First cycle distortion - Graham, what is that?

It tries to decompose the transition process into a spectrum. Proves that this is exactly what affects the sound quality of amplifiers. That's just can not understand that in normal operation, such signals are not received at the input to provoke such a response. There is no overload on SR and NFB has time to track and correct. I don't know how else to explain it more clearly.
Maybe shouldn't just throw insults like that ?

You're right! What if that's all they can do

In parallel with Graham, another author worked on a high fidelity amplifier (only according to the first version of Cyrill Hammer - without NFB), 2005 - quote:
“I recently completed the development of an ultralinear UN. The amplifier has very low distortion typical of feedback systems, but it is built without feedback. " Uni & Vers (M. Kulish)
The work was published in the Russian magazine Radio 2005-12
To measure distortions by the compensation method, the author used a transformer that made it possible to measure the introduced distortion at a level of up to minus 150 dB.

Regarding measurements, he writes:
“By inertia, since the times of Baksandall and Akulinichev, compensation of the main tone by phase adjustment has been applied. In fact, phase compensation should only be used in exceptional, special cases. On the contrary, information about the phase (about the contribution of phase distortions) is no less important than information about the amplitude "

Such measurements remind me of the Hafler method, where phase adjustment is also not used.
The virtual measurement method using a signal delayed by an ideal delay also allows you to estimate the actual distortion introduced.
But do not explain to the hamam - they never get it. Instead of penetrating and understanding, they continue to be rude and insult as in the days of Graham.
Last edited:
Завязывайте с оскорблениями.
Maybe shouldn't just throw insults like that ?

Petr is having fun. He invests 15min a day posting random quotations and five clicks simulation charts spiced with conclusion in provocative tone. Then he gets all day fun watching others attempting to make reason of it.

If you think this is not the case, could you please restate his thesis in one or two sentences without hiding behind links to lengthy articles?

You are right, no need for insults. Such dud posting begs for GIGO approach.
Another post with essentially no content, no point made, just quoting some random stuff.

Actually, I do feel insulted. Spending much time and effort to explain stuff, only to be ignored and see refusing any intelligent discussion, yes that insults me.

It's strange that basic things don't reach anyone.
I'm tired of explaining that the program I'm working with turns the signal of the first period into a periodic signal, regardless of whether there is a phase gap or not.
Here is an example from a textbook of such a signal and a Fourier expansion formula. Everything is painted over 60 years ago, I think there are earlier publications. The first period just characterizes the speed distortions that are not detected by any other testing methods. Spectrum analyzers are not able to measure the instantaneous spectrum of a musical signal - they show an average value - analogous - the average temperature of patients in a hospital.
Distortion meters measure the level of distortion in a steady state.
If someone does not understand basic things, then what Graham unsuccessfully explained in his time, then there is nothing to be angry and offended - pass by, this is not for you.


  • Atabekov_1956_book.png
    328.2 KB · Views: 132
  • Atabekov_1956.png
    89.7 KB · Views: 139
Here is an example from a textbook of such a signal and a Fourier expansion formula.

Periodic non-sinusoidal function fulfilling Dirichlet’s criteria, can be represented in a form of infinite harmonic order – Fourier’s order. Sum of that order equals with the value of f(t) in every point of continuity of that function and in the points of discontinuity it gives the arithmetical average of the left and right limit values of f(t) i.e. ½*[f(t-) + f(t+)].”

It says periodic. First cycle is not periodic, it happens only once.
Petrov seems to be trying to hold a power amp to the same standards as might be applied to an opamp used for precision analog computing up to 20kHz. Looks like that is where the Dostal references come into the picture.

The argument at this point seems to be, most people don't consider that to be necessary and or practical for audio power amps.
AX tech editor
Joined 2002
Paid Member
But Mark, even the best opamp would fail at an unfiltered infinitely-fast step response.
That seems to be his argument: an amp confronted with an infinitely fast step cannot follow that step faithfully. What we all already knew, and understand why.

Its like trying to accelerate a material object from zero to some speed in zero time. You would need an infinite amount of power to do that, which is more than is available in the universe ;-)

Yes, quite right. However, such waveforms are sometimes used in simplified explanations, or simplified models that don't rely on FFT. For example, idealized step and impulse functions can be very useful in Laplace Transform engineering calculations. In that case engineers are presumed to understand the practical limitations of the models.