Tannoy DC and phase response.

[ginetto61]

In a swept sine test the waveguide (woofer) would be stationary during the tweeter portion of the sweep.
Under real-world conditions, the modulation/action of the waveguide would likely have more effect on the THD measures than phase measures

Hi and thanks again for the kind help
So i wonder to what this phase depends in the end
I guess that wideband drivers and panels have a very good phase response
But which would be a multi-way with a very good response ?
Just to keep it as a reference i mean
This phase thing i have problem to understand rightly
maybe for this intrigues me more than other issues.
a FR is easy to understand even for an horse ... but the phase is tricky
Thanks again, gino

 

[speedyhamster]

As the Treble unit is wired backwards the sound wave goes negative ie downwards first, and by the time it has traveled down the horn and meets the bass unit the sound wave is going upwards and in phase with the bass unit. (the same reason active sub-woofers have phase controls so you can match to the main speakers.

 

[ginetto61]

As the Treble unit is wired backwards the sound wave goes negative ie downwards first, and by the time it has traveled down the horn and meets the bass unit the sound wave is going upwards and in phase with the bass unit. (the same reason active sub-woofers have phase controls so you can match to the main speakers.

Hi and thanks a lot for the very helpful advice.
In the meantime i read that maybe i am overestimating the issue of time alignment in the sense that if the delay between the drivers is within few ms the ear is not able to feel this misalignment.
I do not know which is the minimum value for delay that can be felt anyway.
For instance taken from a Stereophile magazine test

Vivid Audio Giya G3 loudspeaker Measurements | Stereophile.com

this speaker here is said to have a perfectly "coherent" sound (i.e. aligned)

In this case by the way all the drivers look to be wired in phase ?

I still think that the "coherence" of a single driver speaker like the Quad ESL63 is unbeatable.
All the frequencies come from the same diaphragm ...
I like this idea of coherence, even if i understand that a wideband driver is a real technical challenge indeed.
Another solution maybe is the use of a digital x-over where it is possible to set the drivers delay finely. This would be my best choice for a DC driver as well.
Thanks a lot again and best regards, gino

 

[AndrewT]

In 1ms a sound wave in (listening room) air travels about 340mm.
I guarantee that you will hear the effect of delaying or advancing one driver by 340mm, behind or in front of it's other matching drivers.

I suspect you will hear a 34mm (100us) of relative delay advance.
Good time alignment probably gets to <30us and maybe a lot less.

 

[ginetto61]

In 1ms a sound wave in (listening room) air travels about 340mm.
I guarantee that you will hear the effect of delaying or advancing one driver by 340mm, behind or in front of it's other matching drivers.

Hi and thank you very much indeed for your valuable advice.
I think that this is indeed the point ... i did not know this.
But why the speaker above has been very well judged everywhere ?
I still think that perfect time alignment like in the panel speakers give a kind of very nice coherence to the sound.
But reports from listening tests say differently ... even obviously not time aligned speakers are well judged

I suspect you will hear a 34mm (100us) of relative delay advance.
Good time alignment probably gets to <30us and maybe a lot less.

... i think that this kind of alignment is only achievable with planar or single driver speakers ... is that low the threshold ?
This changes completely my point of view on the issue.
Do you think that time alignment is so important ?
Thanks a lot again, gino

 

[speaker dave]

In 1ms a sound wave in (listening room) air travels about 340mm.
I guarantee that you will hear the effect of delaying or advancing one driver by 340mm, behind or in front of it's other matching drivers.

I suspect you will hear a 34mm (100us) of relative delay advance.
Good time alignment probably gets to <30us and maybe a lot less.

Are we talking about frequency response changes from shifting one unit relative to another? Or are we talking about the audibility of delay only, as seen as in all-pass filters?

What do you base your 30us criterion on?

Opinion or fact?

 

[AndrewT]

Are we talking about frequency response changes from shifting one unit relative to another?

The audible effect of moving one driver relative to the other/s

Or are we talking about the audibility of delay only, as seen as in all-pass filters?

I know next to nothing about all-pass filters

What do you base your 30us criterion on?

Opinion or fact?

my statement says

I suspect you will hear a 34mm (100us) of relative delay advance.
Good time alignment probably gets to <30us and maybe a lot less.

It is clear to me that I am NOT stating FACT. I am giving my suspicion based on errors of around 10mm to 15mm (or about a third of 34mm) in apparent source location.

BTW,
I missed an "or" between delay advance and it should read "I suspect you will hear a 34mm (100us) of relative delay or advance."

 

[speaker dave]

Thanks for confirming that you were stating opinion rather than fact.

 

[AndrewT]

My statement diod not claim nor show FACTS.

That was clear.

Now you as the professional in speaker designing, should be able to give us the facts.

What distance errors in apparent source locations are required to start hearing the effects of time alignment?

 

[ginetto61]

... What distance errors in apparent source locations are required to start hearing the effects of time alignment?

Hi and i remember i have been told a method to assess this.
Two speakers, one L and one R.
Then the same exact signal impulse sent to both speaker at the same exact time.
Then starting to delay one channel.
Up to some ms no differences are heard. Then the delayed signal sounds weaker. Increasing further the delay the echo effect jumps in.
I think this could be a very good way to assess the time disalignment impact on overall sound.
I do not remember the 1st threshold after which the delayed signal sounding weaker by the way.
Maybe 2 ms ??? i do not remember.
But with a digital dsp the test can be easily done i guess.
Thanks again, gino