Measurements and the Correlation of What We Hear
- Thread starter mep
- Start date
No, not at all, I meant work as in activity, or focus of one's attention. Could be doing it in the future, though ...
To put it bluntly, when you get a system working as well as Roger speaks of, all the idle talk about the significance of various measurements means sweet nothing ...
Frank
When you advocate a position you have to be prepared to deal with the consequences of trying to implement your idea. Let's suppose we are in a political debate. You say, if elected I will end our dependence on foreign oil. I respond that i am against expansion of offshore drilling because of what happened in the Gulf. Furthermore studies show alternative energy are inadequate to fill the gap that foreign oil. That's not a Strawman argument. Without saying it, I have agreed that ending our dependence on foreign oil is a good idea. Too defeat your argument I pointed out that foreign oil is bad but your cure is worse than the problem. The fact is we can correlate some measurements with what we hear. I went directly to the problems associated with that effort. Maybe I should have been clearer.
Sorry it must be written in "engineering code"
It's not in your quote. Here is what you promised. "including adjusting his amps to purposely generate some even order distortion cause it "sounds" better." I'll keep looking.
Of course that is a discontinued amp. Do you a current example?
It's not in your quote. Here is what you promised. "including adjusting his amps to purposely generate some even order distortion cause it "sounds" better." I'll keep looking.
Of course that is a discontinued amp. Do you a current example?
Last edited:
Tom here is what I said:
He is your example of "knowing all there is to know about amps." The fact that they are more expensive is self evident. As far as being better that's his claim. I can only assume he really thinks they are better. If not he is either incompetent (that is at odds with your claim that he is omniscient about all things amplifier related) or that he is engaged in some slight of hand (that would make him corrupt).
Are you privy to any measurements revealing he has purposely added even order distortion to "make it sound better?" Or were you accusing someone else? Would you care to name names?
Here is what you said:
Greg, I am not designing an escape hatch at all. It is important to describe what one means as "better". It may mean ones preference, or may mean better THD spec or nicer cabinet or higher cost parts inside or any number of things. From your references above do I glean that the facts
he presents of lower THD and noise mean "better" to you, or is it simply what your ears prefer? I think the word "better" is fairly hard to pin down in the audio world!
I don't have an opinion either on the amps quality or whether it is better than its predecessor. It's the job of Nelson Pass to prove it is better. It was you who cited Nelson Pass as an expert not me.
As for my position on preference this is a discussion I had with Tim:
Yes many have taken the position that analog at its best is better than digital at its best. They seemingly miss few opportunities to point that out. It is exactly my point. That is stated as an absolute. Your continued characterization of it as a mere preference is dismissive and an incorrect statement of the position.
Saying I like or prefer analog over digital is a statement of preference. So if Myles says I like analog tape best, that is a matter of preference. If he says analog tape is best, that is an absolute to be argued against, if you disagree. Or even to require additional information if you are unpersuaded.
He is your example of "knowing all there is to know about amps." The fact that they are more expensive is self evident. As far as being better that's his claim. I can only assume he really thinks they are better. If not he is either incompetent (that is at odds with your claim that he is omniscient about all things amplifier related) or that he is engaged in some slight of hand (that would make him corrupt).
Are you privy to any measurements revealing he has purposely added even order distortion to "make it sound better?" Or were you accusing someone else? Would you care to name names?
Here is what you said:
Greg, I am not designing an escape hatch at all. It is important to describe what one means as "better". It may mean ones preference, or may mean better THD spec or nicer cabinet or higher cost parts inside or any number of things. From your references above do I glean that the facts
he presents of lower THD and noise mean "better" to you, or is it simply what your ears prefer? I think the word "better" is fairly hard to pin down in the audio world!
I don't have an opinion either on the amps quality or whether it is better than its predecessor. It's the job of Nelson Pass to prove it is better. It was you who cited Nelson Pass as an expert not me.
As for my position on preference this is a discussion I had with Tim:
Yes many have taken the position that analog at its best is better than digital at its best. They seemingly miss few opportunities to point that out. It is exactly my point. That is stated as an absolute. Your continued characterization of it as a mere preference is dismissive and an incorrect statement of the position.
Saying I like or prefer analog over digital is a statement of preference. So if Myles says I like analog tape best, that is a matter of preference. If he says analog tape is best, that is an absolute to be argued against, if you disagree. Or even to require additional information if you are unpersuaded.
garylkoh
WBF Technical Expert (Speakers & Audio Equipment)
I've read through all 24 pages twice this evening hoping to find something useful. Going back to Mark's original post, I've been trying to correlate measurements to what *I* hear for the past few years, and haven't had any luck - once the measurements are "good enough". Let's take 2 examples of something that we can hear.....
I am assuming that for an amplifier the frequency response and phase are flat to +/- 1 dB from 20Hz to 20kHz (and even beyond), THD, IMD are below 0.1%. I think that we know that we already know that we can measure competence. I've got a competent amp - see this post I put up in the beginning of the year:
http://www.whatsbestforum.com/showt...Analog-Dialectic&p=35495&viewfull=1#post35495
The PNWAS members have heard it, and none of them have said that it sounded incompetently designed.
With that amplifier, adding large storage capacitors to the power supply improved dynamic ability and produced a sound that may be described subjectively as "more muscular", "beefy sound", "more meat on the bones", as opposed to "thin". Frequency response into either a dummy resistive load, or a real loudspeaker was +/- 0.1dB from 22 to 18k. I measured at 1W, 10W, and even 20W and 100W into the dummy load.
I figured that with "beefy", I could measure THD using bass frequencies. So, I plotted THD from 20Hz to 20kHz at 20W and then added or took away large capacitors in the power supply. No statistically significant difference.
Adding more small, fast capacitors to the power supply allowed more resolution and detail. I thought that I might be able to find the measurement for resolution and detail at the very low power IMD, but the difference was so tiny that it could have been due to the hook-up I used (alligator clips). Increasing the line resolution from 1 pixel to 3 pixels effectively erased that difference on screen.
Measuring THD and IMD using a real speaker load and a resistor did reveal some difference, but not enough to make any sense of, but it was larger than the difference in IMD adding speedy capacitors to the power supply.
I also tried different IMD frequencies and mixes. I can't remember all the combinations, but SMPTE which specifies 60Hz and 7kHz was one of them. I mixed the two tones at various ratios from 4:1 to 1:1. (I guess that satisfies fas42's test of a bass frequency below 100Hz) DIN spec measured a hair lower at higher power because it's 250Hz and 8kHz.
I also looked at TIM (transient intermodulation distortion) at high power and low power.
After failing using the suite of measurements available on CLIO (the software I use), I tried to derive something different. To measure how well an amplifier resolves micro-dynamic detail (the quiet breathing of the singer while the rest of the band is doing a solo), I figured that if I injected a 80Hz large signal into the amplifier, and then measured the THD of a 1kHz signal, we might be able to see something.
The 80Hz would have to be filtered out somehow, so I built a 4th order passive filter at 80Hz, and ran the amplifier into a dummy load. I compared the THD measurement (the passive crossover increased the THD at the load by about 0.5%) with and without the 80Hz signal. I measured the THD at 1kHz, subtracted the 0.5% from the crossover, and subtracted the 80Hz THD measurement - no statistically significant difference.
After about a month of that, I figured that I was a better listener than a measurer, and did the rest of that part of my design with my ears.
If anyone can come up with a set of measurements that can measure the resolution and dynamic ability of an amplifier, I'm all ears
On the other hand, I *think* that I know what measurement you can make for "image density" or "image sharpness" of an amplifier. Measure the group delay of each channel across frequency and with respect to power. If the group delay tracks well between the two channels, you get a sharp image. If the group delay wavers with power, you lose focus. If the group delay wavers with respect to frequency, you lose PRaT (but only if you believe in PRaT).
I am assuming that for an amplifier the frequency response and phase are flat to +/- 1 dB from 20Hz to 20kHz (and even beyond), THD, IMD are below 0.1%. I think that we know that we already know that we can measure competence. I've got a competent amp - see this post I put up in the beginning of the year:
http://www.whatsbestforum.com/showt...Analog-Dialectic&p=35495&viewfull=1#post35495
The PNWAS members have heard it, and none of them have said that it sounded incompetently designed.
With that amplifier, adding large storage capacitors to the power supply improved dynamic ability and produced a sound that may be described subjectively as "more muscular", "beefy sound", "more meat on the bones", as opposed to "thin". Frequency response into either a dummy resistive load, or a real loudspeaker was +/- 0.1dB from 22 to 18k. I measured at 1W, 10W, and even 20W and 100W into the dummy load.
I figured that with "beefy", I could measure THD using bass frequencies. So, I plotted THD from 20Hz to 20kHz at 20W and then added or took away large capacitors in the power supply. No statistically significant difference.
Adding more small, fast capacitors to the power supply allowed more resolution and detail. I thought that I might be able to find the measurement for resolution and detail at the very low power IMD, but the difference was so tiny that it could have been due to the hook-up I used (alligator clips). Increasing the line resolution from 1 pixel to 3 pixels effectively erased that difference on screen.
Measuring THD and IMD using a real speaker load and a resistor did reveal some difference, but not enough to make any sense of, but it was larger than the difference in IMD adding speedy capacitors to the power supply.
I also tried different IMD frequencies and mixes. I can't remember all the combinations, but SMPTE which specifies 60Hz and 7kHz was one of them. I mixed the two tones at various ratios from 4:1 to 1:1. (I guess that satisfies fas42's test of a bass frequency below 100Hz) DIN spec measured a hair lower at higher power because it's 250Hz and 8kHz.
I also looked at TIM (transient intermodulation distortion) at high power and low power.
After failing using the suite of measurements available on CLIO (the software I use), I tried to derive something different. To measure how well an amplifier resolves micro-dynamic detail (the quiet breathing of the singer while the rest of the band is doing a solo), I figured that if I injected a 80Hz large signal into the amplifier, and then measured the THD of a 1kHz signal, we might be able to see something.
The 80Hz would have to be filtered out somehow, so I built a 4th order passive filter at 80Hz, and ran the amplifier into a dummy load. I compared the THD measurement (the passive crossover increased the THD at the load by about 0.5%) with and without the 80Hz signal. I measured the THD at 1kHz, subtracted the 0.5% from the crossover, and subtracted the 80Hz THD measurement - no statistically significant difference.
After about a month of that, I figured that I was a better listener than a measurer, and did the rest of that part of my design with my ears.
If anyone can come up with a set of measurements that can measure the resolution and dynamic ability of an amplifier, I'm all ears
On the other hand, I *think* that I know what measurement you can make for "image density" or "image sharpness" of an amplifier. Measure the group delay of each channel across frequency and with respect to power. If the group delay tracks well between the two channels, you get a sharp image. If the group delay wavers with power, you lose focus. If the group delay wavers with respect to frequency, you lose PRaT (but only if you believe in PRaT).
I think that you focused on the one sentence in my post that fits your point of view.
Not really. I use subjective reports but I put more emphasis on knowledge, logic, experiment and measurement. I hope that WBF contains more than subjective audio experiences.
When you come down to it, most people a) have financial limits and b) believe that audio gear is subject to the law of diminishing returns. I certainly value making careful choices that fit your budget and that you can live with for a long time. If cheap gear works for a person, why not.
I use $ 350/pair powered AudioEngiine A5s in my home office and find listening to music on them to be enjoyable. My dedicated MusicPC and my personal PC are both connected to a pro audio DAC that I bought for under $ 300 via PCI soundcards that cost $ 129 and $ 80-90. I did the best job I could to understand how much I needed to spend and what the best alternatives at that level of cost were. My choices turned out pretty well and I could listen to the office system for a long time.
My next upgrade for that system will be active pro-audio monitors of the sort you mentioned. I'll be looking for some practical improvements relative to the A5s: larger diameter woofer, higher quality drivers, more power and crossovers in front of the amps. I'll use whatever test results I can find, subjective reports from owners and home listening trials. I will certainly try to calibrate how much I need to spend to get a worthwhile level of improvement.
Bill
Not really. I use subjective reports but I put more emphasis on knowledge, logic, experiment and measurement. I hope that WBF contains more than subjective audio experiences.
When you come down to it, most people a) have financial limits and b) believe that audio gear is subject to the law of diminishing returns. I certainly value making careful choices that fit your budget and that you can live with for a long time. If cheap gear works for a person, why not.
I use $ 350/pair powered AudioEngiine A5s in my home office and find listening to music on them to be enjoyable. My dedicated MusicPC and my personal PC are both connected to a pro audio DAC that I bought for under $ 300 via PCI soundcards that cost $ 129 and $ 80-90. I did the best job I could to understand how much I needed to spend and what the best alternatives at that level of cost were. My choices turned out pretty well and I could listen to the office system for a long time.
My next upgrade for that system will be active pro-audio monitors of the sort you mentioned. I'll be looking for some practical improvements relative to the A5s: larger diameter woofer, higher quality drivers, more power and crossovers in front of the amps. I'll use whatever test results I can find, subjective reports from owners and home listening trials. I will certainly try to calibrate how much I need to spend to get a worthwhile level of improvement.
Bill
Last edited:
Excellent exercise you went through, Gary, that's the sort of process that needs to be extended and developed. I believe that refinement and further investigation will serve you well ...
People may ask, why I'm not trying something like this: a variety of reasons complicate the situation, and if someone else already has the situation well in hand it may be a faster route to getting significant results.
Frank
Exactly the result I would have expected: this is a macro measurement and would have picked up extremely little.
Possibly. Were you looking at spectrum results?
This could go somewhere, and since you're prepared to put some real energy and time into this exercise, may I ask, as a starting point, the following: if you repeated the 80Hz test can you do a spectrum analysis of the result, we're not interested in any total figures like THD, and if so, what precision does your hardware and software setup allow, in terms of measuring a low level peak adjacent to a high level peak?
People may ask, why I'm not trying something like this: a variety of reasons complicate the situation, and if someone else already has the situation well in hand it may be a faster route to getting significant results.
Actually, what you demonstrated is that the ears are an excellent measuring tool, extremely sophisticated in terms of filtering away extraneous detail, and that beat the pants of any device you stick into a point in the wall in getting an accurate assessment of progress.
Frank
Nice post Gary.
The problem I see Gary is that this may give a broad indication-estimation of an affect towards the sound (more critically the measurement will help in assisting in testing the amp's linearity-behaviour-trait IMO).
As an example can you equate those measurements to subtle changes in the partials of an instrument's complex wave, and also the effect on said complex wave on its attack-sustain-decay?
Just to recap this thread is about correlating measurements to what we hear; therefore it is how we hear the music or instrument that has been reproduced by an audio system.
With that in mind, you would need to be able to take the engineering type measurements you mention and then model those to reflect the music's/instrument's waveform.
Because this is an example of what we hear (actual instrument is a trumpet):
Therefore looking at the measurements used for amps-DAC-etc and those you mention, could you honestly say how it would affect the original waveform in the above example?
Especially as what we hear is critically affected by the fundamental-partials, and their temporal and frequency amplitude (energy), in other words how that image would change subtly.
This does not even consider soundstage related setup-config of mics and associated recording-mastering information that is also present in recordings.
So IMO engineering measurement may give a broad indication of how the sound may deviate to what it should be, but it does not truly tell us the outputted waveform of the reproduced instrument and music.
Which is what we hear (the ideal measurement being the actual temporal-frequency domain of a both harmonic and inharmonic complex waveform derived from a real instrument and note-chord).
But as I said engineering measurements are superb at evaluating the linearity-stability-behaviour of an product.
Just to stress, engineering measurements can be as diverse and complex as required, from frequency response to a gaussian pulse,etc.
Also to add technically engineering and science when it comes to measurements cannot be differentiated in the way I am doing, but this is the easiest way to show the variables-factors and requirements each look at when it comes to sounds; audio equipment reproducing the music-sound, or the science of what is sound and the complex waveform applied to instruments-music and how we perceive them in terms of timbre-sensation of tones-texture-instrument differences-etc.
This is just expanding upon my post #53 in this thread, and on some links I have provided in the past.
Just my take on this, which is why for this specific discussion and thread I feel the only relevant measurement must be the scientific approach.
Being; What is sound and measure a "real" (can be derived from waveform synthesis) instrument in terms of temporal and frequency domain and its fundamental-partials, which can be replayed consistently (so would need to be a specific note or chord or multiples over parts of the scale) on the audio gear in question and compared to the original.
Another link I am providing is useful summary as it explains many of the critical factors relating to sound quality; from timbre to vibrato/tremolo.
http://hyperphysics.phy-astr.gsu.edu/hbase/Sound/timbre.html
Definitely worth reading the whole page, especially the lower aspect as it fits in with one of the science papers I linked before that mentions spectral fusion.
To quote a part of that science paper (please appreciate it goes back to the '80s) as it shows interesting aspects of sound quality and its nature:
Thanks and hope the info is of interest, especially in the context of this thread; correlating measurements to what we hear.
And importantly the distinction of the variables-factors and measurement requirements when it comes to sounds; audio equipment reproducing the music-sound, or the science of what is sound and the complex waveform applied to instruments-music and how we perceive them in terms of timbre-sensation of tones-texture-instrument differences-etc
Orb
The problem I see Gary is that this may give a broad indication-estimation of an affect towards the sound (more critically the measurement will help in assisting in testing the amp's linearity-behaviour-trait IMO).
As an example can you equate those measurements to subtle changes in the partials of an instrument's complex wave, and also the effect on said complex wave on its attack-sustain-decay?
Just to recap this thread is about correlating measurements to what we hear; therefore it is how we hear the music or instrument that has been reproduced by an audio system.
With that in mind, you would need to be able to take the engineering type measurements you mention and then model those to reflect the music's/instrument's waveform.
Because this is an example of what we hear (actual instrument is a trumpet):
Therefore looking at the measurements used for amps-DAC-etc and those you mention, could you honestly say how it would affect the original waveform in the above example?
Especially as what we hear is critically affected by the fundamental-partials, and their temporal and frequency amplitude (energy), in other words how that image would change subtly.
This does not even consider soundstage related setup-config of mics and associated recording-mastering information that is also present in recordings.
So IMO engineering measurement may give a broad indication of how the sound may deviate to what it should be, but it does not truly tell us the outputted waveform of the reproduced instrument and music.
Which is what we hear (the ideal measurement being the actual temporal-frequency domain of a both harmonic and inharmonic complex waveform derived from a real instrument and note-chord).
But as I said engineering measurements are superb at evaluating the linearity-stability-behaviour of an product.
Just to stress, engineering measurements can be as diverse and complex as required, from frequency response to a gaussian pulse,etc.
Also to add technically engineering and science when it comes to measurements cannot be differentiated in the way I am doing, but this is the easiest way to show the variables-factors and requirements each look at when it comes to sounds; audio equipment reproducing the music-sound, or the science of what is sound and the complex waveform applied to instruments-music and how we perceive them in terms of timbre-sensation of tones-texture-instrument differences-etc.
This is just expanding upon my post #53 in this thread, and on some links I have provided in the past.
Just my take on this, which is why for this specific discussion and thread I feel the only relevant measurement must be the scientific approach.
Being; What is sound and measure a "real" (can be derived from waveform synthesis) instrument in terms of temporal and frequency domain and its fundamental-partials, which can be replayed consistently (so would need to be a specific note or chord or multiples over parts of the scale) on the audio gear in question and compared to the original.
Another link I am providing is useful summary as it explains many of the critical factors relating to sound quality; from timbre to vibrato/tremolo.
http://hyperphysics.phy-astr.gsu.edu/hbase/Sound/timbre.html
Definitely worth reading the whole page, especially the lower aspect as it fits in with one of the science papers I linked before that mentions spectral fusion.
To quote a part of that science paper (please appreciate it goes back to the '80s) as it shows interesting aspects of sound quality and its nature:
Thanks and hope the info is of interest, especially in the context of this thread; correlating measurements to what we hear.
And importantly the distinction of the variables-factors and measurement requirements when it comes to sounds; audio equipment reproducing the music-sound, or the science of what is sound and the complex waveform applied to instruments-music and how we perceive them in terms of timbre-sensation of tones-texture-instrument differences-etc
Orb
Last edited:
Orb, I would just say you're over-complicating the picture of what an audio signal looks like: no matter how complex the sound, it is ultimately nothing more than a meandering "wiggle", a pure sine wave is just a wiggle that meanders in a very perfect, smooth pattern. At the level that the electronics have to look at what's going on, the audio signal is really very easy to deal with, the problems arise because something disturbs the precise path the meandering takes, or the electronics run out of puff, the power supplies can't keep up with what the circuitry is demanding from them in a clean enough fashion.
Frank
Frank
Thanks Gary. You always bring a reasoned approach to these matters.
It's always difficult to find someone who has the motive, opportunity and ability to bring us the truth. Even more difficult to come up with something an audiophile can use to assemble a real world system.
It's always difficult to find someone who has the motive, opportunity and ability to bring us the truth. Even more difficult to come up with something an audiophile can use to assemble a real world system.
To put this into perspective Frank,
simple task please describe accurately how a piano chord will sound using existing measurements shown on Stereophile for two different amps (1 class A and 1 AB), description in terms of texture,tonality, musicality-how it flows, timbre brightness, lean-cool,etc.
You cannot, and that is what this thread is about; what we hear relating to audio and that is specifically inharmonic and harmonic sounds generated by voices and instruments.
Thanks
Orb
Orb
Notions like "musicality", "flow", "rightness", etc are our attempt to describe our aural perceptions with words, I will grant you, a difficult proposition. Those perceptions however, if they exist at all, are our body responses to physical thus measurable stimuli. Sounds exist in time and space and that all there is to it all measurable. I am somehow unsettled to see how some of us are bending around incredible loops to come up with the notion that something that exists (a stimulus which creates a perception) can't be measured ..i sincerely would have not thought this line of reasoning ..well.. reasonable but .. we are audiophiles ... we believe in stranger things ...
I will answer Gary post later .. It seems at first reading to belong in the "we-can-hear-it-but-can't-measure-it" category ... I disagree with that line of thought...
Notions like "musicality", "flow", "rightness", etc are our attempt to describe our aural perceptions with words, I will grant you, a difficult proposition. Those perceptions however, if they exist at all, are our body responses to physical thus measurable stimuli. Sounds exist in time and space and that all there is to it all measurable. I am somehow unsettled to see how some of us are bending around incredible loops to come up with the notion that something that exists (a stimulus which creates a perception) can't be measured ..i sincerely would have not thought this line of reasoning ..well.. reasonable but .. we are audiophiles ... we believe in stranger things ...
I will answer Gary post later .. It seems at first reading to belong in the "we-can-hear-it-but-can't-measure-it" category ... I disagree with that line of thought...
I've read through all 24 pages twice this evening hoping to find something useful. Going back to Mark's original post, I've been trying to correlate measurements to what *I* hear for the past few years, and haven't had any luck - once the measurements are "good enough". Let's take 2 examples of something that we can hear.....
I am assuming that for an amplifier the frequency response and phase are flat to +/- 1 dB from 20Hz to 20kHz (and even beyond), THD, IMD are below 0.1%. I think that we know that we already know that we can measure competence. I've got a competent amp - see this post I put up in the beginning of the year:
http://www.whatsbestforum.com/showt...Analog-Dialectic&p=35495&viewfull=1#post35495
The PNWAS members have heard it, and none of them have said that it sounded incompetently designed.
With that amplifier, adding large storage capacitors to the power supply improved dynamic ability and produced a sound that may be described subjectively as "more muscular", "beefy sound", "more meat on the bones", as opposed to "thin". Frequency response into either a dummy resistive load, or a real loudspeaker was +/- 0.1dB from 22 to 18k. I measured at 1W, 10W, and even 20W and 100W into the dummy load.
I figured that with "beefy", I could measure THD using bass frequencies. So, I plotted THD from 20Hz to 20kHz at 20W and then added or took away large capacitors in the power supply. No statistically significant difference.
Adding more small, fast capacitors to the power supply allowed more resolution and detail. I thought that I might be able to find the measurement for resolution and detail at the very low power IMD, but the difference was so tiny that it could have been due to the hook-up I used (alligator clips). Increasing the line resolution from 1 pixel to 3 pixels effectively erased that difference on screen.
Measuring THD and IMD using a real speaker load and a resistor did reveal some difference, but not enough to make any sense of, but it was larger than the difference in IMD adding speedy capacitors to the power supply.
I also tried different IMD frequencies and mixes. I can't remember all the combinations, but SMPTE which specifies 60Hz and 7kHz was one of them. I mixed the two tones at various ratios from 4:1 to 1:1. (I guess that satisfies fas42's test of a bass frequency below 100Hz) DIN spec measured a hair lower at higher power because it's 250Hz and 8kHz.
I also looked at TIM (transient intermodulation distortion) at high power and low power.
After failing using the suite of measurements available on CLIO (the software I use), I tried to derive something different. To measure how well an amplifier resolves micro-dynamic detail (the quiet breathing of the singer while the rest of the band is doing a solo), I figured that if I injected a 80Hz large signal into the amplifier, and then measured the THD of a 1kHz signal, we might be able to see something.
The 80Hz would have to be filtered out somehow, so I built a 4th order passive filter at 80Hz, and ran the amplifier into a dummy load. I compared the THD measurement (the passive crossover increased the THD at the load by about 0.5%) with and without the 80Hz signal. I measured the THD at 1kHz, subtracted the 0.5% from the crossover, and subtracted the 80Hz THD measurement - no statistically significant difference.
After about a month of that, I figured that I was a better listener than a measurer, and did the rest of that part of my design with my ears.
If anyone can come up with a set of measurements that can measure the resolution and dynamic ability of an amplifier, I'm all ears
On the other hand, I *think* that I know what measurement you can make for "image density" or "image sharpness" of an amplifier. Measure the group delay of each channel across frequency and with respect to power. If the group delay tracks well between the two channels, you get a sharp image. If the group delay wavers with power, you lose focus. If the group delay wavers with respect to frequency, you lose PRaT (but only if you believe in PRaT).
Gary.....Thank You! Your observations to me is the most informative I have read on this subject in quite a while.
Your experience validates somewhat what I beleive to be the case in all of this conversation. Audio equipment is the sum of design,components,knowledge,and art. That is more so today then yesterday because new technologies have benifited most everything in this business. Also I think there are more break throughs coming as low level noise and distortion and it's effects on reproduction,is more understood by the manufacturer and end user. Bravo
Gary while we have your attention would you care to weigh on why speakers for the most part don't publish distortion figures. Is it that speakers don't lend themselves to easy to those type of specs. like amps? Is the distortion so bad that it's just not a good marketing idea? Or do you feel that distortion can be extrapolated form the present batetry of measurements ? (Hope there are not strawman arguments in there. Smile)
If anyone can come up with a set of measurements that can measure the resolution and dynamic ability of an amplifier, I'm all ears
On the other hand, I *think* that I know what measurement you can make for "image density" or "image sharpness" of an amplifier. Measure the group delay of each channel across frequency and with respect to power. If the group delay tracks well between the two channels, you get a sharp image. If the group delay wavers with power, you lose focus. If the group delay wavers with respect to frequency, you lose PRaT (but only if you believe in PRaT).
I agree, if we can optimise the amplifier,then we can optimise other electronics. Probably a bigger effect is current distortion and it's delivery to the eletronics,but that's another subject. One thing at a time.
IIRC a designer friend of mine used Fourier transform and analysis in circuit design, I always thought his amplifier was exceptional.
http://www.audiorecording.me/fast-fourier-transform-to-view-audio-frequency-spectrum.html
Last edited:
garylkoh
WBF Technical Expert (Speakers & Audio Equipment)
With speakers, it is tough to distinguish between distortion because of the loudspeaker, distortion because of the measuring microphone, and even distortion because of air itself - since the air needs to transmit the signal from the loudspeaker to the microphone. With amplifiers, it is far easier because we know that the transmission - the cable - is perfect even when it is a just zip cord. I've never been able to measure even a smidgen of distortion in any wire.
It's there, I have heard it,it mainly effects resolution and the effects can be startling,no pun intended. Just to clarify (the current that goes through the wire).
garylkoh
WBF Technical Expert (Speakers & Audio Equipment)
Just to recap this thread is about correlating measurements to what we hear; therefore it is how we hear the music or instrument that has been reproduced by an audio system.
With that in mind, you would need to be able to take the engineering type measurements you mention and then model those to reflect the music's/instrument's waveform.
Because this is an example of what we hear (actual instrument is a trumpet):
Therefore looking at the measurements used for amps-DAC-etc and those you mention, could you honestly say how it would affect the original waveform in the above example?
Orb, Thanks.
The problem with a real-world measurement like this is that it is difficult to compare the input and the output. The complexity of the input means that jitter in the measuring instrument would invalidate the result.
The only significant measurements I've seen in this area have been the experiments conducted by Nordost/Vertex - but because they use a PC and soundcard for the measurement, they have been discredited on this forum even though they say that they used math to match the input and the measured output.
All the studies I've read on this subject have been done with listening panels, and this is not the subject of this thread - which is measurements. As Frantz mentioned, these are bodily perceptions stimulated by some physical stimuli.... and if they are physical, there are many here who are sure that if it can be perceived, then they can be measured.
I'm of the camp "I can hear it, but *I* don't know how to measure it, and I would love for someone to tell me how".
Similar threads
