Why evaluating the sound of a single speaker is essential

[Juhazi]

Some people have difficulties understanding relative vs. absolute scoring... Thank you Floyd for patience in explaining!

 

[RickS]

Agreed - multi-language editions would be nice. 1st, 2nd and 3rd editions exist in Chinese translations, so some "foreigners" out there are able to read. Heard any Chinese loudspeakers lately? Truth is that there are a great many, and some are decent - not "high end" though, not where the market is. I have been invited to lecture at Chinese universities, but declined . . . Many Chinese designers still have a bit of the old "golden ear" philosophy, and I am not eager to educate them.

I own the 2nd edition, and it is a great reference. I see there is Kindle edition now so if 4th edition will have as well, plan to purchase. As for other languages, Kindle does not offer translated editions but would allow selected parts to be translated.

 

[DonH56]

I own the 2nd edition, and it is a great reference. I see there is Kindle edition now so if 4th edition will have as well, plan to purchase. As for other languages, Kindle does not offer translated editions but would allow selected parts to be translated.

The 2nd is almost identical to the 1st. The 3rd is significantly different, and it sounds like the 4th will be even better. Definitely on my "got to buy" list.

Glad to see @Floyd Toole back posting again, hope the new digs are shaping up!

 

[Floyd Toole]

I would buy them all if they were translated…

I am not trying to"sell" the new edition. I am "selling" knowledge - and here it is free. I don't need the money from the book, and technical books don't earn much anyway - enough to buy some good wine occasionally.

 

[Esprit]

am not trying to"sell" the new edition. I am "selling" knowledge

I love knowledge and humbly want to learn but English is not my language so learning is even more difficult.
Good afternoon

 

[Esprit]

Heard any Chinese loudspeakers lately?

Many years ago we bought them directly from China.

 

[RickS]

placeholder post demarcating original content from newer.

 

[goat76]

It is particularly disgusting when the center channel is not used at all, as happens in lots of multichannel music and music videos, or it is sent to all three fronts. It is remarkable that we humans have so well adapted to the distorted sound of phantom images on the soundstage that a center channel is criticized for sounding different. It's an upside down world in some ways - all because of the default format: stereo.

The reason for mixing the vocals to the two main front speakers, or all three fronts for multichannel music, isn't because we humans have “adapted to distorted sound of phantom images”. The main problem is that human voices often sound way too isolated and narrow when mixed only to the center speaker, especially for intimate recordings which most vocal recordings tend to be.

If you ever get the chance to compare the same vocal track panned to the center channel vs a phantom center, and panned to all three front speakers, I think it will be a good chance that you will prefer both the phantom center and the three speakers over the isolated sound from just the center speaker. I have compared this myself while mixing in multichannel in a well-set-up system, with a good center channel and a distinct-sounding phantom center. The vocals just sound less congested than what it sounds like isolated to the center speaker.

 

[Tim Link]

The reason for mixing the vocals to the two main front speakers, or all three fronts for multichannel music, isn't because we humans have “adapted to distorted sound of phantom images”. The main problem is that human voices often sound way too isolated and narrow when mixed only to the center speaker, especially for intimate recordings which most vocal recordings tend to be.

If you ever get the chance to compare the same vocal track panned to the center channel vs a phantom center, and panned to all three front speakers, I think it will be a good chance that you will prefer both the phantom center and the three speakers over the isolated sound from just the center speaker. I have compared this myself while mixing in multichannel, in a well-set-up system whit a good center channel and a distinct-sounding phantom center. The vocals just sound less congested than what it sounds like isolated to the center speaker.

I'll have to say I'm with Toole on this one. I think the hard center adds intimacy, immediacy, and presence. If anything, it adds too much of all that compared to stuff panned between the speakers. Totally the opposite of narrownesss or congestion, it's extremely present and vivid. In the case of my small room there's also a problem that the center channel ends up being too close to the listening position compared to the side channels so that it dominates, which on some recordings is actually great. Perhaps if you have a really beamy center it could sound too "closed in" compared to the side speakers, although ideally they would be matched for dispersion. Another potential issue is if the speakers are all wide dispersion and the center is hanging out by itself with no nearby side walls to create early reflections like the side speakers are getting. That can be a real issue and I've heard it. Ideally the side speakers should have plenty of distance from the side walls, or better yet installed in a room that has soffited corners so that the front center section of wall is the same width as the side 45 degree soffit walls so that each speaker is acoustically loaded into the space in the same way. Or, the side wall reflections can just be better controlled and the total width of the listening triangle widened to add envelopment as there will be no hole in the middle problem.
The phantom center is a timbral trainwreck to my ears, a necessary evil when only two speakers are available. It's veiled, dark, and murkey, and it is unachored in space, moving around with you. I hear the timbral effect of it at every high end audio trade show I go to, and any other place its employed except when the speakers are so bad it doesn't matter.
When I tried multi channel music and heard music mixes that were leaving off the center channel, or mixing to all 3 fronts, I gave up on it. I could get better results just using dolby pro logic in movie mode on stereo mixes, and that ain't saying much. I've tried making my own 3 speaker upmixing regimes but found it challenging, so for now am settling on the phantom center after finding surprising benefit from making the listening triangle very wide, which lowers the first (and worst) notch in the comb filtering to a frequency that's less annoying to me.
Funny thing about the multi-channel mixes of old classics, my friend came over while I was listening to some upmixed Fleetwood Mac and asked if it was a newer performance. He said they just didn't seem to have the energy and enthusiam that they used to. I laughed a bit at that knowing it was just a re-mix of the same old material. But then I put on the old stereo version and it surprised me how relatively lifeless the multi channel mix had come to sound. That had nothing to do with use of the center channel, which I found unsatisfactory as well, but more to do with the aesthetic choices the newer mixing/mastering person had made.

 

[Sal1950]

I'll have to say I'm with Toole on this one. I think the hard center adds intimacy, immediacy, and presence. If anything, it adds too much of all that compared to stuff panned between the speakers. Totally the opposite of narrownesss or congestion, it's extremely present and vivid. In the case of my small room there's also a problem that the center channel ends up being too close to the listening position compared to the side channels so that it dominates, which on some recordings is actually great. Perhaps if you have a really beamy center it could sound too "closed in" compared to the side speakers, although ideally they would be matched for dispersion. Another potential issue is if the speakers are all wide dispersion and the center is hanging out by itself with no nearby side walls to create early reflections like the side speakers are getting. That can be a real issue and I've heard it. Ideally the side speakers should have plenty of distance from the side walls, or better yet installed in a room that has soffited corners so that the front center section of wall is the same width as the side 45 degree soffit walls so that each speaker is acoustically loaded into the space in the same way. Or, the side wall reflections can just be better controlled and the total width of the listening triangle widened to add envelopment as there will be no hole in the middle problem.
The phantom center is a timbral trainwreck to my ears, a necessary evil when only two speakers are available. It's veiled, dark, and murkey, and it is unachored in space, moving around with you. I hear the timbral effect of it at every high end audio trade show I go to, and any other place its employed except when the speakers are so bad it doesn't matter.
When I tried multi channel music and heard music mixes that were leaving off the center channel, or mixing to all 3 fronts, I gave up on it. I could get better results just using dolby pro logic in movie mode on stereo mixes, and that ain't saying much. I've tried making my own 3 speaker upmixing regimes but found it challenging, so for now am settling on the phantom center after finding surprising benefit from making the listening triangle very wide, which lowers the first (and worst) notch in the comb filtering to a frequency that's less annoying to me.
Funny thing about the multi-channel mixes of old classics, my friend came over while I was listening to some upmixed Fleetwood Mac and asked if it was a newer performance. He said they just didn't seem to have the energy and enthusiam that they used to. I laughed a bit at that knowing it was just a re-mix of the same old material. But then I put on the old stereo version and it surprised me how relatively lifeless the multi channel mix had come to sound. That had nothing to do with use of the center channel, which I found unsatisfactory as well, but more to do with the aesthetic choices the newer mixing/mastering person had made.

Speaking of modern fully discrete multich mixing, how could a vocal or whatever sent to the center channel not be a cleaner, higher fidelity reproduction, than one getting intermixed with a lot of other information?

 

[goat76]

Speaking of modern fully discrete multich mixing, how could a vocal or whatever sent to the center channel not be a cleaner, higher fidelity reproduction, than one getting intermixed with a lot of other information?

The reason why the vocals are rarely mixed exclusively to the center channel in multichannel mixing is more about achieving the desired width to the source, than anything else. It's also fairly easy to regain the clean and lost fidelity to phantom-centered sound objects using EQ.

Have a listen to a selection of Atmos mixes of your choice, and put your ear close to the center speaker and then any of the front speakers, and you will realize how uncommon it is the mix the vocals exclusively to the center speaker.

 

[NTK]

We have a thread on the misuse of center channel. After all, these are the same people when in the 1980's scientists and engineers gave them the wonderful technology of digital audio, and they rewarded us back with the loudness war.

Dr Olive is not happy with how Dolby Atmos is mixed

So much for the claim I heard from mixers that Dolby Atmos Music translates what the artist intended no matter the setup or space. It might be true for 1 person if they sit in the right seat, but everyone else will hear something the artist did not intend

www.audiosciencereview.com

 

[Sal1950]

The reason why the vocals are rarely mixed exclusively to the center channel in multichannel mixing is more about achieving the desired width to the source, than anything else. It's also fairly easy to regain the clean and lost fidelity to phantom-centered sound objects using EQ.

Yes I should have been more clear, I didn't expect the vocals to be exclusive in the center channel, but was simply referring to one that wasn't a phantom creation of all that was going on in the L & R channels.

We have a thread on the misuse of center channel. After all, these are the same people when in the 1980's scientists and engineers gave them the wonderful technology of digital audio, and they rewarded us back with the loudness war.

If you knew much about Atmos mixing you'd find very very few of the engineers that mix for Atmos are at all involved in loudness war type compression.
That would make no sense at all and is pretty much the home for 2ch mixing being listened to in cars, on cheap home systems and the like.

 

[Floyd Toole]

It's also fairly easy to regain the clean and lost fidelity to phantom-centered sound objects using EQ.

As will be explained in the 4th edition, it is not so simple. The audible flaws are non-minimum-phase (acoustical interference) problems that are not "correctable" with EQ. They differ with listener position and are worst for the listener in the stereo sweet spot. Any EQ attempted to correct for what is heard in the sweet spot becomes an error for anyone else in the room. It is a dilemma requiring adaptation and compromise.

Have a look at this reference for a professional comment:
King, R., Theriault, M., Massenburg, G. (2023). “A Practical Approach to the Use of Center Channel in Immersive Music Production”, Audio Eng. Soc., 155th Convention, Paper 168.

 

[Newman]

As will be explained in the 4th edition, it is not so simple. The audible flaws are non-minimum-phase (acoustical interference) problems that are not "correctable" with EQ. They differ with listener position and are worst for the listener in the stereo sweet spot. Any EQ attempted to correct for what is heard in the sweet spot becomes an error for anyone else in the room. It is a dilemma requiring adaptation and compromise.

It also becomes an error for all content not panned to centre. As explained in the 1st edition, IIRC.

 

[goat76]

As will be explained in the 4th edition, it is not so simple. The audible flaws are non-minimum-phase (acoustical interference) problems that are not "correctable" with EQ. They differ with listener position and are worst for the listener in the stereo sweet spot. Any EQ attempted to correct for what is heard in the sweet spot becomes an error for anyone else in the room. It is a dilemma requiring adaptation and compromise.

Have a look at this reference for a professional comment:
King, R., Theriault, M., Massenburg, G. (2023). “A Practical Approach to the Use of Center Channel in Immersive Music Production”, Audio Eng. Soc., 155th Convention, Paper 168.

The most optimal stereo listening is and has always been a “one-man sport”, so any listening position other than the absolute sweet spot will cause much bigger compromises to the overall stereo sound than the tonality changes happening to the phantom-centered sounds in those off-centered listening positions. I mean, the whole stereo representation practically collapses if you are listening off-center, which should be of a much bigger concern. Therefore, I don't think we should be aiming our efforts at optimizing smaller things like tonality changes for those already compromised listening positions, not when it comes to stereo listening.

The studio standard is an equilateral listening triangle in audio productions, which is also the most common recommendation for home stereo setups of loudspeakers. The possible deviations should therefore be fairly small, or at least small enough from room to room, as long as you are following that standard.

The main stereo problem with phantom-centered sounds is that they often end up sounding a bit dull compared to how they would sound using a center speaker. But that problem will, in most cases, already be taken care of in the mixing stage of the audio production. As an example, if the vocals sound dull to the mixing engineer, that person will probably and most likely do something about it using EQ, even if (s)he doesn't have a clue what caused the "dullness" to begin with. The problem is, therefore, often solved automatically as a "side thing" just by the fact that the mixing engineer aims for a vocal sound that doesn't sound dull, and that is also most likely done by them sitting in an equilateral listening triangle.



Yesterday, I did some testing and comparison between a vocal track I panned to the center speaker versus the same track as a phantom center. As expected, the phantom center lacks a little bit of definition and pinpoint accuracy compared to how it sounds coming from the physical center speaker, but the only thing it takes is a single EQ filter at 1.8 kHz, with a raise of 2.8 dB, and Q 3.0 to regain the lost focus of the phantom-centered vocal track.

I didn't bother trying to make the phantom track sound identical to how it sounds from the center speaker, as I was mostly interested in that particular area around 1.8-2 kHz, which seems to be the most problematic frequency area when these discussions come up.

All three tracks below are in multichannel 5.1, even though just 1 or 2 channels are used. They can be played using Windows Media Player if the sound in Windows is set up in multichannel and connected to an AVR.

Track 1, vocal panned to center speaker: https://www.dropbox.com/scl/fi/bc35...ey=ggzx85943xv87kdfi5vzx03bc&st=rzrs3z9d&dl=0

Track 2, vocal phantom center: https://www.dropbox.com/scl/fi/epfb...ey=rkzim28lybl7dmlxw71u01wxd&st=u43gqurc&dl=0

Track 3, vocal phantom center, with an EQ filter at 1.8 kHz: https://www.dropbox.com/scl/fi/c7s8...ey=fdevbf0i5lkfm9dyzkrallkfz&st=ileos289&dl=0

 

[goat76]

It also becomes an error for all content not panned to centre. As explained in the 1st edition, IIRC.

You are correct, the EQ adjustment can't be done to the loudspeakers themselves; it must be done in the music mix to the actual tracks panned to the phantom center. This is most likely already done by the mixing engineer while just making regular EQ adjustments to form the different sound objects in the mix to his/her liking, with or without the knowledge of why the track sounded a bit dull in the first place.

 

[RickS]

This is a new thread for useful existing content that was originally in another thread.

If you find issues, use Report function to indicate. Posting issues will clutter the thread and the posts will be deleted.

 

[Vladimir Filevski]

True, some speakers are great in mono and stereo, but some are average in mono and stereo. And then, some are average in mono but good in stereo - and the whole system of speaker recommendations based on mono collapses.

Somehow I missed this nonsense, but it is never too late to react accordingly. You are continuing to beat your dead horse, in spite of all hard evidence against your false "logic".

True, some speakers are great in mono and stereo, but some are average in mono and stereo.

False.
Not some, but all speakers which are great in mono are great in stereo also.
Not some, but all speakers which are average in mono, are average in stereo also.

And then, some are average in mono but good in stereo - and the whole system of speaker recommendations based on mono collapses.

False.
Not a single one speaker which is average in mono is good in stereo. It will be just average in stereo, of course.

- and the whole system of speaker recommendations based on mono collapses.

You are (unsuccessfully) trying to manipulate the simple hard evidence - difference between great and average speakers in stereo are smaller than in mono.

 

[pma]

IMO comparing speakers based on measurements only is difficult. As an example - imagine Neumann KH 120 and Wilson Audio Maxx. KH120 has seemingly textbook plots. Would it be preferred in a level matched listening test? Would it even be possible to match the level properly? I do not think so.