[masoman95]

hey, i was wondering what the phase response and group delay graphs mean on my WINisd program, im trying to build a custom box and the amplitude response is flat and everything else is good but i dont know what phase response and group delay mean or do to my subwoofer box, what is the optimum range or area you want these to be in for best sound quality? im a little confused my ported box has more group delay than sealed, and it showing the phase response slope 110deg @ 40hz is this good?

 

[Justin Zazzi]

Group delay is calculated from the phase plot and if you know one, then you know the other one too. Group delay is generally much easier to read since the vertical axis is time measured in (usually) milliseconds, which makes much more sense to most people than degrees of phase.

If you have a phase shift in your system that means some frequencies will arrive at your ear after other frequencies in terms of time. Usually the bass frequencies trail the midrange and treble frequencies but this phenomenon occurs at each crossover too.

There are not many rigorous studies done on the audibility of group delay vs frequency, but the little data that is available suggests a group delay with a duration less than the time it takes for one to two cycles of a frequency to occur is "pretty safe".

So look at your group delay chart at 40hz. One cycle of 40hz will take (1/40)=0.025 or 25 milliseconds. Two cycles at 40hz will take twice as long, so 50ms. If your group delay at 40hz is less than ~25-50ms, then you should be okay.

The "pretty safe" amount of group delay decreases as the frequency increases. Also, more exotic bass enclosures (such as ported) usually have more group delay than simple boxes (such as sealed), like you just learned.

 

[masoman95]

Group delay is calculated from the phase plot and if you know one, then you know the other one too. Group delay is generally much easier to read since the vertical axis is time measured in (usually) milliseconds, which makes much more sense to most people than degrees of phase.

If you have a phase shift in your system that means some frequencies will arrive at your ear after other frequencies in terms of time. Usually the bass frequencies trail the midrange and treble frequencies but this phenomenon occurs at each crossover too.

There are not many rigorous studies done on the audibility of group delay vs frequency, but the little data that is available suggests a group delay with a duration less than the time it takes for one to two cycles of a frequency to occur is "pretty safe".

So look at your group delay chart at 40hz. One cycle of 40hz will take (1/40)=0.025 or 25 milliseconds. Two cycles at 40hz will take twice as long, so 50ms. If your group delay at 40hz is less than ~25-50ms, then you should be okay.

The "pretty safe" amount of group delay decreases as the frequency increases. Also, more exotic bass enclosures (such as ported) usually have more group delay than simple boxes (such as sealed), like you just learned.

ok thanks i understand better, so does that formula work for any frequency? ex: 1/55=0.018ms and by one cycle do you mean like one single note or what do you mean by that? So basically as long as your group delay is less than the 1st and 2nd cycle of that frequency there shouldnt be any audible difference? my group delay at 40hz is 9.5ms, so thats good right? Im using bassbox pro, is the goal basically to get the amplitude response as smooth a slope and as close to 0dB as possible for best quality? my f3 is 28hz and the slope goes no more than 0.45dB above flat line up to 100hz is this ideal?

 

[masoman95]

Group delay is calculated from the phase plot and if you know one, then you know the other one too. Group delay is generally much easier to read since the vertical axis is time measured in (usually) milliseconds, which makes much more sense to most people than degrees of phase.

If you have a phase shift in your system that means some frequencies will arrive at your ear after other frequencies in terms of time. Usually the bass frequencies trail the midrange and treble frequencies but this phenomenon occurs at each crossover too.

There are not many rigorous studies done on the audibility of group delay vs frequency, but the little data that is available suggests a group delay with a duration less than the time it takes for one to two cycles of a frequency to occur is "pretty safe".

So look at your group delay chart at 40hz. One cycle of 40hz will take (1/40)=0.025 or 25 milliseconds. Two cycles at 40hz will take twice as long, so 50ms. If your group delay at 40hz is less than ~25-50ms, then you should be okay.

The "pretty safe" amount of group delay decreases as the frequency increases. Also, more exotic bass enclosures (such as ported) usually have more group delay than simple boxes (such as sealed), like you just learned.

found a really good explanation for group delay and how it affects sound quality, sealed has very little which is why it sounds "tight" and on point with notes, good info https://www.vertexaudio.com/Reading-A-WinISD-Graph-a/269.htm

 

[Justin Zazzi]

ok thanks i understand better, so does that formula work for any frequency? ex: 1/55=0.018ms and by one cycle do you mean like one single note or what do you mean by that? So basically as long as your group delay is less than the 1st and 2nd cycle of that frequency there shouldnt be any audible difference? my group delay at 40hz is 9.5ms, so thats good right? Im using bassbox pro, is the goal basically to get the amplitude response as smooth a slope and as close to 0dB as possible for best quality? my f3 is 28hz and the slope goes no more than 0.45dB above flat line up to 100hz is this ideal?

That formula is not really a formula, but more a back-of-the-napkin rule of thumb kind kind of thing that usually works. It's better than nothing, but not perfectly reliable. So yes you can use it for any frequency, just don't rely 100% on it.

Yes, I mean as long as the group delay for any given frequency is smaller than the time it takes for one cycle of that particular frequency to complete (if you were playing a sine wave at that frequency) then the group delay should not be audible. In theory. Like I mentioned, there have not been very many good listening tests to determine this and people argue about it all the time.

Yes, 9.5ms at 40hz is pretty good.

Using BassBox Pro is pretty straightforward. The goal is to select a woofer and design a box that meets your low frequency requirements. Most people want a flat response down to 20hz, low group delay, minimal hump above 0dB, and a design that the Xmax of the woofer is not exceeded with the power that is planned to be used (ported boxes usually need a sub-sonic filter to help with this). This is a balancing act between enclosure volume, port diameter, and port length.

Some people don't realize at first that subwoofers in a car gain a tremendous boost in the lowest frequencies and a "flat" response all the way to 20hz is not needed. Having an F3 (-3dB point) down to 30hz or even 35hz is sometimes acceptable depending on the kind of music you listen to.

Last, the most important part of the article you linked is:

This higher tuning is what gives ported boxes a bad name as far as their sound quality is concerned, but when ported boxes are designed correctly, this issue doesn’t exist.

There are TONS of fantastic speakers out there (some of the best actually) with a ported design that also sound "tight" and don't have any sound quality issues. Please don't fall into the trap of believing that a sealed box is the only one that can sound "good".