Thanks for attaching the PDF - that was helpful in figuring out which paper you were referring to.
What they did to make "the effective delay negligible" is to use FIR-based processing to implement a minimum phase filter.
If you look at the included graphs, you will see that there are still phase shifts present in the EQ curve.
It's when they start linearising the phase curve that an overall delay penalty is incurred.
I'm looking at pages 14-16 here.
Further down, they actually look at FIR filtering for subwoofers (bottom of P19), where they note the excessively large time delays required to linearise the phase of a subwoofer, rendering it likely useless for live sound.
It looks like you've misinterpretted my offer of measurements. What I meant was that we can compare the phase curves once the frequency responses match, using minimum-phase EQ.
We all know that ported boxes will get louder than sealed. I'm saying that, once the frequency responses match, they'll have the same time-domain performance, too.
Chris
Sure, can do. Here's one of the articles: Spec Wars: Looking Inside Loudspeaker SPL Specifications - ProSoundWeb
Chris
Chris
Hi Chris, you know my motto... 'it takes time to fix time' 
Having played with FIR for both playback and live for several years now, looking for any technique to flatten phase & reduce group delay as low as possible in frequency.....well, all roads end up arriving back at 'it takes time to fix time'
Heck, I even thought i could cheat the time devil by using sidechain limiting to invoke a sub hpf, that simply limited overexcusion below the needed high pass frequency.
It didn't do anything but engage limiting below f3, staged off increasing drive levels.
And damn, once limiting kicked in, phase/group delay looked just like it does when a hpf is in place ! Of course.....
That did end up being a neat trick though, to keep flat phase and eliminate group delay at lower drive levels. Very hi-fi at hi-fi levels, and PA at PA levels
For live, I often use the technique linked in the Eclipse Audio paper........
Like you said, using FIR to replicate minimum phase.
And moving impulse centering to the front of the FIR filter. Or maybe it's better to say move the filter delay, which is normally centered for linear phase, to the front of the filter.
Sometimes for live, I simply use IIR processing for the subs...
Having played with FIR for both playback and live for several years now, looking for any technique to flatten phase & reduce group delay as low as possible in frequency.....well, all roads end up arriving back at 'it takes time to fix time'
Heck, I even thought i could cheat the time devil by using sidechain limiting to invoke a sub hpf, that simply limited overexcusion below the needed high pass frequency.
It didn't do anything but engage limiting below f3, staged off increasing drive levels.
And damn, once limiting kicked in, phase/group delay looked just like it does when a hpf is in place ! Of course.....
That did end up being a neat trick though, to keep flat phase and eliminate group delay at lower drive levels. Very hi-fi at hi-fi levels, and PA at PA levels
For live, I often use the technique linked in the Eclipse Audio paper........
Like you said, using FIR to replicate minimum phase.
And moving impulse centering to the front of the FIR filter. Or maybe it's better to say move the filter delay, which is normally centered for linear phase, to the front of the filter.
Sometimes for live, I simply use IIR processing for the subs...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.