Wow, you engineering knuckleheads sure know how to put a damper on an exciting thread
Hrrmph. That pocket-protector stuff is what got you the NCore in the first place.
Maybe I'll start in on something with lotsa equations. Yer eyes'll cross real good then.
That'll learn ya. @mrpp: That is one of my favourite XKCD strips. I think I'll print it out and pin it up in the kitchen, see how she reacts
@iand: I understand the silicon real estate argument. So far I found it very difficult to conceive a system where this potential economy is actually realised in practice. This is actually why we've held off starting serious digital feedback projects so far. I still have no idea what digitally controlled amp could beat the price and performance of a cheap DSP a la TAS3108, a 6-channel DAC and 6 Ncores.
@chocoholic: That was precisely the nub of the paper. It shows 2 spectra taken from the same signal. Once interpreted as a digital code (result: perfect), once interpreted as an analogue signal (result: rubbish). Speakers don't understand code, so designing the amplifier must indeed be treated as an analogue problem. Since treating it as "digital" yields an unexpected and clearly suboptimal result, why insist on calling it digital and get all confused?
@Juhleren. Your exposé is entirely correct. The problem I have with it is the context in which it appears. Research scientists should be extremely well aware of the frailty of their knowledge because much of what they do is cutting edge and therefore still very much open to falsification. No scientist should ever think their job is done. Good scientists understand this intuitively or they wouldn't be good scientists. However, the longer a hypothesis (and later theory) is tested without ever failing, the safer it becomes. This is not obvious to the layman who reads the same exposé. In isolation it gives the impression that all of science is tentative and anything is equally likely to be disproven at any moment. I have heard this argument put forward all too often by people who doubt science but who, curiously, are prepared to believe just about anything else: "well you openly admit that you're looking for the Higgs Boson because you're not sure about it, doesn't that equally mean you can't be sure that gravity can't be suspended by a meditating buddhist monk?"
A scientist will use this understanding as an impetus to keep refining his insights. For the audio objectivist that means: keep on edge and be prepared to test and measure stuff that you haven't before. The audio subjectivist however, should not deduce from this that all measurements are bunk. Never underestimate just how much stuff (including some that people think of as esoteric) that we can already measure and predict perfectly using existing methods and equipment.
@slowlearner Well that's a polite way of calling offtopic Point taken.
@bbggg: Actually we haven't sent M a sample yet. Good reminder. The company who's showing an NC1200 based amp on the CES is American. The second (I don't think they're showing it yet) as well.
@iand: I understand the silicon real estate argument. So far I found it very difficult to conceive a system where this potential economy is actually realised in practice. This is actually why we've held off starting serious digital feedback projects so far. I still have no idea what digitally controlled amp could beat the price and performance of a cheap DSP a la TAS3108, a 6-channel DAC and 6 Ncores.
@chocoholic: That was precisely the nub of the paper. It shows 2 spectra taken from the same signal. Once interpreted as a digital code (result: perfect), once interpreted as an analogue signal (result: rubbish). Speakers don't understand code, so designing the amplifier must indeed be treated as an analogue problem. Since treating it as "digital" yields an unexpected and clearly suboptimal result, why insist on calling it digital and get all confused?
@Juhleren. Your exposé is entirely correct. The problem I have with it is the context in which it appears. Research scientists should be extremely well aware of the frailty of their knowledge because much of what they do is cutting edge and therefore still very much open to falsification. No scientist should ever think their job is done. Good scientists understand this intuitively or they wouldn't be good scientists. However, the longer a hypothesis (and later theory) is tested without ever failing, the safer it becomes. This is not obvious to the layman who reads the same exposé. In isolation it gives the impression that all of science is tentative and anything is equally likely to be disproven at any moment. I have heard this argument put forward all too often by people who doubt science but who, curiously, are prepared to believe just about anything else: "well you openly admit that you're looking for the Higgs Boson because you're not sure about it, doesn't that equally mean you can't be sure that gravity can't be suspended by a meditating buddhist monk?"
A scientist will use this understanding as an impetus to keep refining his insights. For the audio objectivist that means: keep on edge and be prepared to test and measure stuff that you haven't before. The audio subjectivist however, should not deduce from this that all measurements are bunk. Never underestimate just how much stuff (including some that people think of as esoteric) that we can already measure and predict perfectly using existing methods and equipment.
@slowlearner Well that's a polite way of calling offtopic
Point taken.@bbggg: Actually we haven't sent M a sample yet. Good reminder. The company who's showing an NC1200 based amp on the CES is American. The second (I don't think they're showing it yet) as well.
Last edited:
The SMPS definitely sounds better. I haven't done the test with the Ncore modules but I did test the UcD400HG with the SMPS400 vs the HG supply. The SMPS is definitely better. It measures better too (THD at low frequencies). This is not illogical considering that the SMPS keeps the rail voltages symmetrical. This is discussed at length somewhere in this thread but one can be forgiven for having overlooked it among 1523 other posts
Guys, regarding commercial offerings I suggest you forget about "M" - at least for now...
Go all the way back. To the genesis of our world. There is actually an... anniversary of that right about now!
Bruno, interesting stuff one whole linear vs SMPS thing.
Any chance you examined Ncore's performance using a regulated SMPS? Was there any measurable benefit?
Really?
And here I am thinking people asking for listening impressions were getting in the way of the useful and interesting discussion going on about the technical matters.
Even casework is more interesting (and certainly more useful) than listening impressions from someone I know nothing about, in a system I know nothing about, in a room I know nothing about.
But that's just me.
Seriously people, I like reading listening impressions with my morning coffee and everything and I like to hear stories regarding how Ncore killed the dragon (= $$$$ power amps) that I can tell my friends, but let's not lose perspective.
It's not like buyers are pouring in to post their listening impressions and we're holding them back, either...
They will post if they want to post and when they have something to post, I guess.
Go all the way back. To the genesis of our world. There is actually an... anniversary of that right about now!
Bruno, interesting stuff one whole linear vs SMPS thing.
Any chance you examined Ncore's performance using a regulated SMPS? Was there any measurable benefit?
Wow, you engineering knuckleheads sure know how to put a damper on an exciting thread
Really?
And here I am thinking people asking for listening impressions were getting in the way of the useful and interesting discussion going on about the technical matters.
Even casework is more interesting (and certainly more useful) than listening impressions from someone I know nothing about, in a system I know nothing about, in a room I know nothing about.
But that's just me.
Seriously people, I like reading listening impressions with my morning coffee and everything and I like to hear stories regarding how Ncore killed the dragon (= $$$$ power amps) that I can tell my friends, but let's not lose perspective.
It's not like buyers are pouring in to post their listening impressions and we're holding them back, either...
They will post if they want to post and when they have something to post, I guess.
Last edited:
Yes (my lab supply) and no respectively.
And here I was wondering why the two most recent posts with listening impressions were getting exactly no responses from those asking for them... Ahh forum dynamics
So I guess there is no reason for the added complexity/cost of a regulated SMPS, at least in this application.
Oh, by the way, I'm still going with the bulkiest linear PS I can afford and will bypass the discrete input stage for a tube one (using the oldest toobz I can find).
I'm just making conversation until more people post their listening impressions!
Last edited:
"Probably the only reason why people think not relying on measurements is somehow acceptable is because it's easy to confuse audio with music. I hate to break it but music is art, and audio is engineering. I don't like engineering to get in the way of art, which is why I think it should be undetectable."
Audio is sound and all sound is detectable, because human hearing can pick up very tiny differences - this has been known since 1860 and Wundt's psychophysics experiments with the threshold of perception in Leipzig. The "measurements are everything" argument runs the risk of becoming the old "operation successful, patient died" argument.
The ultimate arbiter for audiophiles who seriously want to re-create music in their home is how "musical" that sound is. For audio "consumers" there are many other arguments like price, convenience, brand, size, etc, and as a "consumer" I'd prefer a digital amp because it consumes less electricity than class A amps and tubes. But as a pro musician who heard live musical instruments at a few feet distance for a good 50 years I want to hear sound that is as close as possible to the timbre and tone colour of live instruments. Engineering is only a means to that end.
The problem as Bruno and everyone knows is that you just can't quantify taste and perception of musicality - it has defied any attempts to unify listener's preferences. Some go for soundstage, others for warmth, others for dynamics, my own preference is timbre etc etc. So frankly we're back to measurement - by default. It doesn't tell us how "musical" a product sounds, but it IS something that comes very close and thank God it is quantifiable.
I still believe that audio is sound - not engineering. Spookily enough, that's what the word means. But I have enormous respect for Bruno and all the other audio pioneers because by engineering they are getting us closer and closer to the goal of reproducing sound.
andy
Audio is sound and all sound is detectable, because human hearing can pick up very tiny differences - this has been known since 1860 and Wundt's psychophysics experiments with the threshold of perception in Leipzig. The "measurements are everything" argument runs the risk of becoming the old "operation successful, patient died" argument.
The ultimate arbiter for audiophiles who seriously want to re-create music in their home is how "musical" that sound is. For audio "consumers" there are many other arguments like price, convenience, brand, size, etc, and as a "consumer" I'd prefer a digital amp because it consumes less electricity than class A amps and tubes. But as a pro musician who heard live musical instruments at a few feet distance for a good 50 years I want to hear sound that is as close as possible to the timbre and tone colour of live instruments. Engineering is only a means to that end.
The problem as Bruno and everyone knows is that you just can't quantify taste and perception of musicality - it has defied any attempts to unify listener's preferences. Some go for soundstage, others for warmth, others for dynamics, my own preference is timbre etc etc. So frankly we're back to measurement - by default. It doesn't tell us how "musical" a product sounds, but it IS something that comes very close and thank God it is quantifiable.
I still believe that audio is sound - not engineering. Spookily enough, that's what the word means. But I have enormous respect for Bruno and all the other audio pioneers because by engineering they are getting us closer and closer to the goal of reproducing sound.
andy
In my view a system should be as transparent as possible so as to reproduce the sound the sound-engineer has designed in his studio, if played in a similar room as a studio. I know this will never be the case, but I think it is a very inefficient empirical way to correct for room deficiencies by playing with "colored" audio components. If ready to correct/modify the spectrum etc, it would, IMHO be better to use DSP based on the measured room acoustics.
I might assume that no one is going to be pleased by a misleading commercial & the thread is about class D amplifier I believe?Wow, you engineering knuckleheads sure know how to put a damper on an exciting thread
Class D is a digital amp that is not workable without DSP does not matter if it has global analogue feedback around or not.
Also claims that PWM amp with switching frequency 300-450KHz can sound as good as HiEnd tube amp while some of tube amps have bandwidth up to 300KHz-500KHz is a an extremely optimistic approach (not realistic from a technic point of view to make it clear).
However it does not mean that all that related to NCORE in particular or NCORE is not an exciting item.
The general discussion regarding class D amplification was started by a link to an external web page with some (IMHO) very provocative general statements.
I guess it depends on how you define "sound" and "detect".
Do you detect the sounds of bats?
I think we are in total agreement on this one.
Well, yes, the word "audio" means sound. So anything having to do with sound. But when used in a context like the name of this forum, it usually means the recording and reproduction of sound. Producing the sound is art, artistry and skill. Reproducing it is technology, and technology is engineering and applied science. For that you do need the tools and skills of the engineer and the scientist, and measurements are important tools.
Suntechnik:
Bruno's (and most other) Class-D amplifiers use an analog signal. It matters not that this analog signal is the duty-cycle of a switching waveform, rather than a changing voltage or current. The duty cycle is still a non-quantised analog representation of the audio waveform. It is NOT a digital signal.
Bruno's (and most other) Class-D amplifiers use an analog signal. It matters not that this analog signal is the duty-cycle of a switching waveform, rather than a changing voltage or current. The duty cycle is still a non-quantised analog representation of the audio waveform. It is NOT a digital signal.
@suntechnik. Go easy on the "IMHO". Your opinions have no semblance of humility whatsoever. You have the most extreme phenomenal gall to try and teach me what class D amplifiers are.
Here's my last polite reply to you (as "polite" goes). Think about this: imagine an engineer who was trained to think mainly in current terms and only sparingly in voltage terms (the opposite of how engineers mostly think). Imagine his oscilloscope had current probes attached and the only voltage probe in the lab was in a cupboard gathering dust. That engineer would have no problem whatsoever designing or testing circuits. The current/voltage duality allows this. This engineer, when studying the signals in a class D amplifier would never see square waves. Instead he would see a continuously variable current alternately conducted by a pair of FETs. The idea of associating this with "digital" wouldn't even occur to this person. Therefore, the association with digital is dependent upon one's point of view. It is subjective. The fact that people think of class D as digital is an artefact of the fact that we mostly think and measure in voltage terms.
Now, I have noticed that you have made no attempt to respond to any of the replies that were given to your previous post. It is extremely disrespectful to expect people to listen to you without bothering to even read their replies in return and respond cogently. You are now simply restating your position and moving on to other items (now it's suddenly bandwidth). This is a one-way conversation: you yelling at the top of your voice with your ears plugged. This is not the behaviour of someone who wants to communicate. It is the behaviour of someone intent either on causing trouble or pressing his view.
I am calling troll.
Other than that, how come you have such strong views? Have you ever built any class D amplifiers yourself to learn about them and to test your views? No? Then eff off.
Here's my last polite reply to you (as "polite" goes). Think about this: imagine an engineer who was trained to think mainly in current terms and only sparingly in voltage terms (the opposite of how engineers mostly think). Imagine his oscilloscope had current probes attached and the only voltage probe in the lab was in a cupboard gathering dust. That engineer would have no problem whatsoever designing or testing circuits. The current/voltage duality allows this. This engineer, when studying the signals in a class D amplifier would never see square waves. Instead he would see a continuously variable current alternately conducted by a pair of FETs. The idea of associating this with "digital" wouldn't even occur to this person. Therefore, the association with digital is dependent upon one's point of view. It is subjective. The fact that people think of class D as digital is an artefact of the fact that we mostly think and measure in voltage terms.
Now, I have noticed that you have made no attempt to respond to any of the replies that were given to your previous post. It is extremely disrespectful to expect people to listen to you without bothering to even read their replies in return and respond cogently. You are now simply restating your position and moving on to other items (now it's suddenly bandwidth). This is a one-way conversation: you yelling at the top of your voice with your ears plugged. This is not the behaviour of someone who wants to communicate. It is the behaviour of someone intent either on causing trouble or pressing his view.
I am calling troll.
Other than that, how come you have such strong views? Have you ever built any class D amplifiers yourself to learn about them and to test your views? No? Then eff off.
Last edited:
@andyjevans: I have the impression we don't disagree much.
At the peril of boring people to death, I am constantly pointing out that the "measurements are all" argument can only be valid when one is prepared to accept that an established set of measurements is probably incomplete. At that stage you have to use your ears to find out if things are happening your measurements didn't predict and then refine the measurements. It all starts with ears. If not, what basis would we have to set the standards we want to meet? We have to start by figuring out what is audible. For instance, if we didn't, how would we know what bandwidth is required? Would 15GHz suffice? 100MHz perhaps? How much distortion can we tolerate? Without starting at the ear, well-meaning designers could end up building ampifiers with 5% distortion and a 500kHz bandwidth and think that because of that bandwidth it must really be a good amplifier. Most of the things we need to know about audibility is by now quite firmly established. A handful of things aren't. For that handful of things you have to do listening tests. For the remainder thankfully we don't.
I think one of the reasons why I like to hammer on the measurement nail is precisely your starting point: really subtle things count. It is weird to see people peddle amplifiers with high distortion figures and concomitant colouration and then find them claiming that some stuff is just too subtle to measure. If that isn't trying to have their biscuit and eat it I don't know what is...
If someone is put off listening by the sound of a particular amp I certainly hope he doesn't buy it. The problem is that when reviewers, who should be better equipped, do the same, you end up with the vicious circle I described with everyone using broken equipment. Once that starts happening in the recording end as well, the audiophile consumer will be the victim. He will have musicality only when his replay equipment plays ball with the equipment used during recording, mixing and mastering.
At the peril of boring people to death, I am constantly pointing out that the "measurements are all" argument can only be valid when one is prepared to accept that an established set of measurements is probably incomplete. At that stage you have to use your ears to find out if things are happening your measurements didn't predict and then refine the measurements. It all starts with ears. If not, what basis would we have to set the standards we want to meet? We have to start by figuring out what is audible. For instance, if we didn't, how would we know what bandwidth is required? Would 15GHz suffice? 100MHz perhaps? How much distortion can we tolerate? Without starting at the ear, well-meaning designers could end up building ampifiers with 5% distortion and a 500kHz bandwidth and think that because of that bandwidth it must really be a good amplifier. Most of the things we need to know about audibility is by now quite firmly established. A handful of things aren't. For that handful of things you have to do listening tests. For the remainder thankfully we don't.
I think one of the reasons why I like to hammer on the measurement nail is precisely your starting point: really subtle things count. It is weird to see people peddle amplifiers with high distortion figures and concomitant colouration and then find them claiming that some stuff is just too subtle to measure. If that isn't trying to have their biscuit and eat it I don't know what is...
I presume you mean to say it's all they have to base their judgment on? That sounds about right
If someone is put off listening by the sound of a particular amp I certainly hope he doesn't buy it. The problem is that when reviewers, who should be better equipped, do the same, you end up with the vicious circle I described with everyone using broken equipment. Once that starts happening in the recording end as well, the audiophile consumer will be the victim. He will have musicality only when his replay equipment plays ball with the equipment used during recording, mixing and mastering.Ah well, the word "music" has multiple meanings too. It can be the actual acoustical output, it can be the act, it can be the score, it can be the study.
Last edited:
A class D modulator converts voltage information into time information. There is no quantization whatsoever. Time resolution of electrons (¿jitter?) is not worse than potential resolution (noise).
Then output stage and output filter perform the opposite conversion, time to voltage.
Someone that doesn't understand the amplitude-time-amplitude conversion has a long learning way to go before he can have a clue about electronics.
There is no inherent distortion involved in such a conversion, only bandwidth limitations (to Fsw/2, aliasing takes place for higher components). Any distortion component measurable at the output of a class D amplifier comes only from aliasing and the inaccuracy of the modulator to exactly mimic the "inverse" of the response of the output stage and filter, like in any linear amplifier... But this is something subject to continuous improvement, there is not a known "improvement ceiling".
Then output stage and output filter perform the opposite conversion, time to voltage.
Someone that doesn't understand the amplitude-time-amplitude conversion has a long learning way to go before he can have a clue about electronics.
There is no inherent distortion involved in such a conversion, only bandwidth limitations (to Fsw/2, aliasing takes place for higher components). Any distortion component measurable at the output of a class D amplifier comes only from aliasing and the inaccuracy of the modulator to exactly mimic the "inverse" of the response of the output stage and filter, like in any linear amplifier... But this is something subject to continuous improvement, there is not a known "improvement ceiling".
Last edited:
That's absolutely true! Well observed. We probably don't disagree much, and our goals are probably identical. I still think that audio is sound, and engineering is a means to that end, so it's just a question of how we phrase it.
Best
andy
This is citation from Bruno's interview 2008 and up to me it shows how somebody who has knowledge and love to music (and sound) enthusiastically build amps. Ears - measurements - ears is only way to make musical equipment.
Bruno Putzeys: The Sound Of Music (Extended Play) - IEEE Spectrum
- Status
- Not open for further replies.