We are just discussing technical details any way right? Like said, if you are using good approach, group delay issues avoided
The conversation got thick when and every time I described what I now know to be start up transient and how group delay effects it. Its obvious phase is apart of group delay but it seems no one here fully understands the amplitude distortion. The only ones to speak confidently around the topic is Dr Geddes and Weltersys....
Its not a big deal right? We have the thresholds of GD to help provide limits.
I described the start up transient and the distortion of amplitude and people start reciting what they've read about phase, a completely different topic which leads to me to believe certain people have a hard time saying "I dont know"
If you describe group delay without the aspect of amplitude, you do not guve a full definition. If you dont know the amplitude part of the equation, thats ooookkkkkk
https://people.eecs.berkeley.edu/~lewis/Ken3.pdf
I did not see the term group delay in this article....so it must be that the transient effects that follow group delay cannot be apart pf group delay, rather the affect, follows/traces GD. Which I highly doubt, since peak energy follows/traces group delay as well....the amplitude characteristics follow time to peak energy...which is based on ms, just like GD....making the ms of GD more relevant than what is being recognized
While some here pretend to know...Google makes short work of it
Group delay is the time delay of the amplitude envelopes....wow...how many times have I said that, and told that I dont know what Im talking about...I need to study more bla bla bla...go back through this thread...you won't see anyone tie group delay to amplitude but me...
The conversation got thick when and every time I described what I now know to be start up transient and how group delay effects it. Its obvious phase is apart of group delay but it seems no one here fully understands the amplitude distortion. The only ones to speak confidently around the topic is Dr Geddes and Weltersys....
Its not a big deal right? We have the thresholds of GD to help provide limits.
I described the start up transient and the distortion of amplitude and people start reciting what they've read about phase, a completely different topic which leads to me to believe certain people have a hard time saying "I dont know"
If you describe group delay without the aspect of amplitude, you do not guve a full definition. If you dont know the amplitude part of the equation, thats ooookkkkkk
https://people.eecs.berkeley.edu/~lewis/Ken3.pdf
I did not see the term group delay in this article....so it must be that the transient effects that follow group delay cannot be apart pf group delay, rather the affect, follows/traces GD. Which I highly doubt, since peak energy follows/traces group delay as well....the amplitude characteristics follow time to peak energy...which is based on ms, just like GD....making the ms of GD more relevant than what is being recognized
While some here pretend to know...Google makes short work of it
Group delay is the time delay of the amplitude envelopes....wow...how many times have I said that, and told that I dont know what Im talking about...I need to study more bla bla bla...go back through this thread...you won't see anyone tie group delay to amplitude but me...
Attachments
Last edited:
I feel like I have woken up in a parallel Universe where the people who have given the time to try and help someone are the problem.
Yes and when you add in the rest of the quoted line I can't see why you think this is some kind of revelation.
"In signal processing, group delay is the time delay of the amplitude envelopes of the various sinusoidal components of a signal through a device under test, and is a function of frequency for each component".
Group delay and phase delay - Wikipedia
It's good to post the source of the information to make it easier for others to see it in context. I'm not sure where your image comes from specifically.
The video where the graphic in the background comes from is here with a lot of maths to explain just the same processes.
Group Delay and Phase Delay - YouTube
This is not a different explanation than all the previous attempts to help you understand it.
It's equally hard for some to say "Sorry I see what you mean now" as it is for some to admit they don't know.
I may not always be right but I only post information that I believe is correct. If I'm not sure then I will stay out of it as it won't be long before Earl, David McBean or someone else comes along to provide the answer with some authority.
"In signal processing, group delay is the time delay of the amplitude envelopes of the various sinusoidal components of a signal through a device under test, and is a function of frequency for each component".
Group delay and phase delay - Wikipedia
It's good to post the source of the information to make it easier for others to see it in context. I'm not sure where your image comes from specifically.
The video where the graphic in the background comes from is here with a lot of maths to explain just the same processes.
Group Delay and Phase Delay - YouTube
This is not a different explanation than all the previous attempts to help you understand it.
It's equally hard for some to say "Sorry I see what you mean now" as it is for some to admit they don't know.
I may not always be right but I only post information that I believe is correct. If I'm not sure then I will stay out of it as it won't be long before Earl, David McBean or someone else comes along to provide the answer with some authority.
Its not a revelation. Theres just people who cant accept that Group delay is the time delay of the amplitude envelopes.
This is the first paragraph of Merlijn VanVeen's article on group delay that mark100 posted at the beginning of the discussion. Look familiar?
"If you lookup group delay on Wikipedia, you'll find the following definition:
"Group delay is the time delay of the amplitude envelopes of the various sinusoidal components of a signal through a device under test (DUT)."
Where various can be thought of as assorted or mixed.
The first time I read that, my initial thought was: "Come again?". I mean seriously, what does that even mean?
During the past three years, this long-dreaded article has been slowly taking shape in my mind. Finally, I have found the courage to write it because frankly, I don't know if I even understand it myself. So here goes nothing".
Merlijn van Veen - Group Delay 101
Seems we have now gone full circle
"If you lookup group delay on Wikipedia, you'll find the following definition:
"Group delay is the time delay of the amplitude envelopes of the various sinusoidal components of a signal through a device under test (DUT)."
Where various can be thought of as assorted or mixed.
The first time I read that, my initial thought was: "Come again?". I mean seriously, what does that even mean?
During the past three years, this long-dreaded article has been slowly taking shape in my mind. Finally, I have found the courage to write it because frankly, I don't know if I even understand it myself. So here goes nothing".
Merlijn van Veen - Group Delay 101
Seems we have now gone full circle
Well isn't what Marshall Leach presented in 1989?
https://www.google.com/url?sa=t&sou...IQFnoECA4QAQ&usg=AOvVaw1P1bc8wJ8LkPMnrraBFy_R
And maybe in this ( in french, for once it'll be to you to put an effort to understand in a foreign language 😉 ) you'll recognize the effect of an high pass filter and why Earl used 'corrected' rather than my initial 'masked' in my latest post:
5.3 Comment filtrer les haut-parleurs ? - francis.audio
Scroll down at least to 'quasi linkwitz ordre 3' at least...
https://www.google.com/url?sa=t&sou...IQFnoECA4QAQ&usg=AOvVaw1P1bc8wJ8LkPMnrraBFy_R
And maybe in this ( in french, for once it'll be to you to put an effort to understand in a foreign language 😉 ) you'll recognize the effect of an high pass filter and why Earl used 'corrected' rather than my initial 'masked' in my latest post:
5.3 Comment filtrer les haut-parleurs ? - francis.audio
Scroll down at least to 'quasi linkwitz ordre 3' at least...
Last edited:
How about 1971 an example from Cable TV, with some useful graphics, I'm sure there would be much earlier examples too 🙂 (I meant full circle within the explanations presented within this thread).
https://www.nctatechnicalpapers.com/Paper/1971/1971-envelope-delay-phase-delay-group-delay-chroma-delay-what-does-it-mean-how-is-it-measured-/download
From page 300
From these examples it is obvious that we must retain the relative positions of the signal components or distortions will result. The delay of the various frequency components, also called a group of frequencies, must be equal for distortion free transmission. If this is not the case we speak of group delay distortion or simply group delay. The group of frequencies, if applied to a modulator, will result in a carrier and associated sideband envelopes. The envelope which contains the modulation must be passed distortion free or we speak of envelope delay distortion. The terms envelope delay and group delay are generally used interchangeably.
For the sake of completeness let us also state the accepted definition of envelope delay: "Envelope delay is the rate of change of the phase versus frequency curve." I've bolded the last line as it seems to make the point that the explanations are the same
https://www.nctatechnicalpapers.com/Paper/1971/1971-envelope-delay-phase-delay-group-delay-chroma-delay-what-does-it-mean-how-is-it-measured-/download
From page 300
From these examples it is obvious that we must retain the relative positions of the signal components or distortions will result. The delay of the various frequency components, also called a group of frequencies, must be equal for distortion free transmission. If this is not the case we speak of group delay distortion or simply group delay. The group of frequencies, if applied to a modulator, will result in a carrier and associated sideband envelopes. The envelope which contains the modulation must be passed distortion free or we speak of envelope delay distortion. The terms envelope delay and group delay are generally used interchangeably.
For the sake of completeness let us also state the accepted definition of envelope delay: "Envelope delay is the rate of change of the phase versus frequency curve." I've bolded the last line as it seems to make the point that the explanations are the same
Last edited:
"Envelope delay is the rate of change of the phase versus frequency curve."
"Group delay is the time delay of the amplitude envelopes."
Those two statements are not equal
So these are two different aspects of Group Delay?
"Group delay is the time delay of the amplitude envelopes."
Those two statements are not equal
So these are two different aspects of Group Delay?
No, they are just different words to describe the same thing, this just posted above
"The terms envelope delay and group delay are generally used interchangeably".
Thats not the same thing said twice...
Frequency and phase do not equal amplitude
Both of these statements describe group delay....then these are two different actions of group delay
Last edited:
If the previous posts and references don't help you see they are different words to describe the same thing then I am out of ideas. I know you don't like to be told to study but at least read through those references above as there is more in them that might help.
Black = Envelope delay is the rate of change of the phase versus frequency curve
Round = Group delay is the time delay of the amplitude envelopes
Tire = Group Delay
Yes tires are black and round
Round = Group delay is the time delay of the amplitude envelopes
Tire = Group Delay
Yes tires are black and round
Frankly, it didn't deviate much from the standard here so it didn't even catch my attention 🙂
Last edited:
Camplo, yes the amplitude envelope is affected why not? You can visualize it by having low and high frequency waves (two sine tones is the simplest) playing same time and record it to the DAW and look the waveform. If you then delay the lower tone the envelope changes (shifts in time) because the higher frequencies seem to ride on the lower tone, the lower tone modulates the highs.
Now, it is called distortion and there are million different distortion sources in a electro mechanical playback system and it is impossible to get rid of them. Only thing we can do to "better" is to keep them distortions below audibility.
The audibility thresholds of group delay, especially below 1kHz, seem to be about ~1/2 cycle of particular frequency. If you have 6ms at 222Hz it is over one cycle and audible but this requires some effort to achieve. It seems to be no issue to stay below the audibility threshold with multiway speakers with no problems. < 3ms group delay at 200Hz is easy. It is only a mere detail what the underlying audible issue is, amplitude problem or what ever. You only have to face the group delay at the very low knee of your system where the response falls inevitably towards 0Hz, rotates the phase and is seen in the group delay and amplitude envelope. Just keep it below audibility, or inject the waveform directly to brain somehow to overcome this and many other sources of distortion.
You have big woofers and I believe you can make a good crossover for the system so you won't have any issues with group delay audibility what so ever. If in doubt, slap in FIR processing and make the group delay flat and that is the best anyone can do in terms of group delay in the playback system. Best would be to feed the brain directly skipping the whole analog electro mechano acoustic stuff completely, but there is no patents on this yet I believe 😉
Now, it is called distortion and there are million different distortion sources in a electro mechanical playback system and it is impossible to get rid of them. Only thing we can do to "better" is to keep them distortions below audibility.
The audibility thresholds of group delay, especially below 1kHz, seem to be about ~1/2 cycle of particular frequency. If you have 6ms at 222Hz it is over one cycle and audible but this requires some effort to achieve. It seems to be no issue to stay below the audibility threshold with multiway speakers with no problems. < 3ms group delay at 200Hz is easy. It is only a mere detail what the underlying audible issue is, amplitude problem or what ever. You only have to face the group delay at the very low knee of your system where the response falls inevitably towards 0Hz, rotates the phase and is seen in the group delay and amplitude envelope. Just keep it below audibility, or inject the waveform directly to brain somehow to overcome this and many other sources of distortion.
You have big woofers and I believe you can make a good crossover for the system so you won't have any issues with group delay audibility what so ever. If in doubt, slap in FIR processing and make the group delay flat and that is the best anyone can do in terms of group delay in the playback system. Best would be to feed the brain directly skipping the whole analog electro mechano acoustic stuff completely, but there is no patents on this yet I believe 😉
Last edited:
Why did you actually start to focus on group delay in particular, camplo?
I blame reports like the one below (quite widespread in the community, I'm afraid), that handle the GD as something special. It really didn't help in better understanding of the matter, especially to those not educated in signals and systems.
https://pearl-hifi.com/06_Lit_Archi...eCleach_Jean-Michel/Horn_Shootout_ETF2010.pdf
I blame reports like the one below (quite widespread in the community, I'm afraid), that handle the GD as something special. It really didn't help in better understanding of the matter, especially to those not educated in signals and systems.
https://pearl-hifi.com/06_Lit_Archi...eCleach_Jean-Michel/Horn_Shootout_ETF2010.pdf
By this statement it seems like that you think that the value of amplitude is relevant to phase. Its not. And therefore GD is not involved either as it is a pure aspect of phase. The use of "amplitude envelope" is to define how to identify a phase position fo a signal, not to indicate that the value of amplitude has any relevance. Phase is in the time domain but needs a reference to the frequency domain (level) in order to be defined. If you cant describe the signal (level/amplitude), how could one talk about phase?
Better ask how one can investigate/measure ones system wrt. to if it is minimum phase in the passband or not. Why - because when you once have identified that your system is MF, you can solely concentrate on FR.
//
Modeling in VCAD, building prototype cabs and FIR filters while taking measurements in REW has helped me tremendously to understand the relationship between FR/phase/GD/time of flight on a practical level. REW allows you to unwrap phase at the press of a button, VCAD will show you what physically aligning the acoustical centers in the Z plane will do to your FR, etc. Personally I really think it’s time better spent than pondering issues that just might exist because of my limited theoretical knowledge.