Fundamental to this question is to define what you mean by "sound quality". Its a very vague term. Volume is a quality of sound, FR is a quality of sound, background noise is a quality of sound... Generally when talking of "sound" you are looking at an entire system, transducers as well as electronics, and we should all know speakers microphones and headphones contribute a lot to a system, usually much more than the amplification.
People need to understand that the sound wave getting into and out of the system has multiple components and is of low wattage, the sum of the components can add up to more watts.
Those waves have a distinct phase.
Now your 0.01% thd amp could distort the phase of many components of the wave in relation to one another and remove some , distort others and add some other components when this given music is sent through the system, an unstable asymmetric wave of more than 5 fundamental with all their harmonics and the whole noise spectrum of the recorded event room/acoustic
But for a single signal it is going to show 0.01% thd at 50 watts let suppose
A well designed amp like the jhl, will only have 10 watts , however it will have preserved the signal and not affected by back emf, and xo, etc
Those waves have a distinct phase.
Now your 0.01% thd amp could distort the phase of many components of the wave in relation to one another and remove some , distort others and add some other components when this given music is sent through the system, an unstable asymmetric wave of more than 5 fundamental with all their harmonics and the whole noise spectrum of the recorded event room/acoustic
But for a single signal it is going to show 0.01% thd at 50 watts let suppose
A well designed amp like the jhl, will only have 10 watts , however it will have preserved the signal and not affected by back emf, and xo, etc
^ I hope that was intended as sarcasm because it isn't exactly true for all humans and for all types of music.
The above illustrates the confusion that can arise from using and trying to make sense out of spectral analysis.
It may help to know that especially at lower frequencies, say, maybe below 1kHz or 1.5kHz, the phase of harmonics is a factor. Phase is also a factor at lower frequencies in relation to linear distortions.
The usual presumption of spectral analysis is that there is an unchanging slightly curved transfer function to an amplifier. The shape of the curvature means that different signal levels may be distorted differently.
The fact that the presumption is an oversimplified model sometimes leads to further confusion. The curvature can change with frequency due to falling loop gain at higher frequencies. The curvature can also change dynamically due to thermal effects at lower frequencies.
Also, a DC servo can audibly affect phase at lower frequencies. In many cut and paste servo designs its often the case that audible linear phase distortion is present at LF. One problem is that a DC servo has to settle fast enough at turn-on, which may result in setting the corner frequency too high. There are ways to fix that but it adds to cost and complexity.
@Baffless - Edited b/c some posts snuck in before I noticed.
re: the salt - I agree, and moreover, I appreciate the added thoughts.
I suppose it's more of a thought experiment than a reality. However, it would be absolutely fantastic if a few people actually and without doubt knew that their brains were processing the information the same way => hearing the "same thing the same way". We can only guess at this point. That's why I often wonder why there can sometimes seem to be so much passion around this amp or that amp or equalization etc. We don't even know if once that perfectly delivered sound wave reaches the pinnae ... that it's even perceived the same from person A to person B.
re: the salt - I agree, and moreover, I appreciate the added thoughts.
I suppose it's more of a thought experiment than a reality. However, it would be absolutely fantastic if a few people actually and without doubt knew that their brains were processing the information the same way => hearing the "same thing the same way". We can only guess at this point. That's why I often wonder why there can sometimes seem to be so much passion around this amp or that amp or equalization etc. We don't even know if once that perfectly delivered sound wave reaches the pinnae ... that it's even perceived the same from person A to person B.
subjective evaluation is by definition useful when you will be the only user of equipment - otherwise not so much
i listened to some well reputed amps and they all sounded horrible to me while my QSC PLX sounds great to me despite not being "audiophile" amp
i don't care what the measurements are for any of the amps in question - i will use what sounds good to me and get rid of the rest
but just because it sounds good to me doesn't mean it will sound good to you and vice versa
we all want things to sound a certain way. the device ( amp, speaker - whatever ) either sounds that way or it doesn't. that's all that should matter to you when the device is for your private use.
if on other hand you're designing a commercial product you can't trust your ears anymore. you would need to use a combination of measurements and blind A/B testing using other people's ears.
of course if all the measurements and blind A/B testing check out but you can't stand the way it sounds then maybe somebody ELSE should sell this product and you should design something that is more uniquely YOURS that way you can build your brand around this sound, provided you can find enough other people who share a similar taste.
i really think all 3 are equally valuable:
1 - objective measurements
2 - blind A/B tests using other people
3 - your own subjective feelings
Robert Greene used "a rider and a horse" analogy to explain the synergy between rational and emotional in his books. He explained that you are doomed to fail if you lean too much on either side - instead you must find a way to rationally channel your passion. same here.
you must always analyze every problem from as many different perspectives as possible then decide what weight to give each one in overall assessment which will ultimately be based on a combination of all of them.
i listened to some well reputed amps and they all sounded horrible to me while my QSC PLX sounds great to me despite not being "audiophile" amp
i don't care what the measurements are for any of the amps in question - i will use what sounds good to me and get rid of the rest
but just because it sounds good to me doesn't mean it will sound good to you and vice versa
we all want things to sound a certain way. the device ( amp, speaker - whatever ) either sounds that way or it doesn't. that's all that should matter to you when the device is for your private use.
if on other hand you're designing a commercial product you can't trust your ears anymore. you would need to use a combination of measurements and blind A/B testing using other people's ears.
of course if all the measurements and blind A/B testing check out but you can't stand the way it sounds then maybe somebody ELSE should sell this product and you should design something that is more uniquely YOURS that way you can build your brand around this sound, provided you can find enough other people who share a similar taste.
i really think all 3 are equally valuable:
1 - objective measurements
2 - blind A/B tests using other people
3 - your own subjective feelings
Robert Greene used "a rider and a horse" analogy to explain the synergy between rational and emotional in his books. He explained that you are doomed to fail if you lean too much on either side - instead you must find a way to rationally channel your passion. same here.
you must always analyze every problem from as many different perspectives as possible then decide what weight to give each one in overall assessment which will ultimately be based on a combination of all of them.
Couple of points in response to Hugh’s original question:
Lets not forget that the ultimate purpose of amplifiers in our hobby is to provide listening pleasure. So be sure to listen to them and see if you enjoy the music. Do you forget about the amplifier and get into the music? That’s a good amplifier. Really good amps make it easy to feel the music and to enjoy it. Lastly, if tube amps didn’t sound good they would not be in the market today.
Lets not forget that the ultimate purpose of amplifiers in our hobby is to provide listening pleasure. So be sure to listen to them and see if you enjoy the music. Do you forget about the amplifier and get into the music? That’s a good amplifier. Really good amps make it easy to feel the music and to enjoy it. Lastly, if tube amps didn’t sound good they would not be in the market today.
Actually the frequency gain increases in Ss amps, the gain at lower freq is compensated by feedback , it has a RIIA curve
I don’t know what ‘frequency gain’ is, or why you think feedback only happens at lower frequencies, but surely you mean RIAA, not RIIA, and that only occurs on phono inputs, and for all of them, not just SS ones.
What?
You seem to actually care about some of this, but I haven't the foggiest of clues how you'd come to that conclusion / make that statement. I don't understand the term "frequency gain".
Maybe break things into smaller pieces. You have a clear idea if you are using the terms feedback and (sic) RIAA.
Let's start with - what exactly do you mean by frequency gain? Perhaps show a graph / chart / hand-drawn scribble.
I have amplifiers with zero feedback that have ruler straight gain across the audible frequency range (as one example). So, no, that's not true for all SS amplifiers.
Phono devices will have an RIAA circuit, yes. That may be perceived to "boost" the low frequencies. However, the only purpose is to bring the intentionally altered equalization 'back to flat'. This is solely because of how the records are manufactured. You can have a tube phono pre-amp or an SS phono pre-amp or even 'hybrids'. They can be built with or without feedback. Depending on the circuit, if the gain is altered, the feedback may be altered (or vice versa). Regardless, they should (let's not debate this) have the same desired RIAA compensation / curve.
Help us understand what you actually mean using different words. I am trying to understand your meaning vs. what you type, but I'm not getting very far. It's particularly interesting when you begin with words like, "actually", which I typically take to mean that you are attempting to correct someone.
Last edited:
For what it's worth, my latest amp sims are SRPP types -- it occurred to me that the structure fixes a problem with amplitude modulation distortion that I was finding with push-pull. By design, P-P cancels even harmonics, that much is pretty clear and everyone usually thinks nothing of it. BUT, this also means there's a tendency for perfect 'nulls' to occur at specific output voltage levels, where cancellation is more perfect than at other voltages.
As the audio signal traverses those voltages, the "benign harmonics" turn into little AM monsters. To calculate the dynamic range of a signal, simply divide the maximum amplitude by the minimum, right? So what do you get if you divide by zero?
The SRPP is not guaranteed to fix this, but what I'm doing is using the N-side as the controlling side, and the P-side is bootstrapped. So if N distorts slightly, P's distortion should only accumulate, without "moire patterns" of destructive interference. As long as P's gain isn't set too high. I've managed to get H2 to cancel quite nicely in simulation, but I don't want to overcook it.
For once, the first few harmonics, 2, 3, 4... seem to follow a steady pattern, whether testing at 400mV out (peak), 4V or 16V, or with various DC offsets. Something that was impossible with push-pull.
Every console I installed since the early 2000s was digital. Content either file based, across an AoIP network or converted to digital at the input. Number of NE5532 in the signal path typically zero.
All that matters in audio production is the number of ICs between the console input and output. In my plants analogue consoles were primarily used as glorified input routers and monitor controllers. Mixing was done in the digital workstation for the last 20+ years. No producer routes individual signals through every input strip, insert loop, sub group and side chain of an analogue console to reach 'hundreds or thousands' of ICs in the path unless they'e aiming at an effect. The overwhelming majority of those 'thousands' idle in unused side paths.
It's time to let this audio myth go.
Sorry I was typing from my cell phone, RIAA.
This is open loop gain of many SS amps, local feedback/combination of multiple devices restores it in input stages.
John Linsley Hood addressed that in his amplifier.
It would be more pleasant if people read the entire text before jumping with a knife and cutting parts and jumping to conclusions such as: you don't understand that.
This is open loop gain of many SS amps, local feedback/combination of multiple devices restores it in input stages.
John Linsley Hood addressed that in his amplifier.
It would be more pleasant if people read the entire text before jumping with a knife and cutting parts and jumping to conclusions such as: you don't understand that.
^ So what did you mean by this statement?
I read it all, tried to understand it, and it makes no sense to me as a whole unless you are referring only to a phono pre-amp.
However, what you stated is not true for SS amps vs. Tube amps (or other) and it may or may not have anything at all to do with feedback.
I am deliberately trying to be pleasant and understand you vs. simply stating that you are "wrong". I'd prefer to understand why you make the statements you are making. You 'learned' it somewhere. Perhaps post a citation that points to where you learned it, so we can all benefit.
I read it all, tried to understand it, and it makes no sense to me as a whole unless you are referring only to a phono pre-amp.
However, what you stated is not true for SS amps vs. Tube amps (or other) and it may or may not have anything at all to do with feedback.
I am deliberately trying to be pleasant and understand you vs. simply stating that you are "wrong". I'd prefer to understand why you make the statements you are making. You 'learned' it somewhere. Perhaps post a citation that points to where you learned it, so we can all benefit.
What are ‘this’, ‘it’, and ‘that’ in these sentences?
You are not making yourself clear: and whinging about your reception does not address that.
You are also not answering the questions you’re being asked here.
Last edited:
Noise from the powersupply which may not be audiable is still there . When one cleans up a suppossidly quiet ps, there effect is a more relaxed feel for the music. And music is emotion, not intellect. We feel the music, we dont think it. Hugh a thumbs up for your 39w se amp.love it, listen every day,beautifull.its a winner.
Agreed - I love mine, too! (In fact, I have two of them, as I need one stereo amp per channel, for my 2-way active spkrs. 🙂 )
But I can't use them over the summer months - simply too much heat output, with 2.5a bias (I built the 4ohm version.). You must live in a colder place than Melbourne. 🙄
harmonic distortion analyzer are valuable in the design of audio circuits Period : .
What I mean is that the answer to your first question is No. The second, yes, with more measurements such as impedance, stability under all frequency and Xo/speakers load and wires, phase, and feedback accuracy whenever it is used so that it doesn't introduce artefacts or/and remove information.
The power supply too is much importance and channel separation.
SE amps have a lot in their favor : can drive asymmetrical complex wave form without much suffering from the power supply, few parts, few added distortion, most harmonics at low wattage is second order.
What I mean is that the answer to your first question is No. The second, yes, with more measurements such as impedance, stability under all frequency and Xo/speakers load and wires, phase, and feedback accuracy whenever it is used so that it doesn't introduce artefacts or/and remove information.
The power supply too is much importance and channel separation.
SE amps have a lot in their favor : can drive asymmetrical complex wave form without much suffering from the power supply, few parts, few added distortion, most harmonics at low wattage is second order.
It's the hammer strike point (of just about any piano, not just grands), 6/7ths the way from one end to the other of the string's speaking length, that maximally reduces generation of the 7th harmonic. The 5th harmonic is (approximately) two octaves and a musical major 3rd up from the fundamental, which isn't nearly as dissonant as the 7th. The highest notes have a hammer strike point even closer to the string endpoint, but up there the 7th harmonic approaches or exceeds the maximum frequency of human hearing.
i have a grand piano 5.8 and I use microphones to tune it sometimes, the fundamental on many spiral warped notes are much lower than the harmonics, I mean the frequency which should be the loudest isn't... (maybe i am exaggerating, but I swear on the initial hit it isn't)
huge THD in a piano!!!
then I listened carefully, and it is true!, the hit is some weird noise, then it oscillate between two harmonics, but the decay is based on the fundamental, however it is not the loudest sound.
huge THD in a piano!!!
then I listened carefully, and it is true!, the hit is some weird noise, then it oscillate between two harmonics, but the decay is based on the fundamental, however it is not the loudest sound.