Interesting aspect of phase

[MDNSC]

Today I saw a video talking about a phase shifted square wave.
In that video, he show cased how a square wave and a 90 degree phase rotated square wave sounds like, which sounds the same in frequency perspective of course.

But interestingly, the phase shifted square needs more dynamic range to sound as the same loudness compare to the original square.

On the other hand, it means that if both signals are at the same true peak, the phase shifted version of the square wave will sound quieter.

Thats an very interesting aspect of phase for me. As we here already know that absolute phase does not affact what we hear in the real world, but will that affact the dynamic of sound in this case?

For example, 2 speakers, one playing the original square wave and the other plays the 90 degree phase shifted square wave to achieve the same sound same loudness level. The speaker will distort more on the phase shifted one because that waveform needs to have more headroom to sound the same as the original square wave.

For more but not sure, does that means that speakers with linear phase operations may have more dynamic advantages to non linear phase speakers? If they both playback this square wave, will the linear phased speaker sounds louder in this case?

 

[markanini]

The speaker will distort more on the phase shifted one because that waveform needs to have more headroom to sound the same as the original square wave.

It's a question of headroom before it reaches the speaker too. Touching digital 1 it's a given. If a limiter used it will be reduced in volume, assuming a perfect limiter.

 

[Hayabusa]

Today I saw a video talking about a phase shifted square wave.
In that video, he show cased how a square wave and a 90 degree phase rotated square wave sounds like, which sounds the same in frequency perspective of course.

But interestingly, the phase shifted square needs more dynamic range to sound as the same loudness compare to the original square.
View attachment 267001
On the other hand, it means that if both signals are at the same true peak, the phase shifted version of the square wave will sound quieter.

Thats an very interesting aspect of phase for me. As we here already know that absolute phase does not affact what we hear in the real world, but will that affact the dynamic of sound in this case?

For example, 2 speakers, one playing the original square wave and the other plays the 90 degree phase shifted square wave to achieve the same sound same loudness level. The speaker will distort more on the phase shifted one because that waveform needs to have more headroom to sound the same as the original square wave.

For more but not sure, does that means that speakers with linear phase operations may have more dynamic advantages to non linear phase speakers? If they both playback this square wave, will the linear phased speaker sounds louder in this case?

so question is, who is doing this shift? The speaker?

 

[MDNSC]

Another video talking about phase rotated square wave, pretty interesting

 

[GXAlan]

This is interesting. @amirm, can you comment? If your phase shifts alter the point of clipping and different amplifiers respond differently to clipping/recovery, could phase shifts be inaudible unless it triggers the amp into clipping?

 

[Hayabusa]

This is interesting. @amirm, can you comment? If your phase shifts alter the point of clipping and different amplifiers respond differently to clipping/recovery, could phase shifts be inaudible unless it triggers the amp into clipping?

seems obvious?

 

[DonH56]

The RMS value of a square wave of amplitude A is simply A. The RMS value of a sine wave (single tone unclipped) is A/sqrt(2) = 0.707*A. (Left for posterity.)

The crest factor (peak to RMS value) of a square wave is 1.0, but the crest factor of a sine wave is 1.414 (sqrt(2)). As you introduce phase shift that distorts the square wave you are changing its crest factor so peaks are higher for the same loudness (RMS value).

Much more to it, but that's a simple explanation.

HTH - Don

 

[Hayabusa]

The RMS value of a square wave of amplitude A is simply A. The RMS value of a sine wave (single tone unclipped) is A/sqrt(2) = 0.707*A. Thus, using RMS values, a perfect square wave is louder than a perfect sine wave of the same amplitude. As you introduce phase shift that distorts the square wave you are lowering its RMS value so it does not sound as loud.

Much more to it, but that's a simple explanation.

HTH - Don

The RMS value should not change if you phase shift some of the harmonics..

 

[DonH56]

The RMS value should not change if you phase shift some of the harmonics

Sorry, had crest factor (peak-to-RMS) in mind, and what I meant to say was crest factor. Corrected my post.