"Cogging" Discussion

[ssully]

What gets 'panned' in software is very much at the mercy of room acoustics and speaker placement in practice.

 

[Sal1950]

Not having identical speakers all around can increase the cogging effect.

 

[ar surround]

If your speaker are placed close enough together there is no or minimal "cogging". If a sound is being panned around the room it is only natural that you would move your head to follow it. With the rear speakers far removed from the front you get the hole in the middle just as you do when the front speakers are too far apart.

Panning is used mainly as a gimmick, while I love the effect it is not a integral part of a typical surround recording, so why worry about it?

I would not want any artificial delay added to my music system!

While making my own 5.1 mixes, I was surprised to find how much stereo panning is actually used in many songs (Who Are You, She's Gone, et al) It seems to be a common method from the era with heavy employment of synths. So I have a choice of panning them strictly across the fronts or backs or having them pan from back to front to back. Depends on my mood. In all cases, I don't hear any "cogging."

Any bets on whether the new Steven Wilson Atmos of Baba O'Reilly will have the intro panning all over the room?

 

[Sonik Wiz]

Panning is used mainly as a gimmick, while I love the effect it is not a integral part of a typical surround recording, so why worry about it?



I would not want any artificial delay added to my music system!

Even tho yours & mine speaker layout is quite different, I agree that this cogging thing is a non-issue. It must be something midi is particularly sensitive to. I do not hear cogging via matrix decode or discrete.



I might disagree with your opinion that side panning is just a gimmick. Anytime you have a left ch only or right ch only into the Surround Master you get a center side pan where left front = left back, or Right front = right back. Ditto of course if you use actual QS decoding, Vario-Matrix or not. If you rotate your head a bit the sound will jump because your favoring the direction of a speaker feed. I don't see that as cogging.



Now for sounds that are actively panned from a rear speaker to a front speaker I think that is the only condition that this cogging discussion is focused on. Many times I have listened to and shared a simple test to demonstrate the value of out of phase blending in QS/Involve decoding. Pretty simple: two matching samples of white noise panned left to right, one normal & one with -7dB opposite phase blending. The normal phased one starts as center left moves a bit to center front & ends ups center right.



Meanwhile the phase adjusted tone starts out precisely left rear>center left>left front>center front, and continues the right side ending in right rear.



Not only is this a great demo of what this out of phase blending does I can also say as it moves from left rear to left front or right front to right rear, it is smooth and continuous. I don't hear any "speaker jumping" or cogging. For that matter on the Quadrafile double test record I can say the same thing for wurbly burbly sounds dynamically panned around, even on CD-4.



Agree also I would never want, nor do I need, any delay on the rear chs.

Original Dolby matrix surround had such terribly poor front/back separation they resorted to utilizing the Hass effect to keep dialog & music more up front, subjectively, to the listener.



No use for that today if your speakers are optimally set up. However I did run across an interesting article about using the Haas effect creatively in stereo recording & mixing:

Haas Effect: What It Is and How It's Used​

The Haas effect is a law which states that when one sound is followed by another with a delay time of 40 ms or less, the two are perceived as one sound.
producelikeapro.com

Well, heck, how did I screw this up? I just wanted to correct a typo & when I saved it incorporated my reply into something from Ken. So just to be clear as mud, the second lengthy quote is not from Ken, it's mine.

 

[MidiMagic]

I have experimented with cogging over the years, as the first time I heard cogging was with the EV decoder. The following is a list of various systems and whether they cog:

Fantasound - yes
Hafler diamond - no
Hafler 3-speaker - no
Scheiber - yes
EV-4 - yes
Dynaquad - less
QS - less
SQ - yes
EV-U - yes
UMX/BMX/UD4 - yes
Discrete/CD-4 - yes
Dolby Surround - no
Dolby PL-I - no
Dolby PL-II - no
Ambisonic = less
Dolby digital (all versions) - yes

 

[MidiMagic]

Even tho yours & mine speaker layout is quite different, I agree that this cogging thing is a non-issue. It must be something midi is particularly sensitive to. I do not hear cogging via matrix decode or discrete.

I might disagree with your opinion that side panning is just a gimmick. Anytime you have a left ch only or right ch only into the Surround Master you get a center side pan where left front = left back, or Right front = right back. Ditto of course if you use actual QS decoding, Vario-Matrix or not. If you rotate your head a bit the sound will jump because your favoring the direction of a speaker feed. I don't see that as cogging.

Now for sounds that are actively panned from a rear speaker to a front speaker I think that is the only condition that this cogging discussion is focused on. Many times I have listened to and shared a simple test to demonstrate the value of out of phase blending in QS/Involve decoding. Pretty simple: two matching samples of white noise panned left to right, one normal & one with -7dB opposite phase blending. The normal phased one starts as center left moves a bit to center front & ends ups center right.

Meanwhile the phase adjusted tone starts out precisely left rear>center left>left front>center front, and continues the right side ending in right rear.

Not only is this a great demo of what this out of phase blending does I can also say as it moves from left rear to left front, or right front to left rear, it is smooth and continuous. I don't hear any "speaker jumping" or cogging. For that matter on the Quadrafile double test record I can say the same thing for wurbly burbly sounds dynamically panned around, even on CD-4.

Agree also I would never want, nor do I need, any delay on the rear chs.
Original Dolby matrix surround had such terribly poor front/back separation they resorted to utilizing the Hass effect to keep dialog & music more up front, subjectively, to the listener.

No use for that today if your speakers are optimally set up. However I did run across an interesting article about using the Haas effect creatively in stereo recording & mixing:

https://producelikeapro.com/blog/haas-effect/

I have been working on the problem. I think a tiny bit of delayed sound on the other side (other ear) is enough to merge the images into one. My experiments have been put on hiatus by the damage to my studio from a falling tree.

 

[stevendive]

I have experimented with cogging over the years, as the first time I heard cogging was with the EV decoder. The following is a list of various systems and whether they cog:

Fantasound - yes
Hafler diamond - no
Hafler 3-speaker - no
Scheiber - yes
EV-4 - yes
Dynaquad - less
QS - less
SQ - yes
EV-U - yes
UMX/BMX/UD4 - yes
Discrete/CD-4 - yes
Dolby Surround - no
Dolby PL-I - no
Dolby PL-II - no
Ambisonic = less
Dolby digital (all versions) - yes

It’s surprising you have ambisonic as ‘less’ rather than ‘no’. In a properly set up system with identical speakers and a (virtual) central listening position, no direction is favoured and there is no audible pull into any loudspeaker.

 

[ar surround]

No use for that today if your speakers are optimally set up. However I did run across an interesting article about using the Haas effect creatively in stereo recording & mixing:

https://producelikeapro.com/blog/haas-effect/

I make use of the Haas Effect quite often:

Mono massed vocals:

Very broad widening across the sound stage using Haas Effect:

 

[par4ken]

Positioning sounds to the sides is different than panning (around the room). Depending on your speaker placement and seating you might hear the side sound from the back or the front instead of to the side. I would not consider that to be "cogging"

That issue is not usually a problem when using closely spaced speakers, when the sound is balanced it images clearly from the sides! Even in a larger room a sound positioned to the sides should sound as if it does come from the side albeit not as clearly as if a speaker was there. At worst you hear the same sound from both speakers ie double stereo. Still sounds great to me!

Also if the rear speakers are placed more to the sides or even slightly upfront instead of behind you the "side" imaging is much better.

Actual "panning effects" of the swirling around the room type is a gimmick the same as the early ping pong stereo records were. I love both effects but they are less and less common now. IMHO not a reason to be concerned if they don't pass the test without any "cogging".

 

[Sal1950]

Actual "panning effects" of the swirling around the room type is a gimmick the same as the early ping pong stereo records were. I love both effects but they are less and less common now. IMHO not a reason to be concerned if they don't pass the test without any "cogging".

With the appearance of Atmos, I believe a multich system that is properly designed and setup has become even more important.
Speakers that are both extremely similar in tonal balance, along with correct positioning and proper use of the DRC system included, all combine to determine how well any and all positioning effects are reproduced.
This maybe the reason I read so much from the "stereo only" clan that they've never been impressed by the surround sound systems they've heard. Getting everything "just right" isn't for the guy that just wants to throw a few speakers down and push the ON button.

 

[MidiMagic]

Positioning sounds to the sides is different than panning (around the room). Depending on your speaker placement and seating you might hear the side sound from the back or the front instead of to the side. I would not consider that to be "cogging"

That issue is not usually a problem when using closely spaced speakers, when the sound is balanced it images clearly from the sides! Even in a larger room a sound positioned to the sides should sound as if it does come from the side albeit not as clearly as if a speaker was there. At worst you hear the same sound from both speakers ie double stereo. Still sounds great to me!

Also if the rear speakers are placed more to the sides or even slightly upfront instead of behind you the "side" imaging is much better.

Actual "panning effects" of the swirling around the room type is a gimmick the same as the early ping pong stereo records were. I love both effects but they are less and less common now. IMHO not a reason to be concerned if they don't pass the test without any "cogging".

I DO consider it to be cogging, because the sound is locked to a speaker instead of where the pan is supposed to be.

I have studied the problem and it comes from how our hearing system locates sounds.

When a sound comes to our ears, the hearing system derives the direction of the sound source from the intensity, time, and phase differences.

When the sound comes from multiple directions, the hearing system uses the loudest instance of the sound in each ear as the main event and considers the other instances as reflections of that sound.

If the sound comes from two directions in front of the listener or two direction behind the listener simultaneously (or near simultaneously), the system uses the various loudness levels from the instances to derive a position between them. This is why level panning works in front of the listener or behind the listener.

If the sound comes from the side, level panning does not work. There are two or three sound sources depending on the setup. The loudest sound arriving at both ears is taken as the direction the sound comes from, with the other speaker source(s) treated as echos because they have different levels and timings.

As the levels change as the sound is panned from one speaker to another, the dominant source changes to another speaker, the level, timing, and phase of the dominant source changes. This sudden change in dominance causes the cogging.

Level panning does not work when both speakers are on the same side of the listener.

Some people seem to be able to hear the pan without cogging. Maybe they have somehow trained their ears to hear side panning, but it doesn't work for me.

 

[par4ken]

I DO consider it to be cogging, because the sound is locked to a speaker instead of where the pan is supposed to be.

I have studied the problem and it comes from how our hearing system locates sounds.

When a sound comes to our ears, the hearing system derives the direction of the sound source from the intensity, time, and phase differences.

When the sound comes from multiple directions, the hearing system uses the loudest instance of the sound in each ear as the main event and considers the other instances as reflections of that sound.

If the sound comes from two directions in front of the listener or two direction behind the listener simultaneously (or near simultaneously), the system uses the various loudness levels from the instances to derive a position between them. This is why level panning works in front of the listener or behind the listener.

If the sound comes from the side, level panning does not work. There are two or three sound sources depending on the setup. The loudest sound arriving at both ears is taken as the direction the sound comes from, with the other speaker source(s) treated as echos because they have different levels and timings.

As the levels change as the sound is panned from one speaker to another, the dominant source changes to another speaker, the level, timing, and phase of the dominant source changes. This sudden change in dominance causes the cogging.

Level panning does not work when both speakers are on the same side of the listener.

Some people seem to be able to hear the pan without cogging. Maybe they have somehow trained their ears to hear side panning, but it doesn't work for me.

We will have to disagree. When the speakers are close together PWM works just fine. The best example is in an automobile, at least the older ones where the rear speakers were mounted on the rear deck. This does not work in the modern cars where the rear speakers are mounted low in the doors, you can't hear them properly in the front you just hear a bit of muffled sound from them.

Anyway using a stereo source and the fadder control you can move the sound from front to rear. In the "balanced" position vocals seem to come from the centre about head level. The left and right sound images are clearly to the sides but are a bit broader image than if an actual speaker was there. What more could you want? I much prefer to listen to stereo that way rather than via only two speakers!

If the speakers are farther apart then you hear double stereo, no defined side images but I am perfectly fine with that situation as well!

 

[Sonik Wiz]

I make use of the Haas Effect quite often:

Mono massed vocals:

View attachment 89151

Very broad widening across the sound stage using Haas Effect:

View attachment 89152

Well that was pretty interesting! Listened to them both a few times. I find the"mono massed vocals" slightly more appealing to me. Do you remember or have a ROT what your delay was at? Based on using some VST presets I've used this sounds like a little more than the Hass amount of delay. More like what would be chorusing.

 

[stevendive]

We will have to disagree. When the speakers are close together PWM works just fine. The best example is in an automobile, at least the older ones where the rear speakers were mounted on the rear deck. This does not work in the modern cars where the rear speakers are mounted low in the doors, you can't hear them properly in the front you just hear a bit of muffled sound from them.

Anyway using a stereo source and the fadder control you can move the sound from front to rear. In the "balanced" position vocals seem to come from the centre about head level. The left and right sound images are clearly to the sides but are a bit broader image than if an actual speaker was there. What more could you want? I much prefer to listen to stereo that way rather than via only two speakers!

If the speakers are farther apart then you hear double stereo, no defined side images but I am perfectly fine with that situation as well!

The point is that if good localisation is wanted for most people then aural cues for direction should not be in conflict and that different people are variously sensitive to these conflicts.

 

[ar surround]

Well that was pretty interesting! Listened to them both a few times. I find the"mono massed vocals" slightly more appealing to me. Do you remember or have a ROT what your delay was at? Based on using some VST presets I've used this sounds like a little more than the Hass amount of delay. More like what would be chorusing.

Yes, it is chorusing. The mono massed vocals still exist in the center of the mix with the chorused version added to the sides. The amount of spread can be controlled by adjusting the volume of the mono center.

As regard to what is ultimately more appealing, it depends on how it sounds when combined with all of the other elements into the 5.1 mix. The mix most definitely sounds different in a room with speakers than with headphones.

In this particular 5.1 mix, there is a dense mono front with the lead vocal and front instruments, so employing the spread chorus adds diversity to the the front sound stage:

And here is the full 5.1 mixed down to stereo:

 

[MidiMagic]

Here is how I did my experimentation:
- I have a Behringer MX2442FX mixer,
- I put my RM encoder insert in the bus 3 and 4 inserts to make buses 3 and 4 encode back signals in RM.
- I connected mics and other sound sources to mixer input buses.
- I sent the odd bus outputs to main left, and the even bus outputs to main right.
- I sent the left and right main outputs to various matrix decoders.

(The mixer in the drawings is not the Behringer, but a miser the venue had.)

1. When I used the EV-4 decoder, the sound image cogs between LF and LB when I turn the panpot and switch between busses 1 2 and busses 3 4 at full pan. Note that I hear it when I turn the panpot at some point, not when I change buses.

2. The same thing happens with QS.

3. The same thing happened with DQ, but it was close to the front speakers.

4. When I used Dolby Surround, the image panned smoothly around the listener.

5. Dolby Pro-Logic worked as well as Dolby Surround.

6. When I used the Hafler Diamond (temporarily changing speaker connections and locations) it did not cog. The image panned smoothly around the listener.

Of course, when I turned my head toward the left or right, any cogging went away from the sides, but appeared in the front and back. in the systems that cog.