Hi Sandrohv, Minek,
Good question.
If we are interested in sound quality then there is another dimesnion.
It requires a bit of detective work to unravel decades of (faulty) ideas.
1) Most power amplifier designers assume that lower distortion is always better period. THD is usually used as the measure; the lower THD the better the sound - it is assumed. But that is nonsense. Massa in 1930 said THD needs to be dropped as a measure of sound quality. So today we have this disconnect between measuring distortion and how good an amplifier sounds.
2) Another dimension is when amplifiers clip, and they do clip more than we like to admit (esp. with low efficiency speakers). Global feedback makes it harder to recover from a clip event. The majority of amps I have seen do not recover cleanly, creating subtle audio artifacts that can dominate the "amplifiers sound". Bob Cordell often uses the phrase "Amplifiers sound differently because they misbehave differently" -- like when clipped or when driving some (unterminated) speaker cables.
3) I mention the Halcro 1994 amplifier in this post Non-switching complimentary output stage as an example of a ultra low distortion amplifier not getting a good review in Stereophile (not a bad review either). The reviewers just liked tube amps, they "sound better to them". Most here are aware that tube amps don't give ultra-low distortion figures -- they can't make ultra-low distortion (standard) tube amps because of the transformers limited bandwidth. It's a useful example of global feedback (or lack of it) influencing the sound of a power amplifier.
Last edited:
Hi Ian,
I guess this reply triggers another question: if we don't use distortion measurement to evaluate/compare amplifiers, what else should we use?
I would still prefer something we can measure and express with numbers, rather than subjective 'opinions'.
I guess this reply triggers another question: if we don't use distortion measurement to evaluate/compare amplifiers, what else should we use?
I would still prefer something we can measure and express with numbers, rather than subjective 'opinions'.
Hi Mikek,
We can use THD as a starting point. (Usually THD is the start and the end point).
Then structure of harmonics using a sinewave is the key to how it sounds. Our hearing has greater sensitivty to higher harmonics when the harmonics are below about 3kHz. The plot below shows tests of average hearing for a 50Hz tone (Ref: Fielder, JAES, Jul 1987, p517-534) and used in Linear Audio Vol.4
The easiest way that I have found to apply this to an amplifier is to take an average slope of 50dB per decade for our hearing. Using a 50Hz signal plot the FFT and check the average harmonic roll-off slope (up to about 3kHz). If the average slope is -50dB per decade or greater then we know the higher harmonics are not going to dominate over the lower harmonics, in which case we only need consider whether we can hear the lower harmonics (2nd or 3rd) or not.
The level for 2nd harmonic to be audible is about 1% (-40dB) and 0.1% for 3rd harmonic (-50dB). That's for the average listener; if your hearing is above average you may add another 10dB. Then add a safety margin of 10dB.
For example, here Non-switching complimentary output stage (plot copied below):
THD here is 0.04% (it's at 20kHz but for argument sake let's assume it is done at 50Hz). It is mostly 2nd harmonic.
So we can confidently say this distortion will be inaudible because it is less than 1% 2nd (and even below 0.1% - that's if you have exceptional hearing).
Also, since in this example it is the output stage distortion, we can confidently say it will sound fine without any global feedback; in fact we can confidently say global feedback won't improve the sound quality of this output stage.
But what if the slope is less than -50dB per decade? Then we may need to add more negative feedback particularly in the 1-10kHz region so higher harmonics are made inaudible (even if the low order harmonics are already inaudible). In this case it's not so clear cut how much extra feedback is needed.
Or have I missed something?
Well don't forget to make the amplifier recover from clipping cleanly, or better still make it soft clip. And all other instabilities are taken care of, like from cables eg a cable termination Zobel at the speaker terminals (see here Bob Cordell's Power amplifier book, and Bob Cordell's Power amplifier book).
This method is dead easy to use and it adds enough certainty to our distortion measuments. Of course we still need to listen as a reality check to build our confidence and trust in this method.
Another example of using this approach is for assessing the known effect of source degeneration resistors in Class-AB MOSFET amps on sound quality, eg here Square Law Amp.
We can use THD as a starting point. (Usually THD is the start and the end point).
Then structure of harmonics using a sinewave is the key to how it sounds. Our hearing has greater sensitivty to higher harmonics when the harmonics are below about 3kHz. The plot below shows tests of average hearing for a 50Hz tone (Ref: Fielder, JAES, Jul 1987, p517-534) and used in Linear Audio Vol.4
The easiest way that I have found to apply this to an amplifier is to take an average slope of 50dB per decade for our hearing. Using a 50Hz signal plot the FFT and check the average harmonic roll-off slope (up to about 3kHz). If the average slope is -50dB per decade or greater then we know the higher harmonics are not going to dominate over the lower harmonics, in which case we only need consider whether we can hear the lower harmonics (2nd or 3rd) or not.
The level for 2nd harmonic to be audible is about 1% (-40dB) and 0.1% for 3rd harmonic (-50dB). That's for the average listener; if your hearing is above average you may add another 10dB. Then add a safety margin of 10dB.
For example, here Non-switching complimentary output stage (plot copied below):
THD here is 0.04% (it's at 20kHz but for argument sake let's assume it is done at 50Hz). It is mostly 2nd harmonic.
So we can confidently say this distortion will be inaudible because it is less than 1% 2nd (and even below 0.1% - that's if you have exceptional hearing).
Also, since in this example it is the output stage distortion, we can confidently say it will sound fine without any global feedback; in fact we can confidently say global feedback won't improve the sound quality of this output stage.
But what if the slope is less than -50dB per decade? Then we may need to add more negative feedback particularly in the 1-10kHz region so higher harmonics are made inaudible (even if the low order harmonics are already inaudible). In this case it's not so clear cut how much extra feedback is needed.
Or have I missed something?
Well don't forget to make the amplifier recover from clipping cleanly, or better still make it soft clip. And all other instabilities are taken care of, like from cables eg a cable termination Zobel at the speaker terminals (see here Bob Cordell's Power amplifier book, and Bob Cordell's Power amplifier book).
This method is dead easy to use and it adds enough certainty to our distortion measuments. Of course we still need to listen as a reality check to build our confidence and trust in this method.
Another example of using this approach is for assessing the known effect of source degeneration resistors in Class-AB MOSFET amps on sound quality, eg here Square Law Amp.
Attachments
Last edited:
Hi Ian,
I actually fundamentally disagree with you. For me, a High Fidelity amplifier should not have sound. As you correctly point out, given that "The structure of harmonics using a sinewave is the key to how [an amplifier] sounds.", an amplifier should have no harmonics.
Mathematically, for me, if I drive an amplifier with V_I(t), then within the audio range, the ideal output ought to be:
V_OUT(t) = A_V * V_I(t) where A_V is the voltage gain of the amplifier.
Now, given no amplifier on Earth does this, and all amplifiers do
V_OUT(t) = A_V * V_I(t) + harmonics + noise
the latter two should be minimized.
The problem as I see it, is that what sounds 'good' for you may sound terrible to me and viceversa. Hence, I think 'sound free' amplifiers are better. Also, they allow the user to insert any distortion that he pleases to the line level signal before amplification and do not alter this distortion (nor the signal). I personally would not do that, since I prefer to listen to the source the way the artist recorded it, but that is personal preference. As a result, if you want to insert harmonics that follow the -50dB scheme, I believe it should be done before amplification by whatever method is best, e.g. could be by square law pre-amp for example.
Having said this, I strongly advocate for Global negative feedback since it is the best way to minimize harmonics. Sadly, NFB does not fix noise.
Now, I would consider amplifiers that have a dialed sound, i.e. that produce a certain harmonic structure, specialty amps. The best example are guitar amps, where guitar players rely on the distortion added by the amplifier. Do they have a place outside guitar amps? Sure, but again to a specialty audience who seek for amplifiers who have certain sound characteristic.
Finally, regarding clipping behavior, I find very few amplifier designs have circuitry to improve recovery and hence why the ugly behavior. Bob's BC-1 has a good attempt which deserves study.
Anyway, I am sure you fundamentally disagree with me, but hey this is Audio, we live to disagree 🙂.
- Sandro
P.S. BTW, I believe it is Minek, not Mikek... the latter is our dear friend Michael Kiwanuka.
I actually fundamentally disagree with you. For me, a High Fidelity amplifier should not have sound. As you correctly point out, given that "The structure of harmonics using a sinewave is the key to how [an amplifier] sounds.", an amplifier should have no harmonics.
Mathematically, for me, if I drive an amplifier with V_I(t), then within the audio range, the ideal output ought to be:
V_OUT(t) = A_V * V_I(t) where A_V is the voltage gain of the amplifier.
Now, given no amplifier on Earth does this, and all amplifiers do
V_OUT(t) = A_V * V_I(t) + harmonics + noise
the latter two should be minimized.
The problem as I see it, is that what sounds 'good' for you may sound terrible to me and viceversa. Hence, I think 'sound free' amplifiers are better. Also, they allow the user to insert any distortion that he pleases to the line level signal before amplification and do not alter this distortion (nor the signal). I personally would not do that, since I prefer to listen to the source the way the artist recorded it, but that is personal preference. As a result, if you want to insert harmonics that follow the -50dB scheme, I believe it should be done before amplification by whatever method is best, e.g. could be by square law pre-amp for example.
Having said this, I strongly advocate for Global negative feedback since it is the best way to minimize harmonics. Sadly, NFB does not fix noise.
Now, I would consider amplifiers that have a dialed sound, i.e. that produce a certain harmonic structure, specialty amps. The best example are guitar amps, where guitar players rely on the distortion added by the amplifier. Do they have a place outside guitar amps? Sure, but again to a specialty audience who seek for amplifiers who have certain sound characteristic.
Finally, regarding clipping behavior, I find very few amplifier designs have circuitry to improve recovery and hence why the ugly behavior. Bob's BC-1 has a good attempt which deserves study.
Anyway, I am sure you fundamentally disagree with me, but hey this is Audio, we live to disagree 🙂.
- Sandro
P.S. BTW, I believe it is Minek, not Mikek... the latter is our dear friend Michael Kiwanuka.
Last edited:
Minek, My sincere apology! It was a typo.
Hi Sandrov,
Seems you have misunderstood what my aim in amplifier design is.
I aim for harmonic levels below audibility (when not clipping). Read my full post to see this.
I assume you understand set theory. If you include the null set (no amplifier 'sound') to how an amplifier 'sounds' then you can understand where I am coming from.
In layman's terms the 'sound' I am aiming for is no added colorations.
But is that realistic? Amplifier designers are forced to consider the inconvenient truth that there are times when an amplifier clips - because our amplifiers do clip when we are listening to them - it happens more than we want to admit. It's partly due to the low efficiency speakers we have since the 1970's when they were made small (strangely, by an Australian by the name of Richard Small).
To minimize this unwanted effect we can use higher efficiency speakers and/or a higher power amp, and/or add soft clipping to the power amp.
The soft clip option is usually rejected by designers aiming for ultra-low distortion because it adds distortion even at 10% of full power, so you no longer get ultra-low distortion up to full power😱.
But in practice with soft clipping you can have clipping more often before you hear clipping. So it is helpful if you don't want to change your loudspeakers OR you don't want a monster amp.
I believe if the soft clip is done right it doesn't add 'sound' to a 'no-sound' amplifier. I haven't done controlled trials to prove this - it's just an 'educated guess' based on some very scant data. Although I am convinced enough from my bench tests to design most of my power amps with soft clip permanently build in as part of the input stage, eg, both my Cube-law amp and CSD amp in Linear Audio (Vol.8 and Vol.13). See my sign links for more on these amps and/or read the Linear Audio articles.
I'll prefer to use 2-factor estimation.
1. Feedback depth at highest audio freq, say 20 kHz for simplicity, combined with unity-gain freq. Now there are no point to design an amp with less than 100 dB at 20k, which paired with 3EF OPS will have straightly unmeasurable distortion. Yes, THD can be measured using compensation methods aka distortion magnifier, but there out of diy interest at those FB depths.
2. Less than 10% step response overshoot with less than 1-2 period oscillation.
Noise level will be defined by input resistanse and we should use inverting topology for cancelling common-mode error which are nonlinear with common mode itself.
That's enough now and accumulates most of the available knowledge.
But...
They better, but they are unsaleable.
Such an amp will sound too clearly to show all the recording/mastering/speaker flattening/ground routing mistakes. First we need to completely refuse unsymmetrical connection and partly noninverting amps.
Also they sound boring.
Distorting amplifier can sound pretty or terribly for different listeners, but they can be combined with speakers distortion and provide acceptable to one results.
Now it makes no sense in developing linear amp. It's better to investigate in GaN and self-oscillating topologies. But not a diy unfortunately.
As far as audio is concerned, this tends to come down to one answer - many audiophiles hear it as sounding better. The most common response to such subjective reports is, the improvement you feel you hear is not scientifically validated, so, is inherently suspect. Which is true, but suspect isn't the same thing as false.
There are a relatively small number of audiophiles who who either strongly suspect, or have concluded, that it's the measurable performance imperfections, the errors, of no global-feedback amplification which is the objective cause of sound perceived as more realistic. To be clear, I'm not referring to pleasant euphonic coloration. I'm referring to a sound which sounds more like live music to human ears. This is the camp in which I've unexpectedly come to find myself.
Once one comes to the above view, many subjective listening reports become rationally explainable which otherwise have not been. Such as the common love for the sound of tube gear, and the preference by many for the sound of non-global-feedback solid-state gear. It seems that the dynamic behavior of certain technical errors are serendipitous utilized by the ear to perceive reproduced music as sounding more like live music. Such as, harmonic distortion which rises and falls smoothly in proportion to signal level.
Conversely, interestingly, this suggests that the more perfect the objective performance of an amplifier, the LESS like live music it will sound! Among the most objectively perfect components is an amplifier utilizing an high amount of global-feedback. Even, perhaps, this is behind the not un-common dissatisfaction/disappointment with sound of objectively perfect digital audio. I don't understand well enough all of the psycho-acoustics at work, it just seems deductively logical to suspect that this is what may be happening.
Hi Ian, regarding soft-clipping, where do you place the start of soft-clipping? I.e. at 80%, 90% of full power.
While I cannot say I am advocate, it just has been the way I have done things, I have always implemented hard clipping with fast recovery. This is the op-amp way. As a result, I would be interested in knowing more on how soft-clipping is implemented.
While I cannot say I am advocate, it just has been the way I have done things, I have always implemented hard clipping with fast recovery. This is the op-amp way. As a result, I would be interested in knowing more on how soft-clipping is implemented.
Hi Ken, it a nutshell what you are saying is that you prefer imperfect amps because they sound better, more like live music. That is your preference and it is perfectly fine.
I belong in a different camp, the camp of making the amp as transparent as possible such that the input signal is modified the least, and if I wanted the input signal to sound more like live music, I would add the effect to the line level signal with an effect box.
I belong in a different camp, the camp of making the amp as transparent as possible such that the input signal is modified the least, and if I wanted the input signal to sound more like live music, I would add the effect to the line level signal with an effect box.
Hi Ian, the take away I get from your post is that there is no need to strive for 0% distortion but for inaudible distortion. The inaudible levels are (from your post):
"The level for 2nd harmonic to be audible is about 1% (-40dB) and 0.1% for 3rd harmonic (-50dB). That's for the average listener; if your hearing is above average you may add another 10dB. Then add a safety margin of 10dB."
Also, given that sensitivity to harmonics increases with frequencies at a slope of ~50dB/dec on average (from your picture and explanation), harmonics ought to decrease at a slope of -50dB/dec or better.
Am I correct with this summary?
Last edited:
Hi, Sandro,
My intention was only to provide an answer to @minek123's question regarding what is the advantage of non-global-feedback. There really is no large objective advantage. So, the advantage to some ears, is an subjective one. I certainly was not intending to suggest that everyone should agree with what any other person hears/prefers.
However, I would like to offer the following philosophical question of extreme opposites. Which of the following would you choose as your home music reproduction system?
1) One that technically is objectively perfect, having no electrical errors detectable by the most sensitive instruments, yet leaves you bored and joyless with any music it reproduces.
-OR-
2) One with easily measurable electrical errors, yet sounds subjectively perfect, meaning, is subjectively indistinguishable to the ears from a live performance.
What I'm asking in the above question is, what do you want most from your music system, intellectual satisfaction, or emotional satisfaction, if you can only have one or the other not both.
Yes I know its a bit confusing, I couldn't find a better way of showing the NIC topology as the feedback resistors would cross over the power rails for the opamp without some snaking about somewhere.
The opamp swings at a fraction of the output swing, about 3/4 of it, quite deliberately to ensure the opamp inputs and outputs are roughly symmetric within the floating supply (the inputs are about 50% swing, the output 100%, so the average is 75%). The zener in across the supply as that protects the opamp from overvoltage.
Hi Ken, my preference is 1.
What I don't like about your phrasing though is that you have added opinion which makes your philosophical question less impartial. In 1 you say: "yet leaves you bored and joyless with any music it reproduces". That is your opinion because what gives me joy could be listening a transparent and accurate reproduction of the input with no errors introduced... maybe for me that is subjectively perfect.
Anyway, this is the why for my choice. 2 reasons:
1. There is no consistency on what subjectively perfect is... hence why it is subjective. Good for you could be really bad for me. Hence what may give you hearing joy, may give me hearing displeasure and viceversa. Therefore, my preference is for an 'impartial amplifier' that adds no coloration. This brings me to reason 2,
2. I can inject sound to give me 'hearing pleasure' by means of an effect box prior to the amp. I.e. I could potentially make the input signal sound like live music by processing the input signal by a multitude of methods.
What I don't like about your phrasing though is that you have added opinion which makes your philosophical question less impartial. In 1 you say: "yet leaves you bored and joyless with any music it reproduces". That is your opinion because what gives me joy could be listening a transparent and accurate reproduction of the input with no errors introduced... maybe for me that is subjectively perfect.
Anyway, this is the why for my choice. 2 reasons:
1. There is no consistency on what subjectively perfect is... hence why it is subjective. Good for you could be really bad for me. Hence what may give you hearing joy, may give me hearing displeasure and viceversa. Therefore, my preference is for an 'impartial amplifier' that adds no coloration. This brings me to reason 2,
2. I can inject sound to give me 'hearing pleasure' by means of an effect box prior to the amp. I.e. I could potentially make the input signal sound like live music by processing the input signal by a multitude of methods.
There is no "better sound", there is only "more accurate" (*). The aim of an amp is to pass the signal unmolested, this is obvious to common sense. Sound accuracy can only be reduced by an amp, never increased, but in practice the failings of a good amp is beyond human perception, and takes measurement to characterize, so long as its working correctly and within its limits. Worry about your room-acoustics and speakers once your amp is below 0.1%, and if your amp is 0.01% or less you're going to be fine.
Yes there are more comprehensive ways to measure amp accuracy, but THD/IMD are basic litmus tests for fundamental performance that are a good starting point.
If its clipping, that's a problem, period. You'll need more voltage headroom to solve that, not throw away the benefits of feedback. [is this not entirely obvious?].
The difference between soft clipping and hard clipping is that soft clipping distorts at lower signal levels - making less efficient use of the rail voltages.
And when we look at the kinds of signals that have a high crest factor and are more likely to clip in a given system, we see that its the extreme transients (rimshots, gunshots, 1812 overture, anvil chorous, that kind of thing). The human ear also clips such sounds so there's little to gain in rendering them better if they are at very high volume compared to average programme level - in fact you might want to limit such extremes to protect hearing.
For big PA systems clipping is unavoidable sometimes, and there soft-clipping/limiting is routinely used but its not an ideal situation. But its done with a limiting circuit designed for the purpose (not a side-effect of circuit weaknesses).
Don't see a point being made here. Stereophile is not a publication I'd trust in any way, they are not about truth.
(*) A truly horrendous sound, if amplified accurately sounds truly horrendous, in this situation an amp that made it sound better would be heavily distorting. Think nails scraped down a blackboard. How would those stabbing notes in Hitchcock's "Psycho" work if they were all mellow and warm??
This is all well and good, but 2nd harmonic distortion goes hand-in-hand with very objectionable 2nd-order intermodulation products, which are definitely not welcome - we rarely listen to single tones and intermodulation between simultaneous tones can fall far outside the perceptually-masked part of the spectrum.
(In general intermodulation products are all over the place, so there's no clear sensitivity slope for them, they are perhaps all equally objectionable)
Fortunately most symmetric amp designs push even harmonics well down, so that 3rd harmonic and 3rd-order IM products tend to dominate in terms of power, and probably dominate IM product generation. This is also true for microphones and speakers of course, they may be the limiting factor in practice.
Well clearly that's not possible is it? Unless it magically makes those in hearing range clinically depressed!
This second choice is entirely possible as electronic test gear far surpasses human perception in detecting non-ideal performance. No reproduced music is "subjectively indistiguishable" from a live performance because our other senses are very important to subjective perception of sound. Add video for much more satisfactory experience, nothing to do with the audio reproduction though.
You can't invoke a counter-factual situation to argue a point of fact, and expect anything but short-shrift.
You offer choice between two imagined things that assume you point of view, Hobson's choice basically.
Its better to go and do some measurements, including perceptual ones with double-blind safeguards against expectation bias, and figure out reality from experiment.
Hi, Sandro, first, thanks for entertaining my question. Not everyone would.
The question is intentionally posed as an extreme of opposites. This is so, to help flush out the philosophy underlying the system choices of person answering the question. I'm unsure what you refer to when you write that I have added 'opinion' to the question. The question only refers to what gives you joy, not me or anyone else. This question is only about your perception of the music you experience with your own ears, via two completely opposite hypothetical systems. No human appreciates music via spectrum analyzer.
Again, the question has nothing to do with anyone's hearing except your own. There is no, my perception versus your perception, or my likes versus your likes. It only assumes only that, for whatever reason, YOU experience only boredom with no musical joy from the fist system, and only experience what sounds, to your ears, like a live performance from whatever music is played via the second system.
Perhaps, you can, perhaps you can't. Either way, it has nothing to do with what the philosophical question is constructed to illuminate. All my question is seeking to determine is whether you most value emotional musical satisfaction, or intellectual parameter satisfaction. Do not assume that one automatically confers the other.
Hi Ken, I understood your question. What I meant with adding your opinion, was inserting the comment "yet leaves you bored and joyless with any music it reproduces". Like Mark said, this comment is pretty ridiculous. If you are getting bored or depressed, the music choice is most likely the culprit.
So let me rephrase your options:
1. One that technically is objectively perfect, having no electrical errors detectable by the most sensitive instruments
2. One with easily measurable electrical errors, yet sounds subjectively perfect
While option 2 tentatively sounds more appealing, the impossibility and inflexibility of such amp voids it. What I mean by this is that there are multiple factors that are part of "subjectively perfect" sound: the music content, my mood of the day, the activity I am doing while listening to music, etc. Given it is subjective, everything is fair game and does not need to be the same at every moment in time.
Therefore, I would prefer an option 1 amplifier that is perfect, and then add sound effects to the input signal to make the output sound the way I want depending on what I feel "subjectively perfect" is at that moment in time. These effects could be to correct for room acoustics, speaker deficiencies, make it sound more like live music, pump the base if I am working out, etc.
So again, option 1 it is.
"All my question is seeking to determine is whether you most value emotional musical satisfaction, or intellectual parameter satisfaction." You could have just have asked directly.
- As a circuit designer and engineer, definitely the latter.
- As a music listener, the former, but in my opinion, I prefer to achieve it via input signal sound processing rather than mangling the amplifier.
So let me rephrase your options:
1. One that technically is objectively perfect, having no electrical errors detectable by the most sensitive instruments
2. One with easily measurable electrical errors, yet sounds subjectively perfect
While option 2 tentatively sounds more appealing, the impossibility and inflexibility of such amp voids it. What I mean by this is that there are multiple factors that are part of "subjectively perfect" sound: the music content, my mood of the day, the activity I am doing while listening to music, etc. Given it is subjective, everything is fair game and does not need to be the same at every moment in time.
Therefore, I would prefer an option 1 amplifier that is perfect, and then add sound effects to the input signal to make the output sound the way I want depending on what I feel "subjectively perfect" is at that moment in time. These effects could be to correct for room acoustics, speaker deficiencies, make it sound more like live music, pump the base if I am working out, etc.
So again, option 1 it is.
"All my question is seeking to determine is whether you most value emotional musical satisfaction, or intellectual parameter satisfaction." You could have just have asked directly.
- As a circuit designer and engineer, definitely the latter.
- As a music listener, the former, but in my opinion, I prefer to achieve it via input signal sound processing rather than mangling the amplifier.
Last edited:
Sandro, I suspect that we are about to go in circles. So, rather than respond to the above, I'll let you have the last word.
Cheers.
Last edited:
Thanks... I guess??? 😕
I think at the end of the day is personal preference. Perfect amp vs. Imperfect amp that sounds good to the owner.
If you can find the imperfect amp whose imperfections sound amazingly good to you, well then it is the perfect amp for you.
I don't think I'll ever find such imperfect amp for me, hence I prefer the perfect amp, and then add those imperfections by signal processing.
I know it sounds like a bit of mouthful, but that is my takeaway of our short discussion. Not sure if you agree.
I think at the end of the day is personal preference. Perfect amp vs. Imperfect amp that sounds good to the owner.
If you can find the imperfect amp whose imperfections sound amazingly good to you, well then it is the perfect amp for you.
I don't think I'll ever find such imperfect amp for me, hence I prefer the perfect amp, and then add those imperfections by signal processing.
I know it sounds like a bit of mouthful, but that is my takeaway of our short discussion. Not sure if you agree.
Last edited: