Lissajous / Phase scope weirdness, meanings / how to?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

user 37518

Well-known member
Joined
Nov 15, 2008
Messages
2,233
Location
-
I posted this thread on another forum, without any success, i guess everyone there is as lost as i am.... anyway here it goes, i hope i have better luck here

For years ive been looking at lissajous or phase scopes and till this day i really havent found an exact way to interpret them, depending on the scope it will be how its layed out, but lets assumume the vertical = mono, horizontal equals 180º out of phase layout which is not the typical XY.

Heres the deal, i really dont understand what the bottom 2 quadrants mean, at first i thought they were like a negative replica of the top 2 quadrants, but then i realized they are not, its obvious that the quadrants are opposed to each other, meaning that at the bottom, the right quadrant means L, while the left quadrant means R.

My first thoughts, is that the top represents the peak of a wave, and  the bottom means the valley of a wave, does this make sense?

In the same manner, i know that if a signal is completely horizontal, it is out of phase, or one side is completely out of phase to the other, but now i have the same issue, the image turned, and now, the rigth (or left) quadrant, is the equivalent of the top and bottom.

Ive been googling this for the las 2-3 days, and theres not much info on phase/lissajous meter reading, except for the obvious description.

So does anyone have a better definition as to what exactly this phase scope meter represents?, im NOT talking about the correlation meter, and im not looking for the textbook definition of "its phase versus amplitude", and all that crap, im looking for a real and easy way to actually read these bastards, common shapes, interpretation, etc.. aswell as the practical meaning of their axis/quadrants.

Ive been experimenting with them, and unless something is extremely out of phase, its still hard to look at the meter and realize something is wrong with the material being monitored, a little bit of experience would mean a lot here.

To add more to the confusion heres a screen shot of the Flux stereo tools phase scope i took while playing certain passage in a song i was checking, the bottom and top quadrants do not match at all, can someone explain what this means?

I read somewhere that this could mean DC offset, but i tried filtering out and it doesnt change a thing, plus, this only happens at certain parts of the song, so i can discard DC, this is a passage where the vocal is left with a piano and drumkit only, in an acoustic arrangement and song.

IDEAS?

scopef.jpg
 
I have a very old Heathkit Audio Scope (AD-1013) that does Lissajous Patterns.
I have never seen a track like yours.
With mine.
An in-phase mono signal is a diagonal line from the lower Left corner to the upper Right corner.
An out of phase mono signal is a diagonal line from the upper Left corner to the lower Right corner.
A max stereo signal is a round fuzzy ball.
Most program material is a very fuzzy wide diagonal line from the lower Left corner to the upper Right corner.
 
This software thingy is NOT a Lissajous screen!
A real Lissajous is as speedskater says.
This thing is a kind of vectorscope, that uses some phase correlation calculation and displays it on a "radar" screen with some remanence.
The way I interpret it, it shows that most of the content (over a period of time) is in-phase, but some of it is out-of-phase (which is not a problem because the energy is lower), with a not-too-wide stereo sperad.
I wonder what the user's manual says...
 
Thanks for your comments guys, Im aware that the image I posted may give the impression that the software is not a lissajous, but it certainly behaves like one, its just that that image is weird, which is why i posted it in the first place, What are you guys looking when you look at a lissajous meter (dont say the screen) i mean what do you guys focus on?, when do you know theres a problem? and i dont mean the obvious like when theres one channel with its polarity completely inverted, but i mean more like in the subtle things.
 
user 37518 said:
Thanks for your comments guys, Im aware that the image I posted may give the impression that the software is not a lissajous, but it certainly behaves like one,
Ok, it may be, but anyway the screen is rotated 45° compared to an XY oscilloscope. Now, if it really displays the L and R instantaneous amplitude, that indicates a highly asymmetrical signal, where the negative amplitude is about 1/3 of the positive.
What are you guys looking when you look at a lissajous meter (dont say the screen) i mean what do you guys focus on?
I look at the phase correlation, checking that the spot draws a 45° cloud (vertical with your software version)
when do you know theres a problem?
When my ears tell me  ;D
and i dont mean the obvious like when theres one channel with its polarity completely inverted, but i mean more like in the subtle things.
There is nothing subtle in a Lissajous when examining common audio program, because the spectrum is too wide and there are often a lot of uncorrelated signals. There are other applications where Lissajous can be more analytic; I used Lissajous a lot for THD evaluation, because it shows the order of products and their phase position.
 
abbey road d enfer said:
user 37518 said:
Thanks for your comments guys, Im aware that the image I posted may give the impression that the software is not a lissajous, but it certainly behaves like one,
Ok, it may be, but anyway the screen is rotated 45° compared to an XY oscilloscope. Now, if it really displays the L and R instantaneous amplitude, that indicates a highly asymmetrical signal, where the negative amplitude is about 1/3 of the positive.
What are you guys looking when you look at a lissajous meter (dont say the screen) i mean what do you guys focus on?
I look at the phase correlation, checking that the spot draws a 45° cloud (vertical with your software version)
when do you know theres a problem?
When my ears tell me  ;D
and i dont mean the obvious like when theres one channel with its polarity completely inverted, but i mean more like in the subtle things.
There is nothing subtle in a Lissajous when examining common audio program, because the spectrum is too wide and there are often a lot of uncorrelated signals. There are other applications where Lissajous can be more analytic; I used Lissajous a lot for THD evaluation, because it shows the order of products and their phase position.

As always, you offer great advice, im aware that the image is very asymmetrical signal, which is why im posting it. Something that still no one has mentioned is what do the lower quadrants represent?, like i said i thought they were just a negative copy of the L & R channels, but they are not, clearly visible in the pic i posted which is very assymetrical, would that mean that the lower 2 quadrants represent the valleys of a wave, while the top 2 quadrants represent the peaks of  a wave? in which case that image would mean that the bottom of my wave is somehow chopped off?
 
user 37518 said:
would that mean that the lower 2 quadrants represent the valleys of a wave, while the top 2 quadrants represent the peaks of  a wave? in which case that image would mean that the bottom of my wave is somehow chopped off?
Well, why don't you make a simple test? Create a rectified signal, either only positive or only negative, and see what the scope does.
 
spreemusik said:
user 37518 said:
would that mean that the lower 2 quadrants represent the valleys of a wave, while the top 2 quadrants represent the peaks of  a wave? in which case that image would mean that the bottom of my wave is somehow chopped off?
Well, why don't you make a simple test? Create a rectified signal, either only positive or only negative, and see what the scope does.
Probably not as simple as you think, because the signal chain doesn't pass DC.
 
user 37518 said:
As always, you offer great advice, im aware that the image is very asymmetrical signal, which is why im posting it. Something that still no one has mentioned is what do the lower quadrants represent?, like i said i thought they were just a negative copy of the L & R channels, but they are not, clearly visible in the pic i posted which is very assymetrical, would that mean that the lower 2 quadrants represent the valleys of a wave, while the top 2 quadrants represent the peaks of  a wave? in which case that image would mean that the bottom of my wave is somehow chopped off?
The quadrants just correspond to the polarity of the signals. Remember the quadrants here are rotated 45° respective to a real o'scope. The upper quadrant represents stereo signals where both L & R are positive. Just the same, the lower quadrant represents stereo signals where both L & R are negative. The East and West quadrants represent signals where the instantaneous amplitudes of L & R are of different signs.
A mono sinewave will be displayed as a straight North-South line, because it oscillates between positive and negative.
As I mentioned earlier, the graph you show indicates that the signals are noticeably asymmetrical; not necessarily because something wrong happened. A typical violin or guitar have naturally very asymmetrical wave forms. Any signal with high 2nd-harmonic content is asymmetrical.
 
Thanks abbey!, i get it now, yeah i always wondered why trumpet/brass waveforms looked so weird. Regarding the quadrants, i guess i was right, top means positive (peak) and bottom means negative ( valley), and left right means just the same but when both signals are of opposite polarity (am i correct)?

 
user 37518 said:
Thanks abbey!, i get it now, yeah i always wondered why trumpet/brass waveforms looked so weird. Regarding the quadrants, i guess i was right, top means positive (peak) and bottom means negative ( valley),
Adademically, it's not correct; both positive and negative maxima are peaks. Valley doesn't belong to instantaneous amplitude. Some use this word for describing the dips in the envelope of a signal, but it doesn't make much sense, because eventually all signals go to zero.
 
true, mathematically speaking theres no distinction between the top and bottom of a waveform so a valley is just a negative peak, but coloquially, a peak is the top or positive side, and a valley is just the negative peak.
 
One last thing, now that you mentioned it, can you describe a bit more how do you use the Lissajous for THD evaluation?
 

Latest posts

Back
Top