Is your sound source (headphone) has so low distortion? or switching on off the harmonic you sense increase adding to that of the source?
For those with no measuring gear Soundcard Scope is brilliant free software,though limited to 20khz. It's THD meter is as accurate as a HP3903B, this with a so so SC. The FFT is useful for tweaking stages to dial in harmonics, EG I make a lot of parameters on my amps adjustable, like LTP CCS bias, adjustable fixed bias etc. After you've done the load lines & prototyped it's nice to be able to tweak for a few more 0.xx % improvement
Andy.
Andy.
I remember everyone working at that publication being deaf. Starting with egg-head 😀
I doubt those blind tests are meanigful ... only if you know your gear well ( speakers + amplifier in your listening space ) and your music you will hear subtle changes . Then you can double check and so on ...
Last edited:
Standing waves in a room can exaggerate LF reproduction, known as the most important point in rejection of a otherwise good pair of loudspeakers. From hifi fora I get the impression people are obsessed with exchanging amplifiers and speakers to get a mix suitable to taste. Next comes cabling and the process repeats from beginning. There is a point one should accept the flaws and live with it.
Designing gear and modding to an (personal) acceptable level is a hobby of its own and should be considered as such. Apply yourself and enjoy while doing so.
Hi ! may i ask which sound card are you using and which operating system ? thanks a lot
Many people are deaf.
Some subtle changes in a sound system will not be heard by those deaf people.
When you can tell me if the clarinetist on a recording that is being played back on your system is using a number 1 reed, and not using a number 1 1/2 reed, please let me know.
(If you happen to be a clarinetist, or married to a clarinetist, you do not qualify for this challenge, recuse yourself).
. . . Then I want to know all the particulars of your system, including the room dimensions and treatment, power mains voltage, temperature, air pressure, what recording and recording media are used, playback equipment of the recording, preamp, amp, and speaker, time of day, what you ate, your physical condition at that time (and anything else that is particular to your system).
Just another subtle sound difference, that is easy for some to hear.
Enjoy the music.
Some subtle changes in a sound system will not be heard by those deaf people.
When you can tell me if the clarinetist on a recording that is being played back on your system is using a number 1 reed, and not using a number 1 1/2 reed, please let me know.
(If you happen to be a clarinetist, or married to a clarinetist, you do not qualify for this challenge, recuse yourself).
. . . Then I want to know all the particulars of your system, including the room dimensions and treatment, power mains voltage, temperature, air pressure, what recording and recording media are used, playback equipment of the recording, preamp, amp, and speaker, time of day, what you ate, your physical condition at that time (and anything else that is particular to your system).
Just another subtle sound difference, that is easy for some to hear.
Enjoy the music.
Last edited:
also when listening to analogue sources ? the much praised sound from analog tapes is coming through solid state circuits 🙄
There can be a reason why tubes are popular in dac output stages
Hi ! ok tube recorders can sound very good
Like Studer, Otari, Ampex without tubes
When people used to listen LPs solid state amps where the norm
Now with digital tubes have come back
Personally i like what just one tube can give in a all solid state chain a lot
I really want to learn how to test for distortion on each stage and overall before I start listening. I work hard at making my amps inoffensive..ok not to hard.. on the poorest digital sources. I like transparency but not if it's going to have me avoiding some video broadcast because they have a bad mic or something. I'm sure the more I learn how to reduce distortion the less cut try and fail I will have to do. I need to learn the rules before I keep breaking them. Like jazz first you must master the instrument before you improvise. I have mostly been improvising so far.. I really love the sound of trans and tape heads. That's the warmth I like.Tape is lovely.
Can't remember details of this study, but the result was that the brain tries to dictate something which had to be logically better and the feelings about that could be really the opposite of that. It could describe this "the better the measurements, the better the sound must be" dogma of the print magazines.
In my all tube, DIY, high res, high efficiency system I can clearly hear when tube audio and tube mixers did the recording and when (in the 1970s) transistor gear began to do the job. For example the Beatles recordings with the Studer C37 machine have excellent transparency and tone imaging. Later, when they switched to Studer A80 (transistor) the sound became boring, flat, less 3D feeling and transparency was much reduced to transistor level.
I'm not not pointing in the direction we all should stop listening to mediocre recordings and buy instead those super discs only, but it is clearly audible what the recording and mixing process was being made with. I don't think thats possible with transistor gear playback system, as everything on it will sound like transistor and therefore couldn't be bettered.
Last edited:
I have the capacity and equipment to measure to perhaps 0.25% THD, depending on how the signal generator feels on a particular day.
I always aim for THD of less than 1%, aiming for 2nd to be -40dB or better.
It isn't always possible, and with a benign spectrum, I'm not sure I can hear the difference between 0.5% and 1%, so I dont chase the 0.001% improvement.
Certainly though, I spend considerable time, measuring THD and tweaking the circuit to get it as low as I can, within the limits of my instruments.
I always aim for THD of less than 1%, aiming for 2nd to be -40dB or better.
It isn't always possible, and with a benign spectrum, I'm not sure I can hear the difference between 0.5% and 1%, so I dont chase the 0.001% improvement.
Certainly though, I spend considerable time, measuring THD and tweaking the circuit to get it as low as I can, within the limits of my instruments.
Is all THD and IM testing done with a non-inductive power resistor?
What about using a real load, like the Elliptical load of an actual loudspeaker?
I think that might make the THD and IM distortion worse in some cases.
And, the distortion will be much more variable versus the frequency of the test tone(s).
I did not say that testing with a real loudspeaker load is easy.
That is up to you.
I am just saying.
What about using a real load, like the Elliptical load of an actual loudspeaker?
I think that might make the THD and IM distortion worse in some cases.
And, the distortion will be much more variable versus the frequency of the test tone(s).
I did not say that testing with a real loudspeaker load is easy.
That is up to you.
I am just saying.
For me, more than measuring URI is not necessary to tune a good amp or system.
If something is wrong with the system (and no system is perfect), my ears will indicate it immediately, nearly after some seconds of listening. And thats, because I have a model of sound stored in the brain which is the perfect model for me.
Question:
The following situation: A chef de cuisine tries to fully satisfy the taste of his customers with a premium dinner (fife star quality). What does he do to achieve best results? Does he have a chemical laboratory in the kitchen, trying to analyse the right mix of ingredients at the right temperature to make for a perfect taste? Nonsense. He goes to the market early to buy the finest ingredients. He cooks the dinner and often tastes it, trying to tweak the result to what he had stored as the best taste he has ever known.
Thats the art of doing it. But if someone has no clue of what to achieve and how to achieve, if he had never tasted something real good, how could he try to achieve? Correct, he would go after measurements. Measurements and measuremts, because there is nothing more he knows.
To achieve a flat response, to achieve low THD, to achieve a wide bandwith, to achieve a short impulse answer and so on. And in the end, he tries to believe that the achieved result MUST sound good, because its the best measurments he was able to create.
And then we come to the test situations I described. People sit in front of their stereo rig and wonder why this sounds so sterile. Why it sounds so emotionless. Why it sounds so flat boring. So boring, that the best would be to switch it off and do some other things.
But wait, all measurements are excellent, this couldn't be the truth says the brain. It measures good, so it has to sound good. And then dispute starts between the brain and the emotional feelings. And that results in listener fatigue.
In the case of restaurant: customer calls the chef de cuisine and declares this dinner to smell and taste not first class, but the same as the industrial production canned beef that could be bought in warehouses for cheap. And the chef said, well, I cooked it in a manner to achieve optimum chemical results. And the customer replies: WTF interests me chemical results? I'm here to enjoy the best chateau briand in my life and you talk about chemical results?
And in the case of my audio rig: WTF is any measurement interesting for me when I try to maximize the taste of enjoument in audio? Its simply of no interest for me, because my ears hear and for that they have to be satisfied. And thats what I measure with the exact same method the chef de cuisine is applying in his kitchen: what is to be tasted has to be measured by taste. What is to be enjoyed by listening has to be tasted and refined by listening.
If something is wrong with the system (and no system is perfect), my ears will indicate it immediately, nearly after some seconds of listening. And thats, because I have a model of sound stored in the brain which is the perfect model for me.
Question:
The following situation: A chef de cuisine tries to fully satisfy the taste of his customers with a premium dinner (fife star quality). What does he do to achieve best results? Does he have a chemical laboratory in the kitchen, trying to analyse the right mix of ingredients at the right temperature to make for a perfect taste? Nonsense. He goes to the market early to buy the finest ingredients. He cooks the dinner and often tastes it, trying to tweak the result to what he had stored as the best taste he has ever known.
Thats the art of doing it. But if someone has no clue of what to achieve and how to achieve, if he had never tasted something real good, how could he try to achieve? Correct, he would go after measurements. Measurements and measuremts, because there is nothing more he knows.
To achieve a flat response, to achieve low THD, to achieve a wide bandwith, to achieve a short impulse answer and so on. And in the end, he tries to believe that the achieved result MUST sound good, because its the best measurments he was able to create.
And then we come to the test situations I described. People sit in front of their stereo rig and wonder why this sounds so sterile. Why it sounds so emotionless. Why it sounds so flat boring. So boring, that the best would be to switch it off and do some other things.
But wait, all measurements are excellent, this couldn't be the truth says the brain. It measures good, so it has to sound good. And then dispute starts between the brain and the emotional feelings. And that results in listener fatigue.
In the case of restaurant: customer calls the chef de cuisine and declares this dinner to smell and taste not first class, but the same as the industrial production canned beef that could be bought in warehouses for cheap. And the chef said, well, I cooked it in a manner to achieve optimum chemical results. And the customer replies: WTF interests me chemical results? I'm here to enjoy the best chateau briand in my life and you talk about chemical results?
And in the case of my audio rig: WTF is any measurement interesting for me when I try to maximize the taste of enjoument in audio? Its simply of no interest for me, because my ears hear and for that they have to be satisfied. And thats what I measure with the exact same method the chef de cuisine is applying in his kitchen: what is to be tasted has to be measured by taste. What is to be enjoyed by listening has to be tasted and refined by listening.
Last edited:
In my case a resistor load only. I have tested with a loudspeaker load, and in all cases I tested THD is worse than with a resistive load.
I now know if x amplifier is quoted as having y % THD, that into a real load the amplifier is much more likely to have 2y% THD.
I don't measure IMD because I do not have the testing equipment.
Nothing wrong with that but it is no dogma as there are many engineering challenges. I recently asked an enthusiastic builder for numbers on his se AD1 amp which tried to power a 104dB PA rig. That was against religious believes because all quality had to come from the less possible overpriced parts. When I suggested my 10W class A UL EL34 PP would be a better fit (power wise) he argued that it must suffer from 3H and cross over distortion. Blissfully ignorant he had no fear of IM distortion. Challenges, challenges 😉 Then again there is much fun to be had getting the most out of a single tube and a single speaker when you accept limitations.
Last edited:
I started making calibrated measurements (DC current) in 1959, just a very simple start.
I have been involved in measurements of electronics ever since.
For my lowest distortion signal source, I have a Denon Test CD.
It can do single tone, and has two different IM test tone pairs (250Hz/8000Hz 4:1) to look at 2nd order IMD; and 11k/12k 1:1 to look at 2nd and 3rd order IMD.
It also has an impulse to do a quick look at the frequency response to 22.05kHz.
And, the very low repetition impulse can safely be used, to test the amplifier into a loudspeaker load (to take a look at the amplifier frequency response into the varying impedance of the loudspeaker).
Then there are spot frequency checks from 21.5Hz to 19,999Hz, to accurately check frequency response and harmonic distortion.
My first look at an amplifier after checking DC voltages and currents,
is to connect a non-inductive resistor, and use a 1kHz tone, looking on the scope display.
Only then do I move on to the FFT to look at 2nd and 3rd harmonic distortion.
I have actually built amplifiers that visually looked very bad on that simple 1kHz tone . . . and guess what, they did not sound good at all.
And some that visually looked OK, but they did not sound good either.
Then there were many that visually looked good, and sounded good.
So Yes, listening is required as the final test.
If it does not sound good, who cares what the measurements say.
Your loudspeakers and your ears are more sensitive than just a simple test into a non-inductive load resistor.
Just my experience and just my opinion.
I have been involved in measurements of electronics ever since.
For my lowest distortion signal source, I have a Denon Test CD.
It can do single tone, and has two different IM test tone pairs (250Hz/8000Hz 4:1) to look at 2nd order IMD; and 11k/12k 1:1 to look at 2nd and 3rd order IMD.
It also has an impulse to do a quick look at the frequency response to 22.05kHz.
And, the very low repetition impulse can safely be used, to test the amplifier into a loudspeaker load (to take a look at the amplifier frequency response into the varying impedance of the loudspeaker).
Then there are spot frequency checks from 21.5Hz to 19,999Hz, to accurately check frequency response and harmonic distortion.
My first look at an amplifier after checking DC voltages and currents,
is to connect a non-inductive resistor, and use a 1kHz tone, looking on the scope display.
Only then do I move on to the FFT to look at 2nd and 3rd harmonic distortion.
I have actually built amplifiers that visually looked very bad on that simple 1kHz tone . . . and guess what, they did not sound good at all.
And some that visually looked OK, but they did not sound good either.
Then there were many that visually looked good, and sounded good.
So Yes, listening is required as the final test.
If it does not sound good, who cares what the measurements say.
Your loudspeakers and your ears are more sensitive than just a simple test into a non-inductive load resistor.
Just my experience and just my opinion.
Last edited: