The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Phase Linear EQ with no pre ringing? Equalizer Plugins
Old 8th November 2018
  #31
Lives for gear
 

If someone has finally created a plugin eq that doesn't flat out suck, what's not to like? I remember when the original versions first came out, fun toys! (but of course not transparent sounding)
Old 8th November 2018
  #32
Gear Nut
 

Quote:
Originally Posted by JP__ View Post
Buying an exotic dongle for demoing is a huge bummer unfortunately (still have a early CodeMeter here, but this doesnt seem to work right?)...
Right. Confirmed at the source. Out of their hands they said, due to CodeMeter Libs.
Old 8th November 2018
  #33
Gear Maniac
 

acourate - https://www.audiovero.de/en/acourate.php - seems to have so pre-ringing compensation. it's a room correction system also used by bob katz. it's not a vst, but it's as far as I understand a program to create digital filter.

not sure what pre ring compensation means exactly here and not sure how good it works. have anyone tried it by chance?
Old 9th November 2018
  #34
Lives for gear
 
macc's Avatar
 

Verified Member
Been having this discussion on a mastering group on Facebook. Copy/pasting my own post...

From what I've read Uli saying, Acourate's PRC works by delaying the freq/s related to the pre-ring, usually which are tied to specific problem areas in the room response (sharp spikes in excess phase, etc). But Acourate's correction is mixed phase, so it doesn't matter in terms of 'it's no longer a linear phase correction', as it isn't intended to be. If, however, you used that approach with an eq, the impulse response would no longer be symmetrical = not linear phase any more. Hmmm.

Disclaimer: this post is mostly conjecture based on inference about Acourate's PRC, and my own inferior knowledge and understanding. Hopefully it makes sense though.
Old 9th November 2018
  #35
Quote:
Originally Posted by macc View Post
Been having this discussion on a mastering group on Facebook. Copy/pasting my own post...

From what I've read Uli saying, Acourate's PRC works by delaying the freq/s related to the pre-ring, usually which are tied to specific problem areas in the room response (sharp spikes in excess phase, etc). But Acourate's correction is mixed phase, so it doesn't matter in terms of 'it's no longer a linear phase correction', as it isn't intended to be. If, however, you used that approach with an eq, the impulse response would no longer be symmetrical = not linear phase any more. Hmmm.

Disclaimer: this post is mostly conjecture based on inference about Acourate's PRC, and my own inferior knowledge and understanding. Hopefully it makes sense though.

But acourate is not an EQ in the traditional sense! Its a filter meant to undo the room's own "filter" (the acoustics).

No acoustics problem is lin phase in the first place, so why even use it?!

Again: Given linear phase specification, the filter magnitude (how flat, how sharp, how complicated) directly dictates the impulse response of the filter. There is ABSOLUTELY no choice, no room for creativity in this corner.

Last edited by FabienTDR; 9th November 2018 at 04:04 PM..
Old 12th November 2018
  #36
Lives for gear
 
macc's Avatar
 

Verified Member
Quote:
Originally Posted by FabienTDR View Post
But acourate is not an EQ in the traditional sense! Its a filter meant to undo the room's own "filter" (the acoustics).

No acoustics problem is lin phase in the first place, so why even use it?!
Hi Fabien,

Not sure if I'm understanding your understanding of my post correctly

I wasn't saying that room effects are lin phase, but that Acourate uses FIR filters that can result in pre-ring when addressing certain things (points with wild excess group delay etc). It has a method for addressing that pre-ring which is fine for room correction, but *if* you are looking for a true LP response (as in this eq) then it cannot be used. That is the only point I was making.



Quote:
Again: Given linear phase specification, the filter magnitude (how flat, how sharp, how complicated) directly dictates the impulse response of the filter. There is ABSOLUTELY no choice, no room for creativity in this corner.
Absolutely, I agree 100%. There isn't any room not to disagree as it's a statement of pure fact
Old 12th November 2018
  #37
Just to better illustrate why ripples pre and/or post the IR are to be expected in all useful cases:












Flat passband = Ripples in the impulse response
No ripples in the IR = No passband.


This is a logical consequence of the Fourier transform.

Min phase only produces post ripples,
and lin phase both pre and post ripples.
The examples shown above are all linear phase.

_
Attached Thumbnails
Phase Linear EQ with no pre ringing?-irripple_1.jpg   Phase Linear EQ with no pre ringing?-irripple_3.jpg   Phase Linear EQ with no pre ringing?-irripple_4.jpg   Phase Linear EQ with no pre ringing?-irripple_2.jpg  

Last edited by FabienTDR; 12th November 2018 at 06:53 PM..
Old 15th November 2018
  #38
Lives for gear
 
gyraf's Avatar
Thanks, Fabien.

As always, I really appreciate and enjoy your explanations from this domain..

Jakob E.
Old 20th November 2018
  #39
Quote:
Originally Posted by FabienTDR View Post

No acoustics problem is lin phase in the first place, so why even use it?!
Exactly! I was baffled when I read on the Sonarworks thread most users seem to prefer the linear phase option. why? (rhetorical)

most acoustical problems are mid-low and low end related, the worst place to use linear phase ...
Old 20th November 2018
  #40
Lives for gear
 

Verified Member
Quote:
Originally Posted by mastervargas View Post
Exactly! I was baffled when I read on the Sonarworks thread most users seem to prefer the linear phase option. why? (rhetorical)

most acoustical problems are mid-low and low end related, the worst place to use linear phase ...
I can't get my head round why that would be "wrong"?

Surely the proof is in the pudding?

i.e. what works to get best sound OUT of the studio when the audio is with the public not what is scientifically the first line of inquiry in the studio?
Old 20th November 2018
  #41
Quote:
Originally Posted by Joe_caithness View Post
I can't get my head round why that would be "wrong"?

Surely the proof is in the pudding?

i.e. what works to get best sound OUT of the studio when the audio is with the public not what is scientifically the first line of inquiry in the studio?
No, the proof is in the listening ... but lets not get into the linear vs min phase debate, it's been done to death. Like you say, whatever helps to get you the best results must be the right way ...
Old 24th November 2018
  #42
Lives for gear
 

Quote:
Originally Posted by Joe_caithness View Post
i.e. what works to get best sound OUT of the studio when the audio is with the public not what is scientifically the first line of inquiry in the studio?
Yes.. but who decides on this "best sound"? People who use Sonarworks correction at linearphase immediately fall into a peculiar category. Ever tried running a simple 909 kick through the system when using linear phase (or their god-awful hybrid stuff)? Yeah.. not pretty with that pre-wobble echo thing.

Like Fabien said, it just doesn't make much sense. The science shows WHY it's bad and it doesn't take much of a golden ear to also HEAR why it's bad.

.. and some comments I've got for mentioning that single 909 kick from some people is "No music starts with just a kick!". Right, no EDM track ever started with only a kick for a solid minute.
Old 26th November 2018
  #43
Gear Maniac
 

I'm also still wondering if anyone really like to use linear eqs? Beside the one special case where you process tracks that have a phase relation that must be retained. Like when you eq one track from a multi mic recording or in parallel chains.

I only use them in this special case. I never heard an advantage in other cases. Be it on tracks without a relation to each other or on a bus.
Old 27th November 2018
  #44
Lives for gear
 

Quote:
Originally Posted by AreYouHuman View Post
I'm also still wondering if anyone really like to use linear eqs? Beside the one special case where you process tracks that have a phase relation that must be retained. Like when you eq one track from a multi mic recording or in parallel chains.

I only use them in this special case. I never heard an advantage in other cases. Be it on tracks without a relation to each other or on a bus.
there is other advantage, phase shift add delay resulting in a less tight sound, linear phase preserve "timing " it can result in a more transparent and tighter sound, the problem is pre ringing but sometimes not sure it s worst than these very sharp itb transients , it depends the sound equed ,the personal taste..

just boost or cut with linear phase in the low mids with a large Q you will ear what i mean , the sound stay toghther , it keep the same tightness , imo this + pre ringing can still sound better than phase eq depending the sounds ( mostly if it s not transient heavy or if you don t mind messing the transients )
Old 27th November 2018
  #45
Lives for gear
 

Quote:
Originally Posted by AreYouHuman View Post
I'm also still wondering if anyone really like to use linear eqs? Beside the one special case where you process tracks that have a phase relation that must be retained. Like when you eq one track from a multi mic recording or in parallel chains.

I only use them in this special case. I never heard an advantage in other cases. Be it on tracks without a relation to each other or on a bus.
I've used linear phase EQ every once in a while (mainly DMG Audio or Nugen Audio SEQ1 Mastering version) but always with very wide bells or gentle shelves. Sometimes when I feel I have a fantastic "phase thing" going on, coming either from outboard or using something like Acustica Audio plugins and I feel minimum-phase EQ ruins it, then I always end up doing the final corrections and balancing with linear phase EQ (or FF Pro-MB set to linear phase mode with 6dB/octave crossovers).

Sometimes one can get really "lucky" and get a EDM track banging just right by having the cumulative effect of various processes rotate the phase in such a way that the track is really energetic and feeling "correct". You can usually verify this by a 180 degree phase swap at the end of the chain and notice how the whole track falls apart and the impact feels weak. This means that when in correct main phase there's a bit of magic going on.
Old 28th November 2018
  #46
I have often used RX for filtering unwanted very low frequencies; below 45 Hz on singer/songwriter stuff, for example. Just select the area below 45 Hz and hit "delete".

But I have no idea what RX is doing when I hit "delete". I assume it is some kind of EQ filtering. Can anyone fill me in on what software such as RX does to accomplish this kind of thing? Thanks.
Old 28th November 2018
  #47
Lives for gear
 
Alexey Lukin's Avatar
 

Verified Member
RX is using a linear-phase filter to delete.
Old 28th November 2018
  #48
Gear Nut
 

Well here's a real world situation the LP haters* may not have considered.

Just did a job to taking a really crappy sound VHS video from TV in the 90's (tape of a fashion dance show), overlay it the full frequency range studio original music track, and make it sound like it was all live. In addition to a multitude of other things, I found it beneficial to HPF the video track with a linear phase HPF. In this instance preringing didn't matter a damn. It was more beneficial to me to have variable control of the video HPF without having to continually adjust the phase rotation vs the studio track.
dogma

* that's a joke, sorta
Old 28th November 2018
  #49
Lives for gear
 
exwel's Avatar
Quote:
Originally Posted by Fred_Abstract View Post
there is other advantage, phase shift add delay resulting in a less tight sound, linear phase preserve "timing " it can result in a more transparent and tighter sound, the problem is pre ringing but sometimes not sure it s worst than these very sharp itb transients , it depends the sound equed ,the personal taste..

just boost or cut with linear phase in the low mids with a large Q you will ear what i mean , the sound stay toghther , it keep the same tightness , imo this + pre ringing can still sound better than phase eq depending the sounds ( mostly if it s not transient heavy or if you don t mind messing the transients )
thats also my experience
Old 28th November 2018
  #50
Gear Head
Question for developers

For example, I have Boom signal. Apply linearphase cut or boost. Render. Get oooBooom with pre-ringing caused by length of delay of linearphase . Is there any method to constantly remove preringing before main signal? I.e. always lookahead before main signal start, and remove preringing when IR delay happen.
It is like I load that ooBooom in editor and manual MUTE oo before Boom, so, I get Boom with linearphase eq pocessing without preringing (but manually impossible to eliminate that inside Booom sample). Any ideas to implement lookahead for such auto preringing removing?
Or this method just leads to minimum phase constant post ringing and phase problems, because of constant removing of linear phase pre ringing?

In any case linear phase will be useful for non transient material like atmosphere pads etc, so no attack smearing, no problems for preringing for non-attack/long attack material.
In dmg equilibrium there is reverse minimum phase (free phase tweaking). Is it for reverse sounds? So only preringing for reverse sounds same as opposite, post ringing for non-reverse sounds? So , minimum phase for Boom, and reversed minimum phase for mooB ?
Old 28th November 2018
  #51
I've learned to generally prevent linear phase filters from touching the audible bandwidth. i.e. the 20-20k range, or maybe 40-16k at extremes.

In doubt, very smooth and shallow curves will also keep side-effects low, but the problem here is, the same very smooth curves reproduced with a min-phase filter would also likely yield an almost linear phase. They would also work well in parallel setups.

Given the latency problem, linear phase often simply isn't worth the trouble.




In summary, for all filter types:

- The edgier the filter magnitude, the stronger the side-effects become. No edges, no trouble.
- Side-effects affect transients (short, very wideband events). No transients, no side-effects.
- Side-effects only happen when transients within audible range get affected by the filter. If no audible events being affected by the filter = no side-effects.

When side-effects appear, min phase tends to sound far more natural (more naturally distorted ).
Old 28th November 2018
  #52
Gear Guru
As a fellow once said "The proof isn't in the pudding old boy, it's in the eating of the pudding".........

$990- buys a LOT of pudding.......
Old 29th November 2018
  #53
Lives for gear
 
polybonk's Avatar
So anyone tried the mythical pudding?
Old 29th November 2018
  #54
Lives for gear
 
JP__'s Avatar
 

Verified Member
No easy demo available, unfortunately. It needs a 50€ dongle to try.
Anyone around with a new Codemeter USB (mine is thefirstgeneration which do not work I think)

Some simple measurements from the developer to proof his states wouldbe helpful too, but theres nothing on the website...
Did they even changed theirclaims from "no preringing" to "no audible preringing"?


PS: what I do not understand, really: is it just a port from the old Algorithmix Orange EQ or is it something new based on that tech?
With the old Orange EQ nibody ever talked about the absence of preringing ever, right?
All in all it still looks highly suspect to me. I would love to hear some honest words from Friedemann here (he posted a statement at fb, but unfortunately without any real proofs again).
Old 29th November 2018
  #55
Lives for gear
 
JP__'s Avatar
 

Verified Member
Quote:
Originally Posted by MondsteinRecords View Post
Hi, my first post here.

I know Christoph and he has a really big mind. He invented a lot of unbelievable audio gadgets. He was the master mind behind the research department at Sennheiser.
Well, long story short, they sent me the dongle (another 50 bucks) last week, so I think in 1 or 2 weeks I will get a beta version. I will do a test and we will see if all the promises are true. But the first thing is: If it sounds better than all other EQs on the market, it will shake up the mastering industry.

Thom
Any news, Thom?
Old 29th November 2018
  #56
Quote:
Originally Posted by FabienTDR View Post
I've learned to generally prevent linear phase filters from touching the audible bandwidth. i.e. the 20-20k range, or maybe 40-16k at extremes.

In doubt, very smooth and shallow curves will also keep side-effects low, but the problem here is, the same very smooth curves reproduced with a min-phase filter would also likely yield an almost linear phase. They would also work well in parallel setups.

Given the latency problem, linear phase often simply isn't worth the trouble.
.
not just the trouble. it sounds worse.
boys and girls, listen to Fabien, he knows what he's talking about.
Old 29th November 2018
  #57
Quote:
Originally Posted by polybonk View Post
So anyone tried the mythical pudding?
you need a CmStick USB dongle from them which is $50. I understand you get that $50 back in form of a voucher IF you buy their product. so either way it will cost you money to try it.
Old 29th November 2018
  #58
Lives for gear
 

Verified Member
Quote:
Originally Posted by bmanic View Post
Yes.. but who decides on this "best sound"? People who use Sonarworks correction at linearphase immediately fall into a peculiar category. Ever tried running a simple 909 kick through the system when using linear phase (or their god-awful hybrid stuff)? Yeah.. not pretty with that pre-wobble echo thing.

Like Fabien said, it just doesn't make much sense. The science shows WHY it's bad and it doesn't take much of a golden ear to also HEAR why it's bad.

.. and some comments I've got for mentioning that single 909 kick from some people is "No music starts with just a kick!". Right, no EDM track ever started with only a kick for a solid minute.
Who decides the best sound for my monitoring?

well...

me?
Old 29th November 2018
  #59
Quote:
Originally Posted by JP__ View Post
Any news, Thom?
Soon I will tell you. It‘s still in beta.
Old 29th November 2018
  #60
Gear Maniac
 

Quote:
Originally Posted by Fred_Abstract View Post
there is other advantage, phase shift add delay resulting in a less tight sound, linear phase preserve "timing " it can result in a more transparent and tighter sound, the problem is pre ringing but sometimes not sure it s worst than these very sharp itb transients , it depends the sound equed ,the personal taste..

just boost or cut with linear phase in the low mids with a large Q you will ear what i mean , the sound stay toghther , it keep the same tightness , imo this + pre ringing can still sound better than phase eq depending the sounds ( mostly if it s not transient heavy or if you don t mind messing the transients )
So far I never used linear phase on individual tracks as low latency also better fits to my workflow.

But I just tried the new fabfilter pro-q3 for fun. I tried the same setting with zero phase and linear on hihats. So only a mid and high range test! I can't stand pre-ringing on the bass.

The linear phase sounds indeed a bit tighter and with better room. I can a/b it blind without a problem on headphones. But maybe the pro-q3 has just a mediocre zero phase implementation? Who knows.

My setting were bell/12db at 4300hz, 7.6db gain, Q 1.

But I also almost blindly trust FabienTDR opinion when it comes to technical explanations.

Not sure what this means for me now.
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump