Is it time for the traditional 1kHz tone test for amp power to go? (brief Amir name drop)

[Blumlein 88]

Hallo Ray, hey folks,
Can you tell me, what the graph says?
For me as a non technical guy, I just want to know, if you find the thoughts of the scientific audiophile intelligent / relevant or not.
What he stated is: with real music / or pink or white noise the modern digital desktop amps don’t perform well.
This is not what I hear, if I compare the Fosi V3 with other amps. I think it sounds like a good tube amp.
But this is just subjective.

Is the scientific audiophile perhaps at least a little bit right?

Not at all. One problem with SA is his measurement method is poor. He may be trying to measure the strength of a signal spread over 20-20khz with a meter that doesn't respond past 400 hz or 1000 hz. No surprise the wideband signal level is well down vs a low frequency tone if that is what you are doing. The problem is not the amp it is how he measured it.

thorsten

The definition of a square wave is a sine wave with decreasing odd order harmonics. Theoretically to infinity. What Ray is showing is simply that with a low frequency square wave and a good enough FFT we can confirm harmonics in the hundreds well up into higher frequencies. You will get spikes in the graph at 30 hz, 50 hz, 70 hz, 90 hz and so on for octaves. So it would be something like a multi-tone signal with a 1000 tones.

Then there is Pkane's software called...Multitone....which does many things, but was originally a multitone test software. You can create all sorts of multitone test signals with about as many tones as you wish.

 

[IAtaman]

Guy is clearly dropping every big name in the audio review world, probably in the hopes to get some reaction and traffic.

 

[ads_cft222]

On what basis did you recommend the amp in the review ? Insufficient power , unstable frequency response, one channel leaking to another … it may be 300$ but the ratio is defined by value/money… where is the value ?

 

[DHT 845]

From what I undestand the issue outlined by the "Scientific Audiophile" the problem with multitone test (i.e. even more important than just 1kHz tone test)
is that it does not make all these tones AT THE SAME TIME (look 8:15 of the review) like in real music.

So please be honest and tell me:

1) Isn't it true (about simultaneity of tones in the test)?
2) Isn't it possible that complex 20-20kHz signal reduces the produced power at a given distortion limit?

 

[sergeauckland]

From what I undestand the issue outlined by the "Scientific Audiophile" the problem with multitone test (i.e. even more important than just 1kHz tone test)
is that it does not make all these tones AT THE SAME TIME (look 8:15 of the review) like in real music.

So please be honest and tell me:

1) Isn't it true (about simultaneity of tones in the test)?
2) Isn't it possible that complex 20-20kHz signal reduces the produced power at a given distortion limit?

Music doesn't do that either.
A music signal can only ever have one voltage value at any one instant in time. It doesn't matter how many tones it is made up of, at any one instant there is only one value of voltage.

A sinusoidal 1kHz signal that swings both positive and negative will exercise the amplifier's transfer function, and harmonic distortion is a measure of how linear that transfer function is. Do the measurement at low, say 30Hz, mid, say 1kHz and high, say 15kHz frequencies, and you will characterise the transfer function including any change in linearity due to limited open-loop bandwidth.

Couple that with a frequency response measurement and a noise measurement, and you'll pretty much cover all the characteristics of an amplifier into one load.

For completeness, repeat into different loads if there's any concern, which is effectively what the Power Cube does.


S

 

[pogo]

And precisely such measurements are not usually made

 

[DHT 845]

Music doesn't do that either.

Really, so why Amrir measures intermodulation distortion 1kHz and 7kHz.
One frequency can influence other frequency that results in increadibly short voltage drop.
But if you would measure for ex. 20 tones at once it would be much more informative. Current in time is limited, 30Hz tone may suck up the potential for clean range of other frequencies. Proof -> high pass inline filter before the amp significantly improoves sense of resolution and separation.

A music signal can only ever have one voltage value at any one instant in time. It doesn't matter how many tones it is made up of, at any one instant there is only one value of voltage.

But the voltage change in very short period of time is more important. There is no music in instant time. Instant time is just the point of start. Like beginning of the vector.

A sinusoidal 1kHz signal that swings both positive and negative will exercise the amplifier's transfer function, and harmonic distortion is a measure of how linear that transfer function is. Do the measurement at low, say 30Hz, mid, say 1kHz and high, say 15kHz frequencies, and you will characterise the transfer function including any change in linearity due to limited open-loop bandwidth.

I do not think the 2nd sentence is true.

Couple that with a frequency response measurement and a noise measurement, and you'll pretty much cover all the characteristics of an amplifier into one load.

The same as above, that is a very reductionist approach.

 

[peng]

Hallo Ray, hey folks,
Can you tell me, what the graph says?
For me as a non technical guy, I just want to know, if you find the thoughts of the scientific audiophile intelligent / relevant or not.
What he stated is: with real music / or pink or white noise the modern digital desktop amps don’t perform well.
This is not what I hear, if I compare the Fosi V3 with other amps. I think it sounds like a good tube amp.
But this is just subjective.
Is the scientific audiophile perhaps at least a little bit right?

thorsten

The part that he simply expressed his opinions about the amps he mentioned is obviously not wrong, because it is totally his subjective opinions. Where he may be wrong is about his so called "uncorrelated" stereo noise test, and with that test rack he claimed his 160 W rated Yamaha "ref" amp dropped to only 8.6 W. Now, there is nothing too wrong about that either consider it is just an internet YT video, but he seemed to be referring that as "facts" (don't recall the exact wording, so may be he meant that to be his opinion too lol...), that would be wrong because if he wanted the results to have any credibility, at the minimum he should define and described how he came up with that noise track, and showed how he came up with the mere 8.6 W, that dropped all the way from 160 W based on a 1 kHz signal test. With that info, then the result does not mean much at all, for example, he did say one thing, that the test signal does not have reference of 0, -10, -30 or whatever dB, so how's that 8.6 W came from, and what did he prove, nothing really!

He seemed to use that test to prove that the 1 kHz test should be dropped because with real music it is 1 kHz, but more like his stereo noise test track. He also seems to know full well that the little amps he mentioned could be fine for people listening from 2 meter and getting spl in the 90 dB range, and that for real music listening, the average amp output could be much less than 1 W, so 8.6 W average is plenty for music listening. Think about this, with music, or his so called test track, without know the details, that 8.6 W, if is an average value, the highest instantaneous value could be 86 W or even 860 W, or higher, sufficient to clip his A-S2100 amp!!

So, right or wrong, I resent people posting such video that could mislead (not deceiving as such, to give benefit of doubt) people who may not have the technical knowledge to deceiver whether his claims are factually relevant to his claims.

Just to clarify my point about his test track, we should know that real world music contents could have dynamic peaks >20 dB, even >30 dB, take a look:

Album list - Dynamic Range DB

So, let's say his test track has 20 dB, then 10 W average would mean the highest peak could be as much as 1,000 W, or more!!!

How he measured matter, by not providing the needed facts, the presentation cold be misleading at best.

By the way, the way he use a digital multimeter to measure voltage is laughable, without stating accuracy that could be expected from his example results. He seemed to be using a low cost DVM (I would use something like a $500 Fluke at the minimum), if he is so keen on stopping the 1 kHz test he should use something better, let alone using the kind of precision instrument that Amir and other reputable reviewers use. Those low cost multimeters surely don't do well at above a few thousand Hz. This is not my main point, just an aside...

 

[sergeauckland]

Really, so why Amrir measures intermodulation distortion 1kHz and 7kHz.
One frequency can influence other frequency that results in increadibly short voltage drop.
But if you would measure for ex. 20 tones at once it would be much more informative. Current in time is limited, 30Hz tone may suck up the potential for clean range of other frequencies. Proof -> high pass inline filter before the amp significantly improoves sense of resolution and separation.

But the voltage change in very short period of time is more important. There is no music in instant time. Instant time is just the point of start. Like beginning of the vector.

I do not think the 2nd sentence is true.

The same as above, that is a very reductionist approach.

It's electronics, not magic.

If we're measuring linearity of the transfer function, then a single tone is fine. Using two tones is just another way of assessing linearity, as there is a mathematical relationship between harmonic and intermodulation distortion.

If you have an amplifier where a "30 Hz tone may suck up potential for clean range of other frequencies", I suggest you throw it away and buy a proper amplifier.

S.

 

[DHT 845]

It's electronics, not magic.

If we're measuring linearity of the transfer function, then a single tone is fine. Using two tones is just another way of assessing linearity, as there is a mathematical relationship between harmonic and intermodulation distortion.

If you have an amplifier where a "30 Hz tone may suck up potential for clean range of other frequencies", I suggest you throw it away and buy a proper amplifier.

S.

No, EVERY emplifier with perform objectively better with line hi-pass. It is only the matter how much better.
Similar issue: I bet there is no amplifier in the world (with traditional PS) that would not perform better if you for example double its electrolytic caps (by adding exactly the same capacitor parallel to all caps). Maybe in 1kHz or some other simple 1-2 tone tests will look the same. But with real music there will be improovement in less overall intermodulation distortion.

 

[peng]

But the voltage change in very short period of time is more important. There is no music in instant time. Instant time is just the point of start. Like beginning of the vector.

That's one of my points too, and so slew rate is important, but like damping factor, it tends to get over blown on the internet audio forums.

Even the low cost opamps D+M, Y, O and other brands used the likes of the NJM8080, 2068, NE5532 etc., in their preamp/processor section can do 5 V/us, that's much faster than needed for 20 kHz to be produced by a 100 W 8 ohms amp.

The V3 used in his example YT video did only 62 W 4 ohms in Amir's test so I would say for music listening, slew rate would not likely be an issue.

5.4: Slew Rate and Power Bandwidth

eng.libretexts.org

Faster, more, higher usually mean "better", not always obviously, but we can say the same about many things, one just have to buy more than enough to satisfy their need, if they need no more than what the V3 or Wiim, than even this YT reviewer said himself, those things can do the job.

I don't know what you meant by "There is no music in instant time"? The point some of us are making, is that whatever waveform of the music signal you are feeding the amp, the amp will just amplify that signal, that signal, will have one unique value at one moment/instant, whatever you like to call it. I assume some people might think the amp will be amplifying multiple signals at any moment? I don't know about that... Instead of me trying to say it in a clumsy way, I remember reading something, yeah on the internet too, so I quickly search and found one of those:

Here's the relevant part of the article:

Elliott Sound Products - Audio Power Amplifier Design Guidelines

ESP - The Audio Pages. Amplifier design techniques, find out how amplifiers work, and how to design them.

www.pe2bz.philpem.me.uk

Measurements Versus Subjectivity
If I never hear of someone complaining that "distortion measurements are invalid, and a waste of time" again, it will be too soon. I am so fed up with so-called experts (where "x" is an unknown quantity, and a "spurt" is a drip under pressure) claiming that "real world" signals are so much more complicated than a sinewave, and that static distortion measurements are completely meaningless.

They are not meaningless, and real world signals are sinewaves! The only difference is that with music, there is usually a large number of sinewaves, all added together. There is not a myriad of simultaneous signals passing through an amp, just one (for a single channel, naturally).

Since physics tells us that no two masses can occupy the same physical space at the same time, so it is with voltages and currents. There can only ever be one value of voltage and one value of current flowing through a single circuit element at any instant of time - if it were any different, the concept of digital recording could never exist, since in a digital recording the instantaneous voltage is sampled and digitised at the sampling rate. This would clearly be impossible if there were say 3 different voltages all present simultaneously.

 

[sergeauckland]

No, EVERY emplifier with perform objectively better with line hi-pass. It is only the matter how much better.
Similar issue: I bet there is no amplifier in the world (with traditional PS) that would not perform better if you for example double its electrolytic caps (by adding exactly the same capacitor parallel to all caps). Maybe in 1kHz or some other simple 1-2 tone tests will look the same. But with real music there will be improovement in less overall intermodulation distortion.

You can believe that if you want. I prefer to believe my measurements which show no such thing.

If the power supply behaves as you describe, it's clearly inadequate. Once there is enough capacitance to hold sufficient charge to provide its rated output into rated loads, then everything else is overkill, and therefore unnecessary.

S.

 

[valerianf]

"Once there is enough capacitance to hold sufficient charge to provide its rated output into rated loads, then everything else is overkill, and therefore unnecessary."
This is the problem with modern amplifiers: a tiny transformer + a small value capacitor are giving a lot of distorsion.
In the old time we had a toroidal transformer + some large capacitors that were providing a real energy reservoir for music transients.
Nowadays, with the very light class AB amplifiers the output power available without distorsion is always very limited when a large current is needed.
Currently there is not any current capability test that is used when benchmarking the audio amplifiers.
It is a mistake.

 

[sergeauckland]

"Once there is enough capacitance to hold sufficient charge to provide its rated output into rated loads, then everything else is overkill, and therefore unnecessary."
This is the problem with modern amplifiers: a tiny transformer + a small value capacitor are giving a lot of distorsion.
In the old time we had a toroidal transformer + some large capacitors that were providing a real energy reservoir for music transients.
Nowadays, with the very light class AB amplifiers the output power available without distorsion is always very limited when a large current is needed.
Currently there is not any current capability test that is used when benchmarking the audio amplifiers.
It is a mistake.

If the amplifier is measured with different loads, that also tests the current capability. The Powercube will also test different phase angles of the load, so suitable tests do exist and are used.

S

 

[pogo]

But not parallel, e.g. 2ch, 3ch, ...

 

[peng]

Currently there is not any current capability test that is used when benchmarking the audio amplifiers.
It is a mistake.

No idea why you think that, Amir has done quite a few such tests on current capability, with the available data on ASR, Sterophile, Audioholics, we should be able to do some benchmarking ourselves. One example is his recent tests on the new Fosi Mono channel amp.

Can it be better, and/or more? Of course, but to say "there is not any......", does not seem to be true, unless I missed, or misread something. I would agree that on "benchmarking", we may have to do something ourselves, but again, the data are there, though not widely.

 

[peng]

Guy is clearly dropping every big name in the audio review world, probably in the hopes to get some reaction and traffic.

The problem is, currently there is no requirement for anyone to pass any electrical/electronics/physics theory/principles knowledge tests!! So anyone can dream something up based on their incomplete (far from...) tiny bit of knowledge, combined with their own misconceptions, and BS (not intentionally I would think...) via YT, and it they present their views, and/or findings, they would in fact get enough traffic to sustain themselves, even make a fortune on such BS videos (again, I assume they meant good).

As much as I tried not to, I contributed to the "traffic" too, but very reluctantly. The thing is, after resisting to say anything about it, then when I failed the temptation to say something, I felt compel to watch the who video, so that I could say something without doing so without watching the whole presentation. I suspect I am not the only who contributed so reluctantly, because without watching the whole thing, I bet many of us could smell something not good lol...

 

[pogo]

One example is his recent tests on the new Fosi Mono channel amp.

But .... Note that these are short term power ratings.
And if I have understood correctly, this is only checked at one frequency.

 

[Sokel]

As devices go smaller and smaller so does the heat dissipation area.
And if we look at chip amps or module amps data sheet for specs the fine writing often includes 100mm heatshinks in the test conditions for the rated output and sometimes even active cooling.

And there come threads of amps probably hitting limits and people we think is biased until we see actual results all over (diyaudio is full).

Wouldn't it be better to up the duration of such test just a little?
It doesn't have to be destructive but just as much.

Thermal imaging would be great too,a LOT of people talk about how their amps go hot only to be answered by others that their's don't.
Of course loads can differ greatly but we're talking about idle sometimes.These people would then seek for other problems,up or down the chain.

 

[peng]

But .... Note that these are short term power ratings.
And if I have understood correctly, this is only checked at one frequency.

As I said, it could be better, and more such tests. As for duration, very true, it could be for short term, or even very short term, hopefully @amirm may chime in and tell us something about the duration he used for those measurements. To me, even 50 ms, i.e. 0.05 second, or less could be long enough for practical purposes, though again, longer price would be nice, no argument there. Still, Amir's tests are rarely (not never, but rare, relatively speaking) done on other bench test sites, so let's recognize that fact!!!

You seem like a NAD fan, so, do you have their test results, that used more frequencies such as 20-20,000 Hz, and for longer duration, even continuous? I like their amps too but I could not find any such current capability tests. Also, such test need to include voltage consideration. There could be such so called high current amps that could drive 2 ohm load on continuous basis but are voltage limited, to the point it may not do well on even 6-8 ohm nominal loads with dips down to below 2 ohms.

The other sad fact is, people often overrate the so called high current requirements while ignoring the nature of speaker impedance that could vary a lot over the audible band, such that in fact the short duration test Amir has been doing may actually be more appropriate than similar tests using longer duration, for real world music listening, as music signal, if that could cause certain speakers to dip below 2 ohm, the duration of such dip would usually be very short anyway.

Same thing as the so called all channel driven tests, lots of such tests, while good to include, are not really helpful for people who are in the process of choosing an amp that can do the job well for them, but without paying for things/features/yeah current/voltage capability they may never need.