2 days of looking @ transfer traces

Re: 2 days of looking @ transfer traces

Thanx but the real challenge is still coming . Getting the rest of the line of speakers to line up with the phase response of this set-up . By the way this is 1 3 way line array speaker .
Measurement was taken close range in the far field reflections will fuck up this nice response because we do not have a anoich(?) chamber . (english and me but i'm getting better at it)
This set-up did see action a month or so ago . The phase response held his ground at a distance of about 20mt but coherence went down a lot because of a not so good venue acousticly .

I still want to compare a 2 way cab first doing the alignment with delay and after that with allpass filters and phase correction and see what the differance is going to be specialy on the impulse response .
Need to talk to the guy that builds the speakers for doing that .
 
Last edited:
Re: 2 days of looking @ transfer traces

Just the first 1 . After that the array but for setting x-over filters and the rest of the processing just 1 . The grill is off because we changed the original low driver for a other model so i had to redo the first settings .
After that the array but in a small room it's not really a good idea to take that as a reference . If the mic gets to close to the array you start to see multiple speakers specialy in the phase response . Also because off the walls being closer to the array than in a concert hall the reflections are a lot stronger in level and it gets pretty hard to make sence out of the measured response .
The first setting of this array saw action about a month ago (and the alignment stil held it's ground but a lot les coherence) but there were no down fills available at that time .
With this measurement i tried to get as little delay as possible in to the system and stil a nice phase response with not to much processing and we wanted to see if the settings could be used on the down fill .
Also instead of delaying the sub i used a 1st order allpass (so 90dg of phase shift) higher up the frequency scale and that also worked pretty nice . In a large venue with the array way up above i probably would end up using delay again but we will see when it happens .
 

Attachments

  • IMAG0580.jpg
    IMAG0580.jpg
    104.5 KB · Views: 0
Re: 2 days of looking @ transfer traces

So i did some more measuring on a 2 way conv. cab. to investigate what the effect on the Impuls response would be when correcting the phase response of this system .

Phase Correction and the effect on the Impulse response « Timo's World all around the globe go to the powerpoint and press play or if it's in dutch press dia voorstelling starten

Alignment with All-pass filters and the effect it has on the Impulse Response « Timo's World all around the globe

As i expected the impuls response got a lot tighter compared to using delay because most of the frequencies are round a bout the same point in time wich also shows in the phase respons of the systems .
I'll post better data @ my blog when i have better data stored on SIM and/or maybe with SMAART 7 with a mac ás i'm making a dicsision on buying a mac/smaart pretty soon now....
 
Re: 2 days of looking @ transfer traces

Nice job Timo!
anyway i can't see the skydrive things :( don't have permissions

Got some questions... why a different slope for the lows hipass (12db/oct) in relation to the highs lowpass (24 db/oct), that should result in a 180° phase mismatch, with more job to do with the allpass to align all... is this correct?

did you try to align with delay and then fine tune with allpass? do you have screenshots? what approach you think is the best for a nice short IR result?

thx a lot!
luigi.
 
Re: 2 days of looking @ transfer traces

Luigi Chelli said:
Nice job Timo!
anyway i can't see the skydrive things :( don't have permissions

Got some questions... why a different slope for the lows hipass (12db/oct) in relation to the highs lowpass (24 db/oct),
I would even choose a 12 db slope on the high driver if it could take that but i had to use the 24db filter in order not to damage the high driver .
The reason for me to use a 12db hp filter : it gives me a larger overlap area (if aligned properly a larger addition area) and less group delay
that should result in a 180° phase mismatch, with more job to do with the allpass to align all... is this correct?
It should if i would use delay to align the system but with allpass filters i'm only delaying a specific part of the frequencies (the part off the phase slope that has a upwurth angle) to get the angles of the phase slope to match .
If i have to reverse polarity i will but only if i have to because the slope matches with a 180° phase difference .

did you try to align with delay and then fine tune with allpass? do you have screenshots? what approach you think is the best for a nice short IR result?
At the last jobs i did where i had the allpass filters available i tried to use as little delay as possible . On 1 job i had the allpass filters available but ended up using delay but this was due to the high driver being in front of the low & mid drivers of the system (this was a big line array loudspeaker from spain) . Sorry no usable screenshots available because of me not saving everything within sim3 .

The best result for a flat phase/tight impuls response for me is try to avoid using delay .


The other way to get a flat phase system is using FIR filters . This is something i can not do because i'm not able to calculate them and it's something i do not fully understand nor can i controll them . All-pass filters is something i can measure and controll so that's the way for me . Also fir filters will give me more latency in the system specialy if they are at lower frequencies so i do not mess with them .

thx a lot!
luigi.
Click to expand...

With kind regards Timo
 
Top Bottom