Ask an Acoustic Engineer (me)

[Jdunk54nl]

A lot of that comes from experience and recognizing the shapes, just like rta graphs. You are probably getting good about guessing that you need (as an example) a boost of 3db at 450hz with a narrower q of 5 when you see a dip around 450hz.

The jl tun software target mode can help you get to know phase and all pass filter stuff together. When you mess with a rta target, you can also see how the phase gets changed because of that. It is a very powerful exercise to play in the target mode and see how all of the IIR/all pass stuff impacts the phase response too. You can actually make a target curve fit measured data, and the phase response or the measured and target will match nearly perfetly as well!

[Kb86]

I learn rta on helix, then rta on rew. Now, do i really need to learn either Tun or Smaart? What do u recommend? I would prefer Rew rta and ear though ��

for example, if i have 0 gain st summing zone 2450hz, mean my tweeter and midrange off around 120 degree. What frequency should i play with 1st all pass filter? Thanks

[Jdunk54nl]

I learn rta on helix, then rta on rew. Now, do i really need to learn either Tun or Smaart? What do u recommend? I would prefer Rew rta and ear though ��

for example, if i have 0 gain st summing zone 2450hz, mean my tweeter and midrange off around 120 degree. What frequency should i play with 1st all pass filter? Thanks

Seriously, play with the jl tun target mode. You can have multiple targets and see how they sum and what all pass filters will do. Smaart is great, but for learning this stuff, the target mode on the JL tun software can't be beat.

[Kb86]

Thank you

[Kb86]

Btw,on Tun, can i measure phase with usb microphone? Or do i need a loopback thing? Thanks

[Jdunk54nl]

Btw,on Tun, can i measure phase with usb microphone? Or do i need a loopback thing? Thanks

To be able to measure phase in real time, you need a loop back some how. JL tun uses Smaart as the code behind jl tun (slightly modified though).

REW can measure phase with a usb mic by it is not in real time and you have to use the sweeps and not pink noise. You also have to have everything right to be able to get useable data out of it. It is a real PITA.

[Justin Zazzi]

https://youtu.be/flj5XmyrLyI

This is mosconi Cto explain how to use all pass filter.

i can not understand why at 15:00 he could find the exact frequency 3500hz to do 1st order apf. Could anyone explain for me pls?.

i only know how to see phase relate using RTA (skizer guide). Dont know how to measure phase.
my right midrange and right tweeter does not sum well at xover zone (2460hz), it does not complete cancellation but sum 0db gain. What does it indicate? What frequency should i attempt to try all pass filter? Thank you.

I only watched a little bit, so I'm guessing a little bit how he found the exact frequency. At that point in the video he was looking for an all pass filter which is a phase wrap of 180 degrees. The center frequency of the all pass filter has a phase wrap of exactly half of 180 degrees, which is 90 degrees. I believe he looked at the phase graph and found where the line crossed 90 degrees, and then looked straight down to the x axis to find that frequency. Now most people like to use round whole numbers so it's very likely that the frequency that was chosen by the designer of that equipment was a nice round number that's easy to type like 3,500hz and not something fussy like 3,397hz.

At least, I think this is what was happening in the video.

[Kb86]

Thank for reply, i understand the same logic. My further question is,

Is it possible to determine the approximate frequency value to use first order all pass filter seeing RTA only?, without being able to see phase graph?

for example, my 2450hz sum at +0db gain. Mean the tweeter and midrange is about 120 degree off.
Now, what frequency should i try to use first order all pass filter?
18xx hz? or 3xxx hz?

Hope im not too much confusing you ����*♂️

[Justin Zazzi]

Thank for reply, i understand the same logic. My further question is,

Is it possible to determine the approximate frequency value to use first order all pass filter seeing RTA only?, without being able to see phase graph?

for example, my 2450hz sum at +0db gain. Mean the tweeter and midrange is about 120 degree off.
Now, what frequency should i try to use first order all pass filter?
18xx hz? or 3xxx hz?

Hope im not too much confusing you ����*♂️

If your speakers sum together flat, then what problem are you trying to solve?

[Kb86]

This is my example. my tweeter and midrange 120 degree shift and sum at 2450hz with 0db gain. on what frequency should i attempt to use first order all pass filter? thank you.