Old 11-26-2014, 12:54 AM   #1
PhaseMaster
Human being with feelings
 
Join Date: Aug 2012
Posts: 21
Default All-Pass Filters

I have used the all-pass filters in ReaEQ to correct phase shift in the HDCD reissue of the King Crimson album, Lizard. (DMG0503)

I have attached a Reaper Project file that contains the ReaEQ settings I used for Lizard. The Reaper Project contains no music. You will have to provide your own copy of the Lizard CD. You should decode the CD before you apply the all-pass filters.

Does ReaEQ provide controls for the amount of phase shift and the Q of each all-pass filter? I could only find controls for center frequency and bandwidth of each all-pass filter. I assume that the gain control is irrelevant with all-pass filters.

Most of the audio CDs in my music collection have different amounts of phase shift in the left and right channels. Can anyone recommend VST all-pass filters that provide independent controls for the left and right channels?
Attached Files
File Type: rpp Lizard-Empty.RPP (2.6 KB, 343 views)
PhaseMaster is offline   Reply With Quote
Old 11-26-2014, 10:18 AM   #2
DVDdoug
Human being with feelings
 
Join Date: Jul 2010
Location: Silicon Valley, CA
Posts: 2,779
Default

Quote:
Most of the audio CDs in my music collection have different amounts of phase shift in the left and right channels. Can anyone recommend VST all-pass filters that provide independent controls for the left and right channels?
How are you measuring phase?

The only reason I'd expect to see phase differences between left & right channels would be for simulated stereo or stereo-widening effects. And, those effects are generally irreversible because we don't know the "formula".

The center-information should be identical and in-phase channel (left & right) and the other information should have mostly random phase relationships.


The Dolby Pro Logic uses left-right phase-differences to steer sound to the rear, but you should only find that in movies (and you can't remove that with an all-pass filter).
DVDdoug is offline   Reply With Quote
Old 11-26-2014, 11:53 AM   #3
serr
Human being with feelings
 
Join Date: Sep 2010
Posts: 12,562
Default

Quote:
Originally Posted by PhaseMaster View Post
I have used the all-pass filters in ReaEQ to correct phase shift in the HDCD reissue of the King Crimson album, Lizard. (DMG0503)

I have attached a Reaper Project file that contains the ReaEQ settings I used for Lizard. The Reaper Project contains no music. You will have to provide your own copy of the Lizard CD. You should decode the CD before you apply the all-pass filters.

Does ReaEQ provide controls for the amount of phase shift and the Q of each all-pass filter? I could only find controls for center frequency and bandwidth of each all-pass filter. I assume that the gain control is irrelevant with all-pass filters.

Most of the audio CDs in my music collection have different amounts of phase shift in the left and right channels. Can anyone recommend VST all-pass filters that provide independent controls for the left and right channels?
So you're saying you believe the tapes (for the CD album transfers in question) were transferred with less than optimal head azimuth alignment. This of course results in one channel being ahead of the other as that channel hits the head first due to the out of adjustment azimuth.

I've certainly seen this from amateur analog tape to digital transfers and done my share of corrections (Nudging one channel back into phase with the other). (Ideally you should dial in the azimuth along with correct speed before it hits the converters.) When faced with a poor transfer, digital correction should be done by nudging the few samples needed to correct the offset and not with a processing plugin like eq.

I haven't really spotted this on commercially released music. I probably haven't gone looking for it either... Although CD's were never really considered an audiophile format and nowadays they're the cheap portable format - so sloppiness doesn't surprise me there.


A way to automate detection of this? Probably not. You need to identify a particular element of the sound that's in both channels and then be able to determine that yes, this bit is indeed supposed to be in phase. In other words, you need to be able to recognize elements that are not supposed to be in phase and not be mislead by that.


And as far as King Crimson Lizard goes, you should treat yourself to the remastered DVDA disc. It has the original stereo mix transferred flat to 24/96. It also has a 5.1 surround remix from the original multitracks by Mr Steve Wilson who did just phenomenal work! Seriously check this out! I'm not normally a fan of remixes years after the fact but this guy is an exception. He has really amazing perception for preserving original idiosyncrasies and intentions while expanding to surround. Seriously check this out. And the remix aside, the 24 bit transfer of the original master will be a stunning upgrade to the chirpy 16 bit CD (or the 20-ish bit HDCD decode) you are finding flaws with now.

Last edited by serr; 11-26-2014 at 02:37 PM.
serr is offline   Reply With Quote
Old 11-26-2014, 01:50 PM   #4
Fergler
Human being with feelings
 
Fergler's Avatar
 
Join Date: Jan 2014
Posts: 5,207
Default

Quote:
Originally Posted by PhaseMaster View Post
Most of the audio CDs in my music collection have different amounts of phase shift in the left and right channels. Can anyone recommend VST all-pass filters that provide independent controls for the left and right channels?
Any plugin will do this if you set the plugin pins in Reaper FX window to left -> left only on one instance and right -> right only on a second instance.
Fergler is offline   Reply With Quote
Old 11-26-2014, 04:45 PM   #5
PhaseMaster
Human being with feelings
 
Join Date: Aug 2012
Posts: 21
Default

Quote:
Originally Posted by DVDdoug View Post
How are you measuring phase?
I measure phase shift with my ears. Phase shift means that different frequencies in the music are delayed by different amounts of time. When phase shift happens, fundamentals and harmonics that should be happening simultaneously are offset by several milliseconds.
PhaseMaster is offline   Reply With Quote
Old 11-26-2014, 10:35 PM   #6
NowoTone
Human being with feelings
 
NowoTone's Avatar
 
Join Date: Mar 2013
Posts: 298
Default

Quote:
Originally Posted by PhaseMaster View Post
Most of the audio CDs in my music collection have different amounts of phase shift in the left and right channels.
Quote:
Originally Posted by PhaseMaster View Post
I measure phase shift with my ears. Phase shift means that different frequencies in the music are delayed by different amounts of time. When phase shift happens, fundamentals and harmonics that should be happening simultaneously are offset by several milliseconds.
Please don't get me wrong, but do you think that is a reliable method? I have a pretty good hifi system and also listen to CDs a lot over my monitors for reference. I haven't really had any audible phase issues apart from one CD. So if most of your audio CDs have that issue, you either have very bad luck with your collection or perhaps something else might be wrong.
__________________
Latest Album made with Reaper: "Growing Old Disgracefully" - Songs for Imaginary Films
Electronica & Psytrance made with Reaper: NowoTone (Artist Profile on Spotify)
NowoTone is offline   Reply With Quote
Old 11-27-2014, 12:10 AM   #7
Sativa
Human being with feelings
 
Join Date: Aug 2014
Posts: 57
Default

Quote:
Originally Posted by PhaseMaster View Post
I have used the all-pass filters in ReaEQ to correct phase shift in the HDCD reissue of the King Crimson album, Lizard. (DMG0503)

I have attached a Reaper Project file that contains the ReaEQ settings I used for Lizard. The Reaper Project contains no music. You will have to provide your own copy of the Lizard CD. You should decode the CD before you apply the all-pass filters.

Does ReaEQ provide controls for the amount of phase shift and the Q of each all-pass filter? I could only find controls for center frequency and bandwidth of each all-pass filter. I assume that the gain control is irrelevant with all-pass filters.

Most of the audio CDs in my music collection have different amounts of phase shift in the left and right channels. Can anyone recommend VST all-pass filters that provide independent controls for the left and right channels?
All-pass filters have nothing to do with left and right phase differences.

http://en.wikipedia.org/wiki/All-pass_filter

For example, an all-pass filter can be used before a compressor so that the compressor is not prematurely activated.

e.g., suppose we have all frequencies aligned in phase at some point. The amplitude of that point will be very large. This could trigger the compressor. If we use an all-pass filter we can "un-align" the phase of each frequency so there is no built up at that point. The magnitude of each frequency is unchanged. The waveform has changed though and it has lower overall amplitude, and hence, as far as compression is concerned, a more smooth behavior.


Of course, it may be entirely possible to use all-pass filters for phase correction between stereo channels. Just, as far as I know, all-pass filters don't really have a Q or phase shift. The whole point is they have a center frequency, which is the frequency that is unchanged. The transfer function then is computed so that the magnitude(not amplitude) is unchanged.

I suppose one could link together to all-pass filters in such a way as to optimize phase relationships between multiple channels. Never really heard about this though.

[I guess I was wrong: http://en.wikipedia.org/wiki/Lattice_phase_equaliser. They have been used for stereo phase correction. Still, not really anything to do with Q]

Last edited by Sativa; 11-27-2014 at 12:29 AM.
Sativa is offline   Reply With Quote
Old 11-27-2014, 01:08 AM   #8
evosilica
Human being with feelings
 
Join Date: Jan 2011
Location: Austria/Europe
Posts: 216
Default

Quote:
Originally Posted by PhaseMaster View Post
I measure phase shift with my ears.
Wow, that's amazing! No wonder you're called PhaseMaster.

Does it bother you a lot that almost any speaker available produces some amount of phase shift? Or do you compensate your files for the phaseshift of your specific set of speakers?

evosilica is offline   Reply With Quote
Old 11-27-2014, 03:06 PM   #9
Fergler
Human being with feelings
 
Fergler's Avatar
 
Join Date: Jan 2014
Posts: 5,207
Default

Quote:
Originally Posted by evosilica View Post
Wow, that's amazing! No wonder you're called PhaseMaster.

Does it bother you a lot that almost any speaker available produces some amount of phase shift? Or do you compensate your files for the phaseshift of your specific set of speakers?

#1 best reason to have headphones around, even if you don't prefer to mix in them.

I doubt he compensates btw, that would be totally the wrong way to go around fixing a problem.
Fergler is offline   Reply With Quote
Old 11-29-2014, 11:32 AM   #10
PhaseMaster
Human being with feelings
 
Join Date: Aug 2012
Posts: 21
Default My Speakers

I use a pleasant sounding mid-fi speaker system that includes its own amplifiers. These speakers were marketed as low end gear. The model is Benwin S-50. It uses single driver satellite speakers and a 6.5 inch woofer. This means I don't have the problems that occur with a crossover in the 2500 Hz range.

Recently another music lover recommended a free VST plugin from Voxengo: MSED. This plugin allows one to invert the polarity of the left and right channels with one click. It also allows one to swap the left and right channels with one click. There are also controls for modifying MS parameters.

When I listen to Joe Gastwirt's remastered versions of Joni Mitchell's albums, MSED is the only DSP I use. However, when I listen to the CD version of Return to Forever's Romantic Warrior, I also use ReaEQ with five all-pass filters.

I still do not fully understand how the all-pass filters work. I understand the center frequency control. I understand that the bandwidth control alters the Q of the all-pass filter. However, I do not understand whether or not the gain control has any effect on the all-pass filter.

When I state that an all-pass filter has adjustable Q, I mean that the steepness of the phase shift slope is adjustable.

Last edited by PhaseMaster; 11-29-2014 at 11:52 AM.
PhaseMaster is offline   Reply With Quote
Old 11-29-2014, 02:24 PM   #11
DVDdoug
Human being with feelings
 
Join Date: Jul 2010
Location: Silicon Valley, CA
Posts: 2,779
Default

Quote:
I have used the all-pass filters in ReaEQ to correct phase shift...

...I measure phase shift with my ears.
I don't believe you are "correcting" phase shift. I think you are simply altering it to get a sound you like. That's OK. Some people like to boost the bass. I like to add a little rear-channel reverb by using a "soundfield" setting on my home theater receiver.

If you have access to a receiver with Dolby Pro Logic, phase-shifting might give some "interesting" results, especially in the "movie mode" or basic "matrix mode". (Pro Logic uses left-right phase differences for surround encoding/decoding.)

I assume you are processing the left & right channels differently? otherwise, an all-pass filter normally has almost no effect on the sound.


But for example, flipping the polarity of one channel causes a "spacey" sound and the bass gets cancelled, whereas flipping both channels together has no effect. Or, if you all-pass one channel but not the other and then combine/blend the channels you'll get EQ-like effects. If you time-delay one channel and then combine them you'll get comb filtering.

Sativa mentioned how the various frequencies will sum-up differently. If you take a typical (compressed) commercial release and all-pass filter it, you will usually increase some peaks (while reducing others) and you have the potential of clipping (distortion). For example if you rip a CD, all-pass filter it and save-it back to 16-bit PCM, chances are it will be clipped (f you don't re-normalize or lower the level).
DVDdoug is offline   Reply With Quote
Old 11-30-2014, 08:32 AM   #12
PhaseMaster
Human being with feelings
 
Join Date: Aug 2012
Posts: 21
Default Time to Reconsider

DVDdoug has given me a lot to think about.

I am still pleased with my DSP settings for Joni Mitchell's studio albums from the 1960s and 1970s. However, I have decided that the album Romantic Warrior by Return to Forever did not benefit from using ReaEQ with five all-pass filters.

I am providing screenshots of the DSP settings I use for Joni Mitchell CDs. I hope that the Reaper community benefits from these examples.
Attached Files
File Type: zip Joni_Mitchell_MSED_Screenshots.zip (773.6 KB, 185 views)
PhaseMaster is offline   Reply With Quote
Old 11-30-2014, 09:14 AM   #13
PhaseMaster
Human being with feelings
 
Join Date: Aug 2012
Posts: 21
Default Absolute Polarity

DVDdoug stated that both channels of a stereo recording should be played back in the same polarity. This is something that everyone agrees on.

However, many people believe that reversing the polarity of both channels at the same time makes a substantial difference in the way music sounds. Here is a link to a reasonably short essay on this subject.

http://www.6moons.com/industryfeatur.../polarity.html

Here is a link to a long article by someone who believes that 92% of CDs play in the wrong polarity on 92% of stereo systems.

http://www.absolutepolarity.com/

I look forward to a spirited debate on this controversial issue.
PhaseMaster is offline   Reply With Quote
Old 11-30-2014, 09:29 AM   #14
whiteaxxxe
Banned
 
Join Date: Jul 2014
Location: United States of Europe, Germany, Mönchengladbach
Posts: 2,047
Default

Quote:
Originally Posted by PhaseMaster View Post
I measure phase shift with my ears. Phase shift means that different frequencies in the music are delayed by different amounts of time. When phase shift happens, fundamentals and harmonics that should be happening simultaneously are offset by several milliseconds.
really? by ear? sorry, I dont believe one word. you correct a King Crimson album by ear??? wow ... hyperselfesteem is not always a good thing. :-((
whiteaxxxe is offline   Reply With Quote
Old 11-30-2014, 11:12 AM   #15
DVDdoug
Human being with feelings
 
Join Date: Jul 2010
Location: Silicon Valley, CA
Posts: 2,779
Default

Quote:
However, many people believe that reversing the polarity of both channels at the same time makes a substantial difference in the way music sounds.
I don't have "golden ears"... I can't hear any difference. From what I understand, certain listeners can hear a difference with certain instruments. So, if there is a difference, I wouldn't call it "substantial".

People believe all kinds of things... Some "audiophiles" claim better speaker cables make a "night and day" difference. 'Till you put them in a blind listening test and then suddenly "something's wrong with the test", or for some reason "all blind listening is invalid", and they can't tell which cable is which.

I believe things like EQ, compression, reverb, different speakers, different microphones, different acoustics, etc., CAN make a substantial difference. Even the volume control can make a difference!



Have you ever listened to white or pink noise, or high-pitch test-tones? Turning your head or moving your head a foot or so makes a difference... Certainly a BIGGER difference than flipping the polarity of both channels.

Have you ever done a Blind ABX Test to see if you are really hearing a difference?

With a full band playing, I'm not sure anybody can hear the difference (in a proper scientific blind listening test). And, maybe the polarity of the guitar is flipped but the vocal is not.

Because filters cause phase shifts, The tweeter is usually "flipped" 180 degrees in a 2-way speaker, and the midrange flipped in a 3-way speaker. This is done to make sure the speakers are in-phase with each other at the crossover point. So, I guess you could say almost everybody is hearing the wrong polarity for part of the frequency spectrum.

Last edited by DVDdoug; 11-30-2014 at 12:50 PM.
DVDdoug is offline   Reply With Quote
Old 11-30-2014, 11:23 AM   #16
whiteaxxxe
Banned
 
Join Date: Jul 2014
Location: United States of Europe, Germany, Mönchengladbach
Posts: 2,047
Default

Quote:
Originally Posted by PhaseMaster View Post
DVDdoug stated that both channels of a stereo recording should be played back in the same polarity. This is something that everyone agrees on.

However, many people believe that reversing the polarity of both channels at the same time makes a substantial difference in the way music sounds. Here is a link to a reasonably short essay on this subject.

http://www.6moons.com/industryfeatur.../polarity.html

Here is a link to a long article by someone who believes that 92% of CDs play in the wrong polarity on 92% of stereo systems.

http://www.absolutepolarity.com/

I look forward to a spirited debate on this controversial issue.
what?

here is a link of people who believe that the universe was created by some superduperhero in 7 days.

the central term is "believe". believers are bullshit-talkers. because someone who believes doesnt know. simple as that.

so I do not read through your esoterik links of things that have to be believed. I dont believe in anything when I can get to know what it is all about. and you cant hear phase differences. quit believing and dive into science, then you know!
whiteaxxxe is offline   Reply With Quote
Old 12-01-2014, 03:27 AM   #17
evosilica
Human being with feelings
 
Join Date: Jan 2011
Location: Austria/Europe
Posts: 216
Default

Quote:
Originally Posted by PhaseMaster View Post
However, many people believe that reversing the polarity of both channels at the same time makes a substantial difference in the way music sounds.
While there can be an audible difference when reversing polarity, it is only audible with certain signals, a pure sawtooth for example. And even then, it just sounds a tiny little bit different and there's definitely not one better/right sounding polarity.
With "real" music signal the difference becomes completely irrelevant and is not substantial at all, unless you want to believe so. But as whiteaxxxe said, believing is not knowing.
If you search the internet for people who believe they can hear special things, you'll find them.
Don't get sucked into audiophoolism, as it makes "engineers" look like idiots and exposes their lack of knowledge.
Knowledge about psychoacoustics and proper ABX testing is the only way to go here.
evosilica is offline   Reply With Quote
Reply

Thread Tools
Display Modes

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


All times are GMT -7. The time now is 09:41 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.