Hi, I want to open a discussion if group-delay is audible under 100Hz.
I doubt it is audible because we can not locate a subwoofer under approximately 200Hz.
As an example if bass travels 6 meter we also have a delay of 18 msec. And we would be able to locate the subwoofer when we could hear that.
Because the group-delay is the result of a resonant system the delay increases when it becomes more resonant. I think when there is poor damping of this resonance the group-delay is audible. Not that we can qualify the delay time it self.
Groupdelay of a closed box and a bass-reflex that is more resonant.
There is a rule of the thum that the delaytime in msec X frequency = <400 the delay isn't audible.
So what is your opinion about that?
I doubt it is audible because we can not locate a subwoofer under approximately 200Hz.
As an example if bass travels 6 meter we also have a delay of 18 msec. And we would be able to locate the subwoofer when we could hear that.
Because the group-delay is the result of a resonant system the delay increases when it becomes more resonant. I think when there is poor damping of this resonance the group-delay is audible. Not that we can qualify the delay time it self.
Groupdelay of a closed box and a bass-reflex that is more resonant.
An externally hosted image should be here but it was not working when we last tested it.
There is a rule of the thum that the delaytime in msec X frequency = <400 the delay isn't audible.
So what is your opinion about that?
Last edited:
Hi,
Group delay is an inevitable consequence of a speakers bass roll-off,
and asking if its audible makes no sense, as of course a speakers
bass roll-off is clearly audible in your room, and modified by your
room, which will effect the real acoustic group delay.
A very misunderstood parameter, that cannot be separated from
the actual response, but can be used to talk about the response.
Maximally flat vented responses are poor IMO in real rooms, in terms
of group delay I could say the peaking of group delay is bad but
that is really only obfuscating the real issues. BTW peaking is bad.
rgds, sreten.
You can hear peaks of group delay below 100Hz.
Group delay is an inevitable consequence of a speakers bass roll-off,
and asking if its audible makes no sense, as of course a speakers
bass roll-off is clearly audible in your room, and modified by your
room, which will effect the real acoustic group delay.
A very misunderstood parameter, that cannot be separated from
the actual response, but can be used to talk about the response.
Maximally flat vented responses are poor IMO in real rooms, in terms
of group delay I could say the peaking of group delay is bad but
that is really only obfuscating the real issues. BTW peaking is bad.
rgds, sreten.
You can hear peaks of group delay below 100Hz.
Last edited:
That should be under 80hz. According to test done by THX, as you go above 80hz, it becomes easier and easier to locate the sub. Hence why they chose 80hz as their standard crossover point for THX components.
We did a double blind test here a year or so ago with an allpass filter in the midrange- for me, it made the image of an acoustic guitar seem slightly disjointed (for lack of a better word). You can easily find out for yourself- using a music editing program like Goldwave, you can put in whatever group delay you like at whatever frequencies you like without changing the frequency response, then compare before and after.
Hi,
at first glance the group delay seems like a measure of how precise and tight low bass may sound. We tested and measured different solutions in a anechoic chamber. Starting from Dipoles with <6ms over CB ~6-10ms going to >>10ms with BR and BPs, presicision becomes lower and the well known boominess comes into play. Now the amplitude responses still differed. A Bassboost for Dipole and CB , equalized to similar amplitude response as BR would introduce group delay of similar dimension. Still though the equalized Dipole as well as the CB sounded cleaner and more natural than BR and BP. This finding would disclude the group delay as the deciding parameter, if there weren´t a difference in ´electrical´ group delay and ´mechanical´ group delay. I haven´t heard of or seen a paper discussing any possible audible differences between electrical and mechanical group delay.
So I´d rather dismiss GD as parameter.
jauu
Calvin
at first glance the group delay seems like a measure of how precise and tight low bass may sound. We tested and measured different solutions in a anechoic chamber. Starting from Dipoles with <6ms over CB ~6-10ms going to >>10ms with BR and BPs, presicision becomes lower and the well known boominess comes into play. Now the amplitude responses still differed. A Bassboost for Dipole and CB , equalized to similar amplitude response as BR would introduce group delay of similar dimension. Still though the equalized Dipole as well as the CB sounded cleaner and more natural than BR and BP. This finding would disclude the group delay as the deciding parameter, if there weren´t a difference in ´electrical´ group delay and ´mechanical´ group delay. I haven´t heard of or seen a paper discussing any possible audible differences between electrical and mechanical group delay.
So I´d rather dismiss GD as parameter.
jauu
Calvin
Hi,
Group delay is not an independent parameter you can adjust,
its a consequence of the bass alignment you use and just
one of the ways of describing the alignments properties.
It easy to hear different bass alignments, thus different GD.
What is not easy is discussing GD correctly in context.
rgds, sreten.
Helmuth, have you seen this ? : Discussion of Group Delay in Loudspeakers
I think GD greater than 10ms below 100Hz is audible.
I think GD greater than 10ms below 100Hz is audible.
Hi Jerome,
I never saw this discussion, it doesn't answer the problem either. They also question and guess what is going on.
As one stated one wave length delay is the max, in simulation the delay doesn't become longer than a wave length.
One can turn it around and say we make a speaker with out delay, so we do not have to answer the question what is allowable. How do we do this.
Only way I know is dipole or closed box to make delay low as possible.
Last edited:
I always go for the SIMPLE answer. 
High inductance on the woofer produces a slow deep bass which is almost like built-in bafflestep compensation. Low Inductance woofers produce fast lighter bass.
Similarly, reflex cabinets produce a feeling of deeper bass, but it lags the midrange and lacks impact on drums. Closed box with low inductance woofers close to a wall is fastest and can have considerable slam but at the price of deep notes.
In general, steep rolloff creates energy storage and delay.
Where it gets complicated is high-inductance woofers are a bitch to filter in the midrange. Trade-offs.
High inductance on the woofer produces a slow deep bass which is almost like built-in bafflestep compensation. Low Inductance woofers produce fast lighter bass.
Similarly, reflex cabinets produce a feeling of deeper bass, but it lags the midrange and lacks impact on drums. Closed box with low inductance woofers close to a wall is fastest and can have considerable slam but at the price of deep notes.
In general, steep rolloff creates energy storage and delay.
Where it gets complicated is high-inductance woofers are a bitch to filter in the midrange. Trade-offs.
Hi Helmut, yes but there are exceptions
Try a sim with a ER18RNX, a BR sim (17L@45Hz) with a nice group delay
I made the box, didn't sound like BR, very fast.
Keep the 10ms in mind, it is something which is admitted by speaker builders. But i think to have low GD is not enough to have precise bass.
interresting
If you compare :
D1001-04 L26ROY
D1004-04 L26RO4Y
I have the explanation of the difference on these woofers.
I think this subject is not so simple, i have some questions, but not about group delay but about speed and bass articulations.
How fast a woofer should be in the bass to reproduce fast transient, to be well defined and articulate ?
Is group delay is a parameter ? Is intermodulation distortion is a parameter ? Is storage energy is one ? I don't have enough experience on bass drivers/baffle to answer these questions
Ooh, new DESIGN series from SEAS.
Looks like they make a High Inductance Metal cone for the Active fans AND a LOW Inductance Metal Cone for the passive crossover fans who add a bafflestep coil around 3mH.
Lighter cones are faster too. Didn't I mention that?
Looks like they make a High Inductance Metal cone for the Active fans AND a LOW Inductance Metal Cone for the passive crossover fans who add a bafflestep coil around 3mH.
Lighter cones are faster too. Didn't I mention that?
Jerome,Hi Helmut, yes but there are exceptions
Try a sim with a ER18RNX, a BR sim (17L@45Hz) with a nice group delay
I made the box, didn't sound like BR, very fast.
Keep the 10ms in mind, it is something which is admitted by speaker builders. But i think to have low GD is not enough to have precise bass.
This kind of BR tuning of the TS parameter I also do see at monacor bassdrivers. It hold the middle between Normal BR and Closed box tuning of the groupdelay.
The GD only rises with higher BR tuning, then you are on the same point as normal BR drivers.
Helmut, I would go further to audibility of group delay
Yes it is audible if you have a huge peak in the group delay. But we want to know the limits. Calvin has gave figures : 6-10ms, higher than 10ms it is not good. I am agree with that.
But the debate is far from finished !
Some People don't like resonant bass because you have a lot of stored energy and cannot be right to them. Can we say resonant bass (BR, TL etc.) is not good ? There are at least 80% of speakers !
Group delay :
Audible means good bass ? Sure no !
Not audible means good bass ? I am not sure.
We have the drivers : a good driver for bass, TS parameter, baffle, electrical parameter etc.
Make good bass is not so easy... the room plays a role but don't say if you have the good condition to make good and high fidelity bass.
Perhaps it is not what you want to explore ?
Yes it is audible if you have a huge peak in the group delay. But we want to know the limits. Calvin has gave figures : 6-10ms, higher than 10ms it is not good. I am agree with that.
But the debate is far from finished !
Some People don't like resonant bass because you have a lot of stored energy and cannot be right to them. Can we say resonant bass (BR, TL etc.) is not good ? There are at least 80% of speakers !
Group delay :
Audible means good bass ? Sure no !
Not audible means good bass ? I am not sure.
We have the drivers : a good driver for bass, TS parameter, baffle, electrical parameter etc.
Make good bass is not so easy... the room plays a role but don't say if you have the good condition to make good and high fidelity bass.
Perhaps it is not what you want to explore ?
Hi,
That is the sort of statement that makes discussing GD impossible.
Take a standard 4th order alignment for my standard default WinISD
driver, Fs = 40Hz, Vas = 58L, Qts = 0.43 you have to up Fs to 75Hz
to get 10mS peak group delay which is an awful bass alignment.
It peaks at 10mS @ 65Hz, which is also the -3dB point - rubbish.
The standard driver peaks at 19mS @ 36Hz - which is far better.
Even better is a good overdamped alignment, changing the
port from 38Hz to 28Hz, GD peaks at 27.5mS @ 22Hz.
At 65Hz standard and overdamped have 5ms and 4mS GD.
The overdamped has around 10mS GD @ 36Hz.
GD numbers inevitably increase with the bass extension, and clueless
comments regarding it as a useful number on its own are very wrong.
rgds, sreten.
Last edited:
Obviously not. It's being discussed right here, isn't it?
Perhaps stating GD in periods instead of time might be better?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.