Two sounds compared

I wanted you to tell me a little bit about pink noise

Wait a second, if you are an electronic engineer and you already know these things, why don't you talk about them? What are we doing, the interrogation between professor and student?

It's not like that, you write it better!

I believe they are all interesting things for the reader and I would gladly talk about them, but I am too lazy to write ...

What interests me instead is isolating a few concepts that apply well to piano life, namely that of timbre or harmonic composition, as you well explained, and that of the low-pass filter, which is able to exclude high-frequency components, paradoxically those that most people's ears do not like.

Timbre because it changes over the years, decays with the aging of the steel, the copper.... Filter because even the soundboard ages, becomes less and less elastic, making the timbre poorer...

Decay of the piano.

I don't understand why today people want to kill a new piano prematurely, with indiscriminate punctures to the heads of the Hammers

And I am referring back to the two uprights listened to!

Let's play a game: poor recording, very close to the Hammers, out-of-tune piano! Does anyone know how to categorize the type, the era, etc.?

Without being able to see it, it is beautiful.

http://yourlisten.com/channel/content/16961405/prova5

Hi Simone, congratulations on your post.

I didn't understand one passage in your first post. I need a hand to follow you.

You say: "Let's analyze the phase correlation to see how the Joint Stereo encoding acted on the spatiality", what do you mean... what is spatiality??

Reading it made me want to try doing an analysis and trying to interpret the results. Can I ask you in detail about the tools you used or others that could be used?

Thanks a lot!!

Joint Stereo is used when two channels of a stereo file are very similar. The less differentiable high frequencies are paired into a single channel, while the low frequencies, which are much more affected by channel misalignment, are split into two channels. The concept of audio space is our brain's ability to interpret from which direction a sound comes. Typically, if we have two speakers in front of us, one on the right and one on the left facing our head and occupying the focal position (the intersection of the perpendiculars passing through the center of the cones), emitting the same signal at the same intensity, our brain interprets a sound that virtually comes from the center. If we turn up one speaker more than the other, for example by increasing the volume on the right, closing our eyes will make it seem as if the sound source is moving from the center towards the right. Another important factor in determining spatiality is the phase of the two signals. In the case of complex signals (which contain many frequencies within a continuous whole), spatiality is given by the phase differentiation of the first channel relative to the second. This is helped by the phase correlator, which is a tool that calculates the cosine obtained from the sum of the angles at the same instant of two signals. We can immediately understand, therefore, that the phase correlator will have a range of values from -1 to +1. If the phase difference between signal channel A relative to B is 0° (there is no phase shift), the cosine is 1. If it remains stuck at 1 for the entire audio file, then we are in the case of a mono signal split across two channels, exactly what happened with the audio of the video we analyzed; if the value oscillates between 0 and 1, we are still fine because it means there is good spatiality—in particular, at 90°, the cosine is exactly 0. When values drop below 0, it is worrying and we must take action because we enter that phenomenon called phase cancellation. As long as the listening remains stereo, there is no more problem, but when the listening is mono, the result is the sum of the channels; however, if the channels move towards out-of-phase—to push it to the extreme, let's say the phase correlator reads -1—it means that in a mono listening we will hear absolutely nothing because the signal from channel A, when added to channel B, has cancelled itself out (two identical and periodic signals out of phase by 180° cancel each other). This is the importance of the phase correlator.

The other plugin I used performs frequency analysis. It basically performs a Fourier series in real time; it is an FFT (Spectroscope).

Fra wrote:

I believe these are all interesting things for those reading and I would gladly talk about them, but I am too lazy to write ...

I know they are interesting things for those reading. As for me, I lack the desire to write as much as I lack the time!

Damn Simò and Fra! Braaaavo!

You've dusted off FFT concepts I had forgotten. Unfortunately, I think they might not be arguments understandable by everyone: in fact, a few steps escaped me here and there. My mechanical background on the FFT is in the study of vibration signals in structures. What books would you recommend for further study of these topics? Or rather, where did you learn and/or studied them?

"If we raise one speaker more than another, for example by increasing the volume on the right, closing our eyes will make it seem as if the sound source from the center is moving towards the right. "

Is the sensation of movement linear with the increase in volume?

The increase in volume is already a non-linear thing with respect to dB SPL. If we have a device that produces a sound pressure of x dB, by adding an audio device of the same power we do not get 2x dB, we get an increase of only 3 dB. If I add another one, I get less than 3 dB because I would need to add 2 (2*2=4) and continuing... 8, 16, 32, 64, 128 systems of the same power to obtain only 3 dB more each time.

I knew this... But reasoning in dB, if I double the dB, I was wondering if the sensation of displacement would be equal to the increase in dB.

I don't think it is linear.

In fact, it isn't... I said it before. We are used to considering our sensations as linear. If we go to measure physically, when it "seems" to us that the power has doubled, we have added 3 dB. Therefore, in this case, if 3 dB were added to only one side, the contribution for us would be double on one side; consequently, the sound would seem to come out not from the center (1/2) but at the 3/4 position on the side where we increased the sound pressure level.

Assuming that one knows that to double the volume one needs to increase by 3 dB, then the sensation of movement becomes linear with this increase in volume. Clearly, if one reasons only based on sensations, even the Shepard tone is infinite...

I was wondering if the ear is able to perceive as a percentage how much the sound moves from one side to the other and, from what you are saying, it seems to me that this is exactly the case. Thank you for your answer!

But imagining a hypothetical listener not positioned in the center between the two speakers but in front of (let's say) speaker B, thus forming a right-angled triangle between the speakers and the listener... how much should I increase the volume of speaker A (the other one) to give the impression of symmetrical stereophony?

Beautiful question... mmm I wouldn't even know where to start answering.

micamahler wrote:

Assuming that one knows that to double the volume one needs to increase by 3 dB, then the sensation of movement becomes linear with this increase in volume. Clearly, if one reasons only based on sensations, even the Shepard tone is infinite...

I was wondering if the ear is capable of perceiving, in percentage, how much the sound moves from one side to the other, and from what you are saying, it seems to me that this is indeed the case. Thank you for your answer!

But imagining a hypothetical listener not positioned in the center between the two speakers but in front of (let's say) speaker B, thus forming a right-angled triangle between the speakers and the listener... how much should I increase the volume of speaker A (the other one) to give the impression of symmetrical stereophony?

Well, reasoning based on sensations, the Shepard tone is not infinite at all. After 4 or 5 times, one realizes that it is composed of two sounds that vary their amplitudes in relation to the frequency they are at. And so, after listening to the loop 4 or 5 times, you understand the trick and manage to perceive the two sounds. Sensations often go hand in hand (aside from units of measurement) with physical reality, so we must never underestimate what we perceive, because in 90% of cases it directs us toward the solution to the problem. In the remaining 10%, there is that band where the ear is bypassed for physical reasons (such as the resonance of the auditory canal, for example).

In the particular case of speaker positioning, as you have posed it, the discussion is truly long and complex. Right now I don't have the material time to answer you properly. As soon as I can, I will. Unfortunately, it is a long matter because I need to post some figures to explain it decently.

I was just saying, the feeling it gives you is deceptive; of course, you eventually realize that it's not infinite, like the barber pole and all the other tricks...

Thanks in advance for the response to come!

Excuse my ignorance, but what is a shepard tone?

I discovered it while reading a book and I almost had a heart attack!

Have you ever seen those optical illusion examples, like Escher's infinite staircase? Well, the Shepard scale is an illusion comparable to that, only it is auditory in nature.

It consists of an ascending or descending scale that never ends, or at least one has the perception that it does.

Our auditory system gets fooled XD

The way it is constructed isn't quite clear to me, but if the scale is ascending, as you go up with the notes, the volume decreases and they are replaced by notes with the frequency of the previous octave.

I recommend listening to György Ligeti "the devil's scale" where the composer exploited this effect.

Ciao

I think I understood the meaning well, but I don't quite understand how such an effect could be achieved.