Thanks for pointing that out you are of course right and REW just plotted what I told it toREW is wrong on this.
To display phase one needs to set a reference point in time.
What RWE is doing :
It takes the IR of the Low Pass and finds the peak, uses this as reference to plot the phase of the LP.
Then it uses the peak of the HP´s IR ( which is different ) to plot that phase.
To compare two phase traces, one always needs the same reference time ( delay ).
Why is this easy to spot:
In simulated HP/LP there is no time of flight involved, therefore the phase trace should always be going down ( except for REW)
The phase plot of REW for the LP generates negative Group delay, which is second best to time travel
I swapped the low pass and high pass around for it to give a positive delay number rather than request time travel. I took a shortcut and didn't estimate the IR delay I just made REW find the match point for me.
This is the impulse difference
Then in the alignment tool no delay is necessary, sorry if I confused anyone. The relative differences still apply just not the specific delay numbers. This is another example of why time locked measurements are necessary.
Attachments
Interresting! So in other words, For best results: Active xover for driver aligned and FIR is needed for group delay systems?
Thx Ro808,
I read it, then reread the link ernperkins gave in #208, trying to get a better handle on power response and lobing.
Must admit, the more i read papers like these, the more I get convinced that steep linear phase xovers are the way to go. I wish someone would put a similar paper out on them for a reality check.
And if one wants to control directivity to further manage the off axis response a waveguide is needed. Multiple entry horn is pretty much the only thing that can achieve the goal of 1/4wl spacing and controlled directivity. A true point source
I think so too....there just doesn't appear to be another way to get to 1/4wl spacing, with directivity.
Other than electrostats i guess,
or maybe line or ring arrays using a boatload of various sized drivers.
Synergies sure look like the best alternative to me...getting ready to post another one.
Last edited:
Thank you fluid for the good illustration. This illustrates was exactly what I was trying to describe.
Im my case, I can only do the Phase-shift in steps of ~11us or ~3.66mm, which means that I end up in a combination of the first and second picture.
In order to correct this I tilt the baffle, which corresponds to the third picture, which takes me to a combination of second and third picture in order to make the lobe horizontal.
An extra challenge is that I have two crossovers to align (W-M, and M-T), which makes the tilting a bit more tricky, since both are affected at the same time.
As bentoronto points out it can be discussed how accurate/puristic you need to be when aligning the phases/timing.
In my experience it pays off to spend some time on this, and the reward is a more relaxed sound, but must of course be balanced towards all other imperfections.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.