness.
First, let's take a look a set of equal-loudness curves:
Where the curves dip lowest is where the ear is most sensitive. While the exact shape of these curves changes with level, we can clearly see that the ear is especially sensitive in the 2-5 kHz region, with sensitivity peaking at about 4 kHz. Just so happens this is the region where excess energy is often described as "harshness". So, how does this excess energy sneak into our systems?
I said earlier that improperly used dome tweeters were the culprit. In my opinion that's often where this excess energy in the 2-5 kHz region comes from, even if they measure perfectly flat through that region. How is that possible? Because that "perfectly flat" curve is an on-axis curve, and the excess energy is happening off-axis. Let me describe what's happening using an example:
Suppose we have a 6.5" two-way with our ubiquitous 1" dome tweeter, crossover frequency 2.5 kHz. At the crossover frequency, the woofer is just starting to beam - that is, its radiation pattern has narrowed to about 90 degrees (-6 dB limits). But the 1" dome tweeter wants to be nearly omnidirectional, because its diaphragm is so small in relation to a wavelength. Well the baffle keeps that from happening, acting as a 180 degree "horn" if you will, but the tweeter still has strong off-axis energy to the sides. And that strong off-axis energy holds up quite well from the crossover point up to ballpark 5 kHz. So we have excess off-axis energy in that 2-5 kHz region, which is the worst place for us to have excess energy.
Now maybe you're thinking, just treat the room and absorb it. The problem is, acoustically absorptive material is much more effective at absorbing shorter wavelengths than longer ones. So we are likely to end up with a room that is dead because we've killed all the high frequency reverberant energy, but haven't done as good a job of absorbing the energy in that 2-5 kHz lower treble region.
One possible solution would be to listen nearfield, such that the direct sound is much louder than the reverberant sound. Nearfield listening has its plusses and minuses, and is not my first choice, but it is a useful work-around for excess off-axis energy at the bottom end of the tweeter's range. In a "normal" setup most of the energy that reaches our ears is reverberant field energy, so the reverberant field has a very strong effect on perceived tonal balance, but nearfield listening in effect turns up the volume of the direct or first-arrival sound, so that it dominates the perceived tonal balance.
Imo a better solution, if you don't want to use a nearfield setup, is to design in an on-axis dip in that lower treble region, which will reduce the excess off-axis energy from the tweeter's wide-at-the-bottom radiation pattern. This is of course a compromise, but so was that nice flat on-axis response - we just didn't realize it at the time.
Unfortunately it is not possible for our hypothetical 6.5" two-way to have the correct spectral balance on-axis and off-axis at the same time. In order to accomplish that, we need a tweeter section that has constant directivity, or close to it. If we can keep the tweeter's pattern from flaring at the bottom end of its range, we won't have that unwanted harshness-causing off-axis energy.
Obviously the excess off-axis energy at the bottom end of the tweeter's range is not the cause of harshness in every case, but I think it happens fairly often. And combined with some other source of harshness, it's even worse.
Another coloration that I'm sensitive to, for whatever reason, is something I call "barkiness". To my ears, many otherwise promising speakers sound a bit "barky". I think this is due to excess energy somewhere in the 200 Hz to 900 Hz region, which is a secondary region of increased aural sensitivity, as shown in the curves above. In this case I don't think the cause is excess off-axis energy; I think it's often a woofer cone/surround resonance issue, but that's not the topic of this thread.
Okay, what causes harshness? Improperly used dome tweeters. But explaining how and why will take a few paragraphs. And this won't be an all-encompassing answer of course, but hopefully it will describe one possible source of harshness.
First, let's take a look a set of equal-loudness curves:
Where the curves dip lowest is where the ear is most sensitive. While the exact shape of these curves changes with level, we can clearly see that the ear is especially sensitive in the 2-5 kHz region, with sensitivity peaking at about 4 kHz. Just so happens this is the region where excess energy is often described as "harshness". So, how does this excess energy sneak into our systems?
I said earlier that improperly used dome tweeters were the culprit. In my opinion that's often where this excess energy in the 2-5 kHz region comes from, even if they measure perfectly flat through that region. How is that possible? Because that "perfectly flat" curve is an on-axis curve, and the excess energy is happening off-axis. Let me describe what's happening using an example:
Suppose we have a 6.5" two-way with our ubiquitous 1" dome tweeter, crossover frequency 2.5 kHz. At the crossover frequency, the woofer is just starting to beam - that is, its radiation pattern has narrowed to about 90 degrees (-6 dB limits). But the 1" dome tweeter wants to be nearly omnidirectional, because its diaphragm is so small in relation to a wavelength. Well the baffle keeps that from happening, acting as a 180 degree "horn" if you will, but the tweeter still has strong off-axis energy to the sides. And that strong off-axis energy holds up quite well from the crossover point up to ballpark 5 kHz. So we have excess off-axis energy in that 2-5 kHz region, which is the worst place for us to have excess energy.
Now maybe you're thinking, just treat the room and absorb it. The problem is, acoustically absorptive material is much more effective at absorbing shorter wavelengths than longer ones. So we are likely to end up with a room that is dead because we've killed all the high frequency reverberant energy, but haven't done as good a job of absorbing the energy in that 2-5 kHz lower treble region.
One possible solution would be to listen nearfield, such that the direct sound is much louder than the reverberant sound. Nearfield listening has its plusses and minuses, and is not my first choice, but it is a useful work-around for excess off-axis energy at the bottom end of the tweeter's range. In a "normal" setup most of the energy that reaches our ears is reverberant field energy, so the reverberant field has a very strong effect on perceived tonal balance, but nearfield listening in effect turns up the volume of the direct or first-arrival sound, so that it dominates the perceived tonal balance.
Imo a better solution, if you don't want to use a nearfield setup, is to design in an on-axis dip in that lower treble region, which will reduce the excess off-axis energy from the tweeter's wide-at-the-bottom radiation pattern. This is of course a compromise, but so was that nice flat on-axis response - we just didn't realize it at the time.
Unfortunately it is not possible for our hypothetical 6.5" two-way to have the correct spectral balance on-axis and off-axis at the same time. In order to accomplish that, we need a tweeter section that has constant directivity, or close to it. If we can keep the tweeter's pattern from flaring at the bottom end of its range, we won't have that unwanted harshness-causing off-axis energy.
Obviously the excess off-axis energy at the bottom end of the tweeter's range is not the cause of harshness in every case, but I think it happens fairly often. And combined with some other source of harshness, it's even worse.
Another coloration that I'm sensitive to, for whatever reason, is something I call "barkiness". To my ears, many otherwise promising speakers sound a bit "barky". I think this is due to excess energy somewhere in the 200 Hz to 900 Hz region, which is a secondary region of increased aural sensitivity, as shown in the curves above. In this case I don't think the cause is excess off-axis energy; I think it's often a woofer cone/surround resonance issue, but that's not the topic of this thread.
I would add zero crossover distortion to this. Sadly this can be added anywhere from the production to the reproduction chain. Old solid state devices with low switch speed/bandwidth and low bias were particularly susceptible. Fortunately this is getting rarer and rarer these days. It's actually getting pretty hard to find grainy solid state amps these days even in mass market gear. Cheapo opamps however still seem prone in source and line level electronics.
When doing editing work in the past this was a Major thing to watch out for. It is something that is very audible even at low level. This explains my bias towards triodes and ZNFB designs. As I said though that has changed the last few years where so many AB and even D designs that switch on their rails so fast and do so with such great timing that you get that sense of flow rather easily.
When doing editing work in the past this was a Major thing to watch out for. It is something that is very audible even at low level. This explains my bias towards triodes and ZNFB designs. As I said though that has changed the last few years where so many AB and even D designs that switch on their rails so fast and do so with such great timing that you get that sense of flow rather easily.
In the 11th post of this thread, Diapason wonders why truly flat sounds bright. I believe the reason may have to do with high frequency absorption by air. Out where the audience sits, the highs have been attenuated somewhat by passing through a lot of air. Up where the microphones (usually) are, those highs are still in full force. So when what the microphones picked up is reproduced through a "flat" system, it sounds bright to most of us. It is less likely to sound bright to a musician who is used to listening at the source.
Okay, what causes harshness? Improperly used dome tweeters. But explaining how and why will take a few paragraphs. And this won't be an all-encompassing answer of course, but hopefully it will describe one possible source of harshness.
First, let's take a look a set of equal-loudness curves:
Where the curves dip lowest is where the ear is most sensitive. While the exact shape of these curves changes with level, we can clearly see that the ear is especially sensitive in the 2-5 kHz region, with sensitivity peaking at about 4 kHz. Just so happens this is the region where excess energy is often described as "harshness". So, how does this excess energy sneak into our systems?
I said earlier that improperly used dome tweeters were the culprit. In my opinion that's often where this excess energy in the 2-5 kHz region comes from, even if they measure perfectly flat through that region. How is that possible? Because that "perfectly flat" curve is an on-axis curve, and the excess energy is happening off-axis. Let me describe what's happening using an example:
Suppose we have a 6.5" two-way with our ubiquitous 1" dome tweeter, crossover frequency 2.5 kHz. At the crossover frequency, the woofer is just starting to beam - that is, its radiation pattern has narrowed to about 90 degrees (-6 dB limits). But the 1" dome tweeter wants to be nearly omnidirectional, because its diaphragm is so small in relation to a wavelength. Well the baffle keeps that from happening, acting as a 180 degree "horn" if you will, but the tweeter still has strong off-axis energy to the sides. And that strong off-axis energy holds up quite well from the crossover point up to ballpark 5 kHz. So we have excess off-axis energy in that 2-5 kHz region, which is the worst place for us to have excess energy.
Now maybe you're thinking, just treat the room and absorb it. The problem is, acoustically absorptive material is much more effective at absorbing shorter wavelengths than longer ones. So we are likely to end up with a room that is dead because we've killed all the high frequency reverberant energy, but haven't done as good a job of absorbing the energy in that 2-5 kHz lower treble region.
One possible solution would be to listen nearfield, such that the direct sound is much louder than the reverberant sound. Nearfield listening has its plusses and minuses, and is not my first choice, but it is a useful work-around for excess off-axis energy at the bottom end of the tweeter's range. In a "normal" setup most of the energy that reaches our ears is reverberant field energy, so the reverberant field has a very strong effect on perceived tonal balance, but nearfield listening in effect turns up the volume of the direct or first-arrival sound, so that it dominates the perceived tonal balance.
Imo a better solution, if you don't want to use a nearfield setup, is to design in an on-axis dip in that lower treble region, which will reduce the excess off-axis energy from the tweeter's wide-at-the-bottom radiation pattern. This is of course a compromise, but so was that nice flat on-axis response - we just didn't realize it at the time.
Unfortunately it is not possible for our hypothetical 6.5" two-way to have the correct spectral balance on-axis and off-axis at the same time. In order to accomplish that, we need a tweeter section that has constant directivity, or close to it. If we can keep the tweeter's pattern from flaring at the bottom end of its range, we won't have that unwanted harshness-causing off-axis energy.
Obviously the excess off-axis energy at the bottom end of the tweeter's range is not the cause of harshness in every case, but I think it happens fairly often. And combined with some other source of harshness, it's even worse.
Another coloration that I'm sensitive to, for whatever reason, is something I call "barkiness". To my ears, many otherwise promising speakers sound a bit "barky". I think this is due to excess energy somewhere in the 200 Hz to 900 Hz region, which is a secondary region of increased aural sensitivity, as shown in the curves above. In this case I don't think the cause is excess off-axis energy; I think it's often a woofer cone/surround resonance issue, but that's not the topic of this thread.
Duke,
Thanks for the great explanation! Your understanding and an ability to communicate complex scientific and engineering ideas in plain language are second to none. I wish you would participate here more.
Amen! And in addition to talking you should listen to the music their gear makes!!!!
My take is that the high resolution systems are just telling it like it is.
I, now more than ever,believe that the power to our gear is really at an all time low.
My recent experiences with ERS cloth and QV2's has confirmed that dedicated lines, power conditioning,upgraded fuses and expensive power cords are great ways to clean things up, and let more music and less hash come thru those high resolution systems.
The culprit isn't the gear, don't shoot them they are just the messengers.
The high resolution devices let you hear the noise that's always present because most of us rely on the grid to power our gear.
So, in my opinon,until you clean up the power and do as much treatment as you can, some gear will always be regarded as being too harsh, sterile or edgy.
The gear is doing it's job, it's ultra revealing of the music but also the noise from the power delivery we are feeding it.
As I stated, until a couple of days ago, I felt I had done just about all I could do to treat my gear to clean power.
As I've pleasantly found out, you can do a bit more and be rewarded for your efforts.
I am very pleased I tried the QV2's and the ERS cloth tweaks.
I believe I am hearing my gear the way it was designed to sound-,very clean , detailed, resolving and free from harshness and no listener fatigue.
When you start to really fine tune your power
I, now more than ever,believe that the power to our gear is really at an all time low.
My recent experiences with ERS cloth and QV2's has confirmed that dedicated lines, power conditioning,upgraded fuses and expensive power cords are great ways to clean things up, and let more music and less hash come thru those high resolution systems.
The culprit isn't the gear, don't shoot them they are just the messengers.
The high resolution devices let you hear the noise that's always present because most of us rely on the grid to power our gear.
So, in my opinon,until you clean up the power and do as much treatment as you can, some gear will always be regarded as being too harsh, sterile or edgy.
The gear is doing it's job, it's ultra revealing of the music but also the noise from the power delivery we are feeding it.
As I stated, until a couple of days ago, I felt I had done just about all I could do to treat my gear to clean power.
As I've pleasantly found out, you can do a bit more and be rewarded for your efforts.
I am very pleased I tried the QV2's and the ERS cloth tweaks.
I believe I am hearing my gear the way it was designed to sound-,very clean , detailed, resolving and free from harshness and no listener fatigue.
When you start to really fine tune your power
Great observation on the issue of the state of power from Chuck Lee.
Agreed that with very massively increased urbanisation and just so many more i and e devices out there and the explosion of inexpensive appliances around our homes with their cheap smps is likely creating a subsequently much dirtier/noisier electrical environment for our increasingly sensitive audio systems.
Factor in a generally significantly improved loud speaker design over the last decades and also increasingly more resolving components with more digital devices as part of that and you have a perfect formula for heightened sensitivity to a cocktail of electrical noise for our systems.
In particular the kinds of low level but ugly electrical noise that give that hardness in the sound in particular those involved with the higher order odd order harmonics.
After focussing on noise minimisation in my system with the use of Shunyata and Stillpoints one of the most noticeable improvements has been in the extraordinary removal of harshness in a system while managing to maintain the vitality, beauty and inherent musical structure in the music making.
More and more I find myself in agreement with those that believe the solution to harshness isn't in any veiling of the noise but rather by use of a comprehensive system approach to electrical and mechanical isolation to first minimize it.
The battle against harshness can just as successfully be fought in the establishment of better quality system infrastructure and isolation rather than just always chasing it within the design of the primary components alone.
Agreed that with very massively increased urbanisation and just so many more i and e devices out there and the explosion of inexpensive appliances around our homes with their cheap smps is likely creating a subsequently much dirtier/noisier electrical environment for our increasingly sensitive audio systems.
Factor in a generally significantly improved loud speaker design over the last decades and also increasingly more resolving components with more digital devices as part of that and you have a perfect formula for heightened sensitivity to a cocktail of electrical noise for our systems.
In particular the kinds of low level but ugly electrical noise that give that hardness in the sound in particular those involved with the higher order odd order harmonics.
After focussing on noise minimisation in my system with the use of Shunyata and Stillpoints one of the most noticeable improvements has been in the extraordinary removal of harshness in a system while managing to maintain the vitality, beauty and inherent musical structure in the music making.
More and more I find myself in agreement with those that believe the solution to harshness isn't in any veiling of the noise but rather by use of a comprehensive system approach to electrical and mechanical isolation to first minimize it.
The battle against harshness can just as successfully be fought in the establishment of better quality system infrastructure and isolation rather than just always chasing it within the design of the primary components alone.
If you want to kill AC noise problems, the best line conditioners were made by Elgar. One of their larger units (model 3006) was capable of 28Amps continuous, while at the same time putting out a very low distortion sine wave. No high end audio conditioner can do that, plain and simple.
Elgar saw the conditioner market shrinking years ago and got out of the niche- now they make AC power sources, which are not the same thing at all. Here is one of the smaller AC conditioner models that they made, it can run a complete stereo if the amps are not too big. IME the units are mechanically noisy and so should be wired to the breaker box and then feed the line that goes to the audio room.
http://www.ebay.com/itm/Elgar-Model...928?pt=LH_DefaultDomain_0&hash=item35df4477a0
Once you hear what one of these does all the other AC line tweaks and such will go by the wayside.
Elgar saw the conditioner market shrinking years ago and got out of the niche- now they make AC power sources, which are not the same thing at all. Here is one of the smaller AC conditioner models that they made, it can run a complete stereo if the amps are not too big. IME the units are mechanically noisy and so should be wired to the breaker box and then feed the line that goes to the audio room.
http://www.ebay.com/itm/Elgar-Model...928?pt=LH_DefaultDomain_0&hash=item35df4477a0
Once you hear what one of these does all the other AC line tweaks and such will go by the wayside.
