Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Multi-Way Conventional loudspeakers with crossovers

rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool
rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 18th November 2019, 02:03 PM   #2851
mark100 is offline mark100  United States
diyAudio Member
 
Join Date: Dec 2010
rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool
Quote:
Originally Posted by fluid View Post
Exactly, if the bandwidth is large enough the two end up being the same.


1) I agree with the phase part and I see the point you are making, but I don't think you are looking at the whole picture. The problem comes when you try to keep the phase flat but let the amplitude fall. You fix one problem and create another. Flat phase without flat amplitude is just the flipside of your idea of adding group delay to a recording from a speaker, two wrongs don't make a right



2) A simple way is to record it and have REW generate an minimum phase version. Valid at the point the microphone was placed.




3) The phase will follow the amplitude response and group delay will be there. Sampling theorum shows that you can recreate the input if the sampling frequency is high enough. If something only produces sound in a limited range increasing the bandwidth won't change the sound, the change will come if you limit the bandwidth below where the sound is being produced. Like in the graphs Byrtt showed earlier.
Good stuff fluid

1) Oh, I also totally agree that you can't have flat phase and magnitude rolloff.
Sorry if i ever gave the impression i linearize phase past flat magnitude response. Heck, I've never used more than 6144 taps (at 48kHz) and currently have dropped down to 4096 taps on all my designs.
The trace below shows the normal phase lag on one of my favorite sub designs, a push-push slot loaded double 18" reflex. Green raw. Blue processed obviously.

2) Natural sound recording and generate transfers?...I didn't know REW could do such? Have you tried it, does it work well?

3) Yeah, i get that as long as the reproduction capability's bandwidth exceeds the natural sound's or recorded sound's bandwidth, all is good. And I get there are components of a natural sound that propagate from different start times, and with different freq spectra.
What i don't get is how that pragmatically matters to reproduction.
It seems to me, we build a speaker for how low we want response to go, and then linearize both mag and phase all the way to where low magnitude rolloff begins.
I'm not sure how you could linearize phase past mag rolloff anyway. Can grillions of taps do that?

I get the sense that maybe we're all close to saying the same thing, once we fully understand each other.......?
Attached Images
File Type: jpg transfer push push mar 28 raw and proc R.jpg (148.1 KB, 261 views)
  Reply With Quote
Old 19th November 2019, 01:40 AM   #2852
fluid is offline fluid  Australia
diyAudio Member
 
Join Date: Jan 2009
Quote:
Originally Posted by mark100 View Post
Good stuff fluid

1) Oh, I also totally agree that you can't have flat phase and magnitude rolloff.
Sorry if i ever gave the impression i linearize phase past flat magnitude response.
Graphs can help a lot because it is easy to misunderstand the terms people are using.

Quote:
Originally Posted by mark100 View Post
2) Natural sound recording and generate transfers?...I didn't know REW could do such? Have you tried it, does it work well?
Sorry if I gave that impression, REW can't record directly, but it should be possible to import an impulse of the sound, or import a file as a frequency response with magnitude and phase, then REW can generate a minimum phase version to see if there is any excess phase. I haven't tried it.

Quote:
Originally Posted by mark100 View Post
What i don't get is how that pragmatically matters to reproduction.
It seems to me, we build a speaker for how low we want response to go, and then linearize both mag and phase all the way to where low magnitude rolloff begins.
I don't know that it does matter if the rolloff is low enough.

Quote:
Originally Posted by mark100 View Post
I'm not sure how you could linearize phase past mag rolloff anyway. Can grillions of taps do that?
Easily and with not many taps if you correct phase only. As an example in rephase on the filters linearization tab you can select a box rolloff to compensate, one of the sealed ones work well for this. Don't change the magnitude just the phase and see how few taps you need to get a perfect match. I tried it with 4096 and it was exactly the same as with 65536. Phase manipulation by itself does not need that many taps. If you use IIR EQ to correct magnitude, you could do almost any phase correction with your 4096 taps.

Quote:
Originally Posted by mark100 View Post
I get the sense that maybe we're all close to saying the same thing, once we fully understand each other.......?
I think so
  Reply With Quote
Old 20th November 2019, 03:46 PM   #2853
mark100 is offline mark100  United States
diyAudio Member
 
Join Date: Dec 2010
rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool
fluid, thx for this ongoing dialog and posts....I've never been able to grasp what BYRTT and wesayso were saying about their low end phase preference till now..

@ BYRTT and wesayso, sorry it took me so long to see what you guys were saying about preferring minimum phase down low. I do too, below flat response.

It just never crossed my mind that anyone would try to linearize phase below the frequency where flat sub response ends.
And that led me to misunderstanding, thinking since you preferred minimum phase over linear phase down low, that you were saying you preferred minimum phase even up into frequencies of flat response. But now I see
  Reply With Quote
Old 20th November 2019, 10:07 PM   #2854
Oabeieo is offline Oabeieo  United States
diyAudio Member
 
Join Date: Jan 2017
Location: Denver
Okay so not to sound like “that guy”

In the case of my box that rolls off at 45hz but has room gain down below 20hz

Do I add the box linearization or not?

Or do I simply just eq flat (whatever the responce is and call it done and leave only the filter linearization?
Thanks in advance
  Reply With Quote
Old 21st November 2019, 02:32 AM   #2855
BYRTT is offline BYRTT  Denmark
diyAudio Member
 
BYRTT's Avatar
 
Join Date: Dec 2009
rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool
Quote:
Originally Posted by Oabeieo View Post
Okay so not to sound like “that guy”

In the case of my box that rolls off at 45hz but has room gain down below 20hz

Do I add the box linearization or not?

Or do I simply just eq flat (whatever the responce is and call it done and leave only the filter linearization?
Thanks in advance
No box linearization for the Q0,83 @45Hz stopband only linerize subs LP filter : )

Here is a block diagram example of 4-way system where i used numbers for your sub because you told those numbers, the rest is for example. As we were around pages back to hinder destructive ripple calculate final target slope per transducer to be a cascade that includes the side band transducers slopes plus stopbands exactly as in block diagram. As you see stopbands are of minimum phase and not linearized so we end up a true system sum minimum phase curve as was it a single point wide band transducer 45Hz(Q0,83)-20kHz(Q0,707).

rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool-o_1a-png

Curves for above block diagram, lows stopband will extend enormous into your car so find out that probably bit ragged curve and smooth it best you can, probably also you need set a highpass in 15-20Hz area to get it musical or use a linkwitz transform to transfer Q0,83 @45Hz up or down to whatever better stopband that blend smoother with the car cabin, all the room gain corrections should be global and of minimum phase.

rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool-o_1b-png
Attached Images
File Type: png O_1a.PNG (22.8 KB, 260 views)
File Type: png O_1b.PNG (45.5 KB, 335 views)

Last edited by BYRTT; 21st November 2019 at 02:38 AM.
  Reply With Quote
Old 21st November 2019, 03:57 AM   #2856
Oabeieo is offline Oabeieo  United States
diyAudio Member
 
Join Date: Jan 2017
Location: Denver
Quote:
Originally Posted by BYRTT View Post
No box linearization for the Q0,83 @45Hz stopband only linerize subs LP filter : )

Here is a block diagram example of 4-way system where i used numbers for your sub because you told those numbers, the rest is for example. As we were around pages back to hinder destructive ripple calculate final target slope per transducer to be a cascade that includes the side band transducers slopes plus stopbands exactly as in block diagram. As you see stopbands are of minimum phase and not linearized so we end up a true system sum minimum phase curve as was it a single point wide band transducer 45Hz(Q0,83)-20kHz(Q0,707).

Click the image to open in full size.

Curves for above block diagram, lows stopband will extend enormous into your car so find out that probably bit ragged curve and smooth it best you can, probably also you need set a highpass in 15-20Hz area to get it musical or use a linkwitz transform to transfer Q0,83 @45Hz up or down to whatever better stopband that blend smoother with the car cabin, all the room gain corrections should be global and of minimum phase.

Click the image to open in full size.

Thank you!!

Aah the light bulb moment finally!

I think I get it....as per pages back I thought I got it, I think it is now making sense especially with what you and fluid and mark100 were saying earlier.

Now I get the cascade! I never thought about the ripple before like that, I’ve always seen it as a artifact that I wasn’t smart enough to solve. I tryed that with an old measurement and it’s stellar, yes It’s starting to make sense. Can’t wait to try this!

Thank you.

Last edited by Oabeieo; 21st November 2019 at 04:00 AM.
  Reply With Quote
Old 21st November 2019, 04:36 AM   #2857
Oabeieo is offline Oabeieo  United States
diyAudio Member
 
Join Date: Jan 2017
Location: Denver
The only thing I think I’ll have to account for is I like the way the stopband sounds on certain drivers rolloff. The blending is desired especially in car to help build a soundstage and it’s components to help hide the location of the drivers and to help promote stage height. Some of these “problems” I think are of benefit in a car where your trying to get the sound off the floor and to be perceived as coming from up higher and farther back than the actual driver locations.

I think I’m going to have to experiment, maybe raise or lower crossover points in areas where I want or don’t want interaction in the stopbands on each driver.

Ultimately I’m going to have to learn this out of the car so I can get the hang of it than try to adopt it and make it work.


Does anyone know how we perceive sound height?What amplitude/phase/reflection characteristics help make sound appear higher up than it actually is. Or what is the mechanism that we hear height.

Links welcome any reading material please.


I basically just fiddle with delays between pairs +/- maybe .3ms and listen for the stage to rise, which has worked quite well , but I’m definitely changing some of the overall impulse in ways that some may think looks like a mess.

It would be really cool if I could learn how to dial the midrange -250hz-3khz to have good height and than talor the other bands around it and get a smoother overall shape.

I think knowing how we hear height is what will help me the most figure this out

I promise last question for at least 4weeks, I’ll be playing with BYRTTs methods for at least a month before I try something else .


Thanks in advance.

Last edited by Oabeieo; 21st November 2019 at 04:39 AM.
  Reply With Quote
Old 21st November 2019, 03:52 PM   #2858
BYRTT is offline BYRTT  Denmark
diyAudio Member
 
BYRTT's Avatar
 
Join Date: Dec 2009
rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool
Hi Oabeieo now in reality principles are easier said than done, 2 way probably a okay workload but 3 or 4 ways are a big mouthfull of work in reality to get the acoustic right with tons of measurements and quality checks verse textbook targets.

Had below target curve laying around : ) and as far i remember its a car cabin target house curve, i hang it on below in a zip folder, but please research if its true use is for car cabin. House curves can be dialed in as a final act using global EQ, but can also be incorparated in per pass band target curves and a tip to create one is in Rephase set the string of all the cascaded filters and save as frd file and when imported in REW you now have a overlay target to live measurement sweeps and if you want to incorate the hanged on target curve to per passband target curve then go on "ALL SPL" tab "Controls" in REW and times to two curves each other (A x B).
Attached Images
File Type: png O_1c.png (7.1 KB, 20 views)
Attached Files
File Type: zip Oabeieo_target.zip (52.8 KB, 4 views)

Last edited by BYRTT; 21st November 2019 at 03:58 PM.
  Reply With Quote
Old 22nd November 2019, 03:15 AM   #2859
Oabeieo is offline Oabeieo  United States
diyAudio Member
 
Join Date: Jan 2017
Location: Denver
Quote:
Originally Posted by BYRTT View Post
Hi Oabeieo now in reality principles are easier said than done, 2 way probably a okay workload but 3 or 4 ways are a big mouthfull of work in reality to get the acoustic right with tons of measurements and quality checks verse textbook targets.

Had below target curve laying around : ) and as far i remember its a car cabin target house curve, i hang it on below in a zip folder, but please research if its true use is for car cabin. House curves can be dialed in as a final act using global EQ, but can also be incorparated in per pass band target curves and a tip to create one is in Rephase set the string of all the cascaded filters and save as frd file and when imported in REW you now have a overlay target to live measurement sweeps and if you want to incorate the hanged on target curve to per passband target curve then go on "ALL SPL" tab "Controls" in REW and times to two curves each other (A x B).

Thanks again BYRTT,

Wow that target is almost identical to my Dirac live target that sounds the best in the car, I’m not not using Dirac live and have grown out of it, I can beat what it can do manually in the car with rephase in most aspects. (I could go on and on why I live Dirac but the things I didn’t like as few as they are were dealbreakers for me and I like the freedom over the impulse so I can break the rules as I wish)

That target is pretty much exactly what I’m using. Especially using horns in the car that play down to 1k that have a two axis flare that Dirac has a hard time resolving.


You guys have no idea how much of a help you are, I super appreciate the time from everyone. My car now is literally a one in 10million cars to listen to. It’s very special and have tons and tons of hours into it. It’s a lot of fun.

Rephase has made it that one step above Dirac that I just couldn’t get with tonality and stage height with kick panel 8s and under dash horns, the stage is above the dash , vocal nicely centered and pushed far back. Thank you rephase!

I tried that crossover setup today, second go at it from pages back this time I got it working a little better. The midrange sounds much better and more in harmony with the bass. Thank you. I like it a lot


As far as sound height , I have scoured the web and can’t find any articles.getting my stage high was trial and error, I want to learn how to go right to it as I switch gear a lot

Last edited by Oabeieo; 22nd November 2019 at 03:20 AM.
  Reply With Quote
Old 22nd November 2019, 04:19 AM   #2860
mitchba is offline mitchba  Canada
diyAudio Member
 
mitchba's Avatar
 
Join Date: Jul 2011
Location: Sunshine Coast
Quote:
Originally Posted by Oabeieo View Post

Does anyone know how we perceive sound height?What amplitude/phase/reflection characteristics help make sound appear higher up than it actually is. Or what is the mechanism that we hear height.

Links welcome any reading material please.

Thanks in advance.
Hello, you may be interested in trying this:

Online LEDR™ Sound Test | Listening Environment Diagnostic Recording Test
  Reply With Quote

Reply


rePhase, a loudspeaker phase linearization, EQ and FIR filtering toolHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
FIR linear phase plugin for MiniDSP? diyjb01 miniDSP 17 9th June 2016 01:35 PM
FIR filter design tool for Loudspeaker magnitude equalization ttmusic Software Tools 3 24th May 2013 08:30 PM
FIR Filtering experiences Olombo PC Based 8 10th February 2013 03:45 PM
AVX based FIR VST, crossover / EQ / DRC and delay KOON3876 PC Based 97 26th November 2012 07:18 AM
Phase EQ using FIR filters Grasso Multi-Way 2 2nd July 2003 10:37 PM


New To Site? Need Help?

All times are GMT. The time now is 09:31 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Copyright ©1999-2020 diyAudio
Wiki