Flattening Phase Response using FIR Filtering via MiniDSP OpenDRC-DA8
ฝัง
- เผยแพร่เมื่อ 28 ส.ค. 2024
- I show how to flatten the phase response of your drivers using FIR filtering. This is possible with the MiniDSP OpenDRC-DA8 Digital Signal Processor/DAC.
![ฮักเทื่อสุดท้าย - บอย ศิริชัย [OFFICIAL MV]](http://i.ytimg.com/vi/-1PoPoQ-B-k/mqdefault.jpg)
hey man your an absolute legend, you've helped me out again, gonna call you up on those solidworks lessons you offered me at some point, need to turn the point source/line source horn I have in my head into an actual plan. i loved the look of your point source horns btw, but sadly it won't work for me because i'm trying to design something for pro audio with allot higher spl goals, synergy horns are nice but i was trying to come up with something that avoids the bandpass element of them. am gonna build your horn for home use at some point though. sorry if i seem a bit overexcited, but really appreciate what your doing for the community :)
alex kirk I do quite a bit for the pro sector. I haven’t done anything synergy related but the summer Rain it totally suitable for high SPL if you use the venerable B&C 8PE21 which provides a stunning 110dB sensitivity @1watt. That driver has a very low QTS for horn loading applications.
What amazing software, thankyou for showing us.
Environ 60 600 000 résultats (0,45 secondes)
Résultats de recherche
Résultat de traduction
Français
Anglais
Very good video
You could make a video on the open DRC DA8, to linearize the amplitude of the response curves, to linearise the phase, to adjust the volumes and the delays .... it would be the tutorial which is missing for the beginners like me :)
Problem with this kind of setup , this Coaxial setup is Wave Guide Vibration , so you have a Horn essentially and its vibrating back and forth . So your physical air for HF driver is not stable, this introduces Frequency instabilities , and it sounds like when you talk trough Ventilator , creates Frequency modulation .
Now this shows how there is no Perfect speaker , everything is a balance . However i can make a DSP to correct for this , but its not cheap . Its an Invention . Few Speaker companies are working on it already , still not patented so if you are interested let me know .
Make sure your frequency response is flat before you try to make phase flat , your make it non minimum phase
Good video tho ,
Minimum phase should look (almost) like an inverted frequency response .
Nicole Denver interesting...I’ll be sure to give it a try. I’m a little disappointed in the OpenDRC-DA8 because it can’t do low crossover points below around 5kHz because of the number of taps. I think MiniDSP should clearly state this on their literature.
Joseph Crowe's DIY Speaker Building
Opendrc can make crossovers at 50hz
Try linearizing your crossover instead of making linear phase crossovers
How many taps does the da8 have and how many can be allocated across channels and is it 48khz or 96khz
I use 2 2x4hds as Dacs and fir crossovers with 1024t on each channel in a 4way
And a Dirac ddrc in front of them on toslink
For eq , I can make all kinds of crossovers with 2x4hd at 96k
I would use a opendrc on each speaker if you want full flexibility, but it’s not needed ,
I think your trying to do too much,
You want mixed phase filters , use regular iir peq for as much as you possibly can, than once your normal peq and iir crossovers are tuned and frequency response is flat and crossovers are behaving themselves (acoustic slopes match electric slopes) and system is sounding very good and your magnitude is flat than measure and add your phase correction only on non minimum phase parts of the phase. If it’s a ring in box or some wired driver behavior than correct in fir, once phase is corrected it dosent matter too much if your crossovers are linear , it’s the sum of all the waves from the speakers on each side that combines and makes a new response that is where fir comes in.
Do nearfield measurements and work your way back to your seat and start gateing and offsetting the ir for time of flight delay to mic.
Or make your crossovers linear , especially the sub. I would use fir on the sub and midbass integration if I was limited on taps or buy more opendrcs and go deep down rabbit hole (but a fun hole to be in) :-)
Nicole Denver wow thanks for the info! This info isn’t available online, at least not that I’ve been able to find.
@@therealandyncrystal i know this is an old comment but if possible, could you explain what you mean by this? "add your phase correction only on non minimum phase parts of the phase."
how did you get to view the phase plot like that, nomilised with the frequency above? I can unwrap it but its not nearly as horizontal as that, mines almost verticle for any of my speakers. Is that generated minimum phase we are looking at? I can't get a similar view for shit on any speaker data I have on REW (subs, midbass, mids or tweets) ??
I found this technique more predictable when using lower q fir filters. Higher q rarely worked as expected. Don’t know why.
Yes, this. I think a q
Phase is considered only on Xover frequency , because of summing , if there is no summing like here is the case , there is no change in response . Your Ear doesnt care about phase , so if the Xover is set to 2Khz , then everything above that is non important , you can play with it but you will hear no change
True, but phase changes the waveform shape. Imagine listening to a square wave synth.
@@manamimnm yes it does , but its not a square wave , bass unit creates sinusoidal pressure medium in front of mid high driver , and then its sounds travels along the bass sound. You can imagine this as Amplitude modulation , and it sounds exactly like that , take a song , amplitude modulate it at say 100hz in some DAW , and you will hear exactly that . If you listen only those speakers you cant detect it easy , but A and B fast switching , you will notice , i did . And beyond that Measuring microphone does not point towards a speaker but its tangent or 90 degrees , in this case vertical . Sound crosses the capsule and changes pressure in it , while if you point it at the speaker , sound will enter in and cause phase shifts and all kinds or reflections ,and stuff , it makes it even horizontality but less . But at least measuring instrument is used , i promote that , because Sound is similar to Visual Art , if you dont know what to look for you dont see it or appreciate it , with sound its more pronounced since you cant see it , effectively you are enable to hear stuff if you dont know what to listen for , and when this is done , its done with 1 speaker , always only 1 , even if you have a stereo pair . You test stereo Pair only to see the sound at listening position , for other measurements always one , always in the center of the room because reflections are of lowest volume there , if you want to know more about your speaker . I have measuring gear , and time and time again , i fail , tests show me one speaker is better then the other when i though the opposite but only after careful examination , things fall in to place . For example i have one philips speaker , dull sounding , i was thinking its something wrong , it dull sounding . But later when i saw the measurements , i saw that is the one that is the flattest or most true to the signal that was sent to the amplifier .
Hi, great tutorial on how to use rephrase and the open Drc. But I'm not sure if the application example was that helpful. Would it be better to show an example of flattening the 180deg phase rotation in a LR2 crossover between a tweeter and woofer? Thank you!
I tried for a short while to phase correct crossovers and to actually create fir crossovers....but in my experience there are nowhere near enough taps available in the minidsp to do that.
@@plcamp1 Can you just do the phase? and use the xover in dsp to save taps
How do I measure tweeter phase on a 2 way passive speaker? If I measure close up will the woofer mess with the measurement? Also should I do the sweep from dead center at the crossover point??
TURN off physically woofer :)))