done with powered subs

[Roc Rinaldi]

Jay,

Yes porting adds group delay around the tuning of the vents/PR's. I know this...also to get the best of both worlds(low group delay and extension).A larger box tuned very deep is needed,tuned below 16 or even below 14Hz. Ideal tuning is low,where group delay effects on SQ are low (in the 25Hz-80Hz range where it is critical to have low group delay forthe all known musicality )and the ears are forgiving of deviations below 18Hz.

OK guys, what the heck is "group delay?"

I think I know what "extension" is.

A larger box sounds like it is a good idea but believe me sometimes a smaller box can be more pleasurable.

[Jay481985]

Jay,

Yes porting adds group delay around the tuning of the vents/PR's. I know this...also to get the best of both worlds(low group delay and extension).A larger box tuned very deep is needed,tuned below 16 or even below 14Hz. Ideal tuning is low,where group delay effects on SQ are low (in the 25Hz-80Hz range where it is critical to have low group delay forthe all known musicality )and the ears are forgiving of deviations below 18Hz.

below 32 hertz your ear really doesn't hear a difference is high or low group delay. Group delay is the wave coming in at different timing to put it in a very very simplistic way. Also how big is the box going to be, seriously dual lms5400 need atleast 2, 6inch flared ports atleast 80 inches deep to tune at 80 hertz

[mas]

Jay,

Yes porting adds group delay around the tuning of the vents/PR's. I know this...also to get the best of both worlds(low group delay and extension).A larger box tuned very deep is needed,tuned below 16 or even below 14Hz. Ideal tuning is low,where group delay effects on SQ are low (in the 25Hz-80Hz range where it is critical to have low group delay forthe all known musicality )and the ears are forgiving of deviations below 18Hz.

below 32 hertz your ear really doesn't hear a difference is high or low group delay. Group delay is the wave coming in at different timing to put it in a very very simplistic way. Also how big is the box going to be, seriously dual lms5400 need atleast 2, 6inch flared ports atleast 80 inches deep to tune at 80 hertz

Nope. Not exactly.

Group delay is "the rate of change of phase". In other words, it is the first derivative of the phase equation... The slope of the phase response curve at a given point.

Phase 'wrap' is a normal occurrence in dynamic drivers. But abrupt changes in the slope of the phase curve - how smoothly the phase is changing - are not normal nor desirable!

Exactly what group delay is and what its effects are does not easily lend itself to a trivial 25 word answer. []

If you need some additional info, PM me and I can send some.

[DrWho]

OK guys, what the heck is "group delay?"

Despite what Mark says, I think I can do it in 25 words or less []

Group delay is a measure of how spread out the group of input signals

becomes. Think of the input to your sub that contains a 16Hz bass

guitar note (achieved through an octave drop) and for the sake of

analogy lets assume that the higher harmonics of that note (the

frequencies that make it sound like a bass guitar) occur at one octave

intervals (so 32Hz, 48Hz, 64Hz, etc...). This will actually sound like a crappy synthesizer, but let's just call it a bass guitar.

When the bass player plucks his 16Hz note, all of the harmonics arrive

at specific intervals related to how the body of the guitar interacts

with the strings and all that crap. For the sake of analogy, lets

just assume that all of the harmonics arrive at the same time. So at time=0,

you have 16, 32, 48, and 64 Hz content. (The analogy works because any time-offsets in the original instrument need to arrive at their specific intervals and if we assume that all of this frequency content arrives at the same time, then we can calculate the same level of offset between that of the system versus that of the desired result.)

Now send this signal into your subwoofer where there is lots of group

delay. In fact, let's say that the group delay is 0 at 48 and 64Hz, 10ms at

32Hz, and 40ms at 16Hz. What comes out is 48 and 64 Hz for the first

10ms until the 32Hz signal joins in, and then finally 30ms after

that the 16Hz signal arrives. Then when the bass player stops the note,

the 48 and 64Hz content stops immediately, the 32Hz takes

another 10ms to finish and then finally 16Hz continues for another 30ms after

that.

So if the original "group" of frequencies lasted 4ms (basically the duration

of the note was 4ms), then the output from the subwoofer is going to have a

duration of 44ms! So if we got a bassist playing a lot of fast notes,

the lower frequency content is going to start piling up on top of the newer high frequency content - making things sound

extremely muddy and weird (muddy from all the extra noise and then

weird from the comb-filtering and timbre shifts that ensue).

Ok, so in 25 words or less:

"The group of frequencies at the input get spread out in time at the

output. Basically, lower frequencies are arriving after the higher

frequencies." []

Though I absolutely despise the terms of "fast" and "slow" - a person

that is describing a subwoofer as "slow" is often hearing the increased

group delay.

So what causes the group delay? Basically, it takes time for the

subwoofer driver to move - and since low frequencies require more

excursion, it takes the driver a longer amount of time to move far

enough to create the output for the lower frequencies. Ported subs are especially bad

because not only does the driver need to move, but the air inside the

cabinet needs to vibrate, which then causes the port to vibrate, which

then creates the output from the port. It takes time for the sound to

travel through the cabinet and out the port which increases the group

delay of the system. And again, more air needs to move for the lower

frequencies. Just keep in mind that the delays associated with the port

are only happening at the tuning frequency...not at frequencies above

the tuning point.

If you want to think of everything in terms of the derivative of phase,

then you simply need to observe that in order for the phase to change,

that there needs to be a delay. Think of phase as defining where on the

sine-wave you're at - you may be reproducing a 10V P-P voltage, but the

signal is always oscillating somewhere inbetween - which is going to be a function

of time and phase. Shifting the phase is the same thing as shifting the

time. I have a feeling this won't make any sense without pictures

though...

You said 25 paragraphs, right? lol

[TheEAR]

Dr WHO,

Very very good description of group delay.

And ....

Your following statement...

"Though I absolutely despise the terms of "fast" and "slow" - a person that is describing a subwoofer as "slow" is often hearing the increased group delay."

Spot on. And part of the lack of musicality in a good few subs. This is why a sealed will always be more musical than a ported.This taking as a given both are of high quality and well designed. Group delay can be pushed in the depths,but there are limits doing so.

[Jay481985]

Dr WHO,

Very very good description of group delay.

And ....

Your following statement...

"Though I absolutely despise the terms of "fast" and "slow" - a person that is describing a subwoofer as "slow" is often hearing the increased group delay."

Spot on. And part of the lack of musicality in a good few subs. This is why a sealed will always be more musical than a ported.This taking as a given both are of high quality and well designed. Group delay can be pushed in the depths,but there are limits doing so.

unless you equalize the sealed subwoofer like a fool that is trying to maximize the excursion which brings group delay. So if you have a sealed enclosure eq to its life, vs a ported that does not dip below the tuning frequency they should be around the same in group delay. And since the vented subwoofer is more efficent, you do not have to move the cone as much (unless it goes below tuning freqency where it can over excursion, but a good designed subwoofer should have high pass filters) so you get less group delay time due to less excursion, but then again you have to account for the port which adds time to the ear. But as DrWho said its all about compromises.

[DrWho]

The group delay will actually be the same for a sealed alignment with EQ that matches the frequency response of a ported system. The only difference is that the ported system requires less cone-excursion and therefore can play louder. It's really not "less" excursion in the sense of the total system though - the ports/passive radiators are actually providing the extra displacement. So ultimately the advantage of porting is that more displacement can be achieved for less money.

In other words, ported is always better than sealed. The only trick is designing the ported system to not sacrifice any of the performance of a sealed system.

One side point: modifying the T/S parameters of a driver to achieve a given frequency response is equivalent to using electronic EQ. Basically, you can think of building the EQ into the driver or building the EQ into the amp as having the same effect. It's all part of the total system and affects the phase response (and thus group delay) in the same way.

So basically, the only way to improve the time-behavior of a system with a given frequency response is to reduce the cone excursion of the system. This can be achieved by increasing the total radiating surface area (more and bigger drivers), or by making the system more efficient - like horn loading. Moving to smaller high-excursion drivers is not the way to go - at least in terms of the time-behavior of the system. I can't deny how good some of these smaller systems sound, but I just know they could sound oh so much better if they weren't so small. I guess the conclusion is that group delay distortion is less annnoying than frequency response abberations or even harmonic distortion (to an extent).

[jdm56]

Thanks, Who. I've never understood what group delay is, and now I do. Thank you very much. The old dog just learned a new trick!

[TheEAR]

My idea of a great subwoofer is around 18cu ft net interlan volume,using dual TC SOunds LMS-5400 18" woofers opposed,vented uning three 6" ports flared to 11".Tuned to 12Hz,powered by a 4KW amp.Cone excursion is within the Xmax down to 11Hz! With 4KW input!

You get outstanding output to 14Hz,low group delay in the audible range(down to 20Hz).With no room gain factored it is linear to below 14Hz @ almost 120dB! And with room gain you add almost 10dB to this,this unit will trounce any commercial sub.I am set on this sub.

I also have my dual TC-3000 15" woofers,again will be mounted in a ported cabinet,tuned to 12Hz again.12 cu ft net volume,venting using dual 6" ports flared to 11",powered by a 2KW amp.gain cone excursion is below the Xmax.

In both designs the porting will be made a la HO/Turbo one part outside the box. As ports will be long.

Both exibit low group delay in the audible range,where group delay skyrockets is down very deep . Where it does not matter much.

[Jay481985]

My idea of a great subwoofer is around 18cu ft net interlan volume,using dual TC SOunds LMS-5400 18" woofers opposed,vented uning three 6" ports flared to 11".Tuned to 12Hz,powered by a 4KW amp.Cone excursion is within the Xmax down to 11Hz! With 4KW input!

You get outstanding output to 14Hz,low group delay in the audible range(down to 20Hz).With no room gain factored it is linear to below 14Hz @ almost 120dB! And with room gain you add almost 10dB to this,this unit will trounce any commercial sub.I am set on this sub.

I also have my dual TC-3000 15" woofers,again will be mounted in a ported cabinet,tuned to 12Hz again.12 cu ft net volume,venting using dual 6" ports flared to 11",powered by a 2KW amp.gain cone excursion is below the Xmax.

In both designs the porting will be made a la HO/Turbo one part outside the box. As ports will be long.

about 60 inches give or take, you can actually make it around 56 if you want to still tune at 14 hertz, the thing is even after the sound/air leaves the port it still acts like there is some porting. Kinda like a bullet coming out of a barrel still is pretty straight before the air hits it. btw if you do it ala ho/turbo you must compensate for the curvature. the curve will reduce air transfer and restrict flow acting like the port is smaller than it really is so you want to make the curve as wide as possible.

Both exibit low group delay in the audible range,where group delay skyrockets is down very deep . Where it does not matter much.

to be honest group delay is not so important 50 hertz down but yes group delay skyrockets below the tuning point. then again we can argue we cannot much hear below 20 hertz so all is basically mute on group delay below 20 hertz

btw I am actually impressed with the port design, I would not have thought it woud model as well