New speaker design by NOOB

[Adam Frandsen]

Not to repeat myself but speaker design is hard. You may be underestimating that difficulty.

Over my career I have worked with a handful of skilled loudspeaker design engineers, and had the honor of knowing a few giants. Most had decades of experience that informed their design choices.

By all means build something. You will learn from that experience.

JR

I already incorporated a lot of the acoustic elements talked about below in my speaker box design like baffle size, baffle thickness, compartment volumes, wavelengths, horizontal as well as vertical placement, tweeters center for optimal result, listening height, floor reflections, room size of 5x5m. Not saying the outcome is necessarily right, and I ended up with 200kg per speaker, which is on the heavy side.

https://www.linkwitzlab.com/Fitz/acoustics-hearing.htm

 

[pucho812]

Westlakes ultimate design, the sm-1, the 300,000 dollars a pair design clocks in at 900 pounds each and 6 feet tall each. It contains:

• Dual 18" woofers
• 12 midbass
• 2" horn loaded compression midrange
• 1/2" compression super tweeter

It’s a 5 way crossover design.

It was no easy task to design such a speaker. It’s even harder to ship them.

 

[Adam Frandsen]

I've noticed that some members of this forum are very enthusiastic about DSP-based systems and seem to believe that ambitious projects are best left to the professionals. While I respect all viewpoints, I believe it's important to remember that there's room for everyone, so I think I will make a 9-way speaker system using DSP and a 9 channel power amp. The build will be like this:

1. Sub-Bass (15 Hz to 60 Hz): One could consider a specialized subwoofer capable of reaching down to 15 Hz for this lowest range. While larger drivers might offer superior performance, the smaller size would help maintain the overall system's compactness.
2. Mid Bass (60 Hz to 200 Hz): A 6.5-inch or 8-inch woofer or mid-woofer could potentially be suitable for this range, accommodating the lower range of most musical instruments and male vocals.
3. High Bass (200 Hz to 300 Hz): For this range, a 5-inch or 6.5-inch mid-woofer or large midrange driver might be a good fit, handling the upper bass and lower midrange frequencies.
4. Low Midrange (300 Hz to 500 Hz): Theoretically, a 4-inch or 5-inch midrange driver could work well here, covering the lower midrange frequencies.
5. Midrange (500 Hz to 2,000 Hz): A dedicated 3-inch or 4-inch midrange driver could be used for this critical range, which is key for music and voice reproduction.
6. High Midrange (2,000 Hz to 4,000 Hz): A 2-inch or 3-inch midrange or small tweeter could potentially handle this range, covering the upper midrange frequencies.
7. Low Treble (4,000 Hz to 6,000 Hz): One could use a 1-inch dome tweeter for this range, contributing to the "brightness" or "presence" of the sound.
8. Mid Treble (6,000 Hz to 10,000 Hz): The same 1-inch dome tweeter could also theoretically handle this range, adding to the "sparkle" or "airiness" of the sound.
9. High Treble (10,000 Hz to 40,000 Hz): A specialized super tweeter, capable of reaching up to 40,000 Hz, could be used for this highest frequency range, contributing to the "detail" or "resolution" of the sound.

So, to those who might scoff at this idea, I say: isn't it fun to dream? And to those who might be inspired by this concept, I say: let's keep pushing the boundaries and see what we can create (or which buttons we can push )

Now, back to my analogue build. I look forward to continuing to learn and grow with the help of this community, and I welcome all constructive and respectful feedback.

 

[pucho812]

Whatever floats your boat.

 

[Adam Frandsen]

Now I have rethought my initial idea, and will likely be doing something like this:

1. Subwoofer (Low Frequency Driver): Handles frequencies up to around 84.9 Hz. Given the long wavelengths of these low frequencies, a larger driver is needed to move the larger volumes of air. A 12" or 15" subwoofer would be good choices. The cutoff ensures that the entire bass vocal range is handled by the same driver, the woofer.
2. Woofer (Low-Mid Frequency Driver): Handles frequencies from 84.9 Hz to 350 Hz. This range includes the fundamental frequencies of many musical instruments and the lower range of male voices. A 6.5" or 8" woofer would be suitable for this range. It's above the highest note of the bass voice and below the lowest note of the soprano voice. It's also below the "sweet spot" for orchestral music (around 500-600 Hz).
3. Midrange Driver: Handles frequencies from 350 Hz to 3500 Hz. This is a critical range for both music and voices, containing most of the fundamental frequencies of orchestral instruments and the human voice. A 4" or 5" midrange driver would be ideal for this range. Cutoff is above the highest note of the tenor voice and the highest note of the soprano voice. It's also within the range of frequencies where the ear is particularly sensitive (around 2000-5000 Hz). Here will also ideally be a 3db dip switch for around 3-4khz, to account for HRTF directionality, now with the crossover being where it is, this may be impossible, and I will have to find another way to mitigate this problem.
4. Tweeter (High Frequency Driver): Handles frequencies from 3500 Hz up to around 40,000 Hz. This range includes the highest notes of the piccolo and violin, as well as the harmonics of lower-pitched instruments and voices. A 1" dome tweeter would be a liekly choice for this range.

I will still be using 4th-order-linkwitz riley, l-pad for the tweeter, zobel networks for impedance correction and some form of equalization for the bass. The enclosure for this will be more advanced than just a box, after having read more incl. what was available at linkwitzlab.

 

[MidnightArrakis]

I've noticed that some members of this forum are very enthusiastic about DSP-based systems and seem to believe that ambitious projects are best left to the professionals. While I respect all viewpoints, I believe it's important to remember that there's room for everyone, so I think I will make a 9-way speaker system using DSP and a 9 channel power amp. The build will be like this:

1. Sub-Bass (15 Hz to 60 Hz): One could consider a specialized subwoofer capable of reaching down to 15 Hz for this lowest range. While larger drivers might offer superior performance, the smaller size would help maintain the overall system's compactness.
2. Mid Bass (60 Hz to 200 Hz): A 6.5-inch or 8-inch woofer or mid-woofer could potentially be suitable for this range, accommodating the lower range of most musical instruments and male vocals.
3. High Bass (200 Hz to 300 Hz): For this range, a 5-inch or 6.5-inch mid-woofer or large midrange driver might be a good fit, handling the upper bass and lower midrange frequencies.
4. Low Midrange (300 Hz to 500 Hz): Theoretically, a 4-inch or 5-inch midrange driver could work well here, covering the lower midrange frequencies.
5. Midrange (500 Hz to 2,000 Hz): A dedicated 3-inch or 4-inch midrange driver could be used for this critical range, which is key for music and voice reproduction.
6. High Midrange (2,000 Hz to 4,000 Hz): A 2-inch or 3-inch midrange or small tweeter could potentially handle this range, covering the upper midrange frequencies.
7. Low Treble (4,000 Hz to 6,000 Hz): One could use a 1-inch dome tweeter for this range, contributing to the "brightness" or "presence" of the sound.
8. Mid Treble (6,000 Hz to 10,000 Hz): The same 1-inch dome tweeter could also theoretically handle this range, adding to the "sparkle" or "airiness" of the sound.
9. High Treble (10,000 Hz to 40,000 Hz): A specialized super tweeter, capable of reaching up to 40,000 Hz, could be used for this highest frequency range, contributing to the "detail" or "resolution" of the sound.

So, to those who might scoff at this idea, I say: isn't it fun to dream? And to those who might be inspired by this concept, I say: let's keep pushing the boundaries and see what we can create (or which buttons we can push )

Now, back to my analogue build. I look forward to continuing to learn and grow with the help of this community, and I welcome all constructive and respectful feedback.

Don't forget to consider "acoustic phase-alignment" of all of the drivers when mounting them together within your speaker cabinet enclosure!!!

/

 

[michinger]

Hey Adam, I did not read the whole text here, but your plan seems very ambitious. Building myself speaker for more then 25y. and built more then 100pieces (mostly not engineered myself) I can tell you that it is much more complicated then you think.

Also you need a place for measurements (down to the mids you can do at home) and a lot of knowledge in electronics.
A DSP Speaker is much more easy (not better then clever analogue designs) and even then you need to measure.

First have a look to this two magazines (only in german): HobbyHifi and Klang&Ton there you can learn all the basics.
Also you need measuring software like: ARTA, REW or similar
For simulations the best is to start with Boxsim or minimum with BASScad (for beginners)

 

[MicUlli]

Hi Adam,
dont feel overloaded by all the comments in this thread. All participants want to help you. And a lot of them have great experience in acoustics and electronics..
30% sound quality is done by the speaker itself, 70% depend on room issues. Optimizing room acoustics is therefore more effective than optimizing the speaker only. As you explained your living room is 5x5 m. I suppose the room height is around 2,6 m. The lowest standing wave frequency is therefore around 34 Hz and gives a boost of at least 6 dB. In conjunction with a closed design with HP2-corner at approx. 55 Hz you can get a considerably flat response down to 30 Hz. For this construction a 6,5" or 8" bass chassis would be fully sufficient..
I strongly recommend startin with a "small" speaker design to avoid cost and frustration

 

[Adam Frandsen]

Hey Adam, I did not read the whole text here, but your plan seems very ambitious. Building myself speaker for more then 25y. and built more then 100pieces (mostly not engineered myself) I can tell you that it is much more complicated then you think.

Also you need a place for measurements (down to the mids you can do at home) and a lot of knowledge in electronics.
A DSP Speaker is much more easy (not better then clever analogue designs) and even then you need to measure.

First have a look to this two magazines (only in german): HobbyHifi and Klang&Ton there you can learn all the basics.
Also you need measuring software like: ARTA, REW or similar
For simulations the best is to start with Boxsim or minimum with BASScad (for beginners)

My German is decent, but not fluent, I will try and see if I can digest it. I understand that a DSP speaker is somewhat easier to achieve good results with, it would also beneficial for tweaking, should I want to move the speakers. But, I like to play records and to me it defeats the purpose of a true analogue source, to have it go through a digital process or even a class D amp with binary switching. I am sure they can sound wonderful and perhaps measure better, and maybe I will consider it down the road, should I truly fail in creating something good and analogue. I build my own subs with scan speak drivers and a plate amp, where all I did was charge rca lead wires and speaker wire leads to correct quality and gauge. The boxes were very simple, but calculated using online tools. They sound wonderful! I have them facing my main speakers set to 180 degrees, and it works really well. I know building a full speaker is far more difficult, but I hope to get there.

 

[Adam Frandsen]

Hi Adam,
dont feel overloaded by all the comments in this thread. All participants want to help you. And a lot of them have great experience in acoustics and electronics..
30% sound quality is done by the speaker itself, 70% depend on room issues. Optimizing room acoustics is therefore more effective than optimizing the speaker only. As you explained your living room is 5x5 m. I suppose the room height is around 2,6 m. The lowest standing wave frequency is therefore around 34 Hz and gives a boost of at least 6 dB. In conjunction with a closed design with HP2-corner at approx. 55 Hz you can get a considerably flat response down to 30 Hz. For this construction a 6,5" or 8" bass chassis would be fully sufficient..
I strongly recommend startin with a "small" speaker design to avoid cost and frustration

Yes, will get a USB measuring mic and REW and do my best to achieve good results.

 

[thor.zmt]

Yes, will get a USB measuring mic and REW and do my best to achieve good results.

Adam, you got of very good free avise in this thread.

The very first thing I will suggest again is that you download vituixcad and learn to use it.

It is a tool that can help you to see if things you want to do have a chance of working.

A simulator does guarantee results, but it a good way to show you what will not work and allow you to iterate rapidly.

Now Ulli quipped:

30% sound quality is done by the speaker itself, 70% depend on room issues.

I will rephrase that as: "30% or less of what you hear is ON-AXIS Response, 70% is OFF-AXIS response (which some people mistake for problems caused by the room, when it is in fact poor design of the speaker)"

I saw you re-evaluated setup...

Now I have rethought my initial idea, and will likely be doing something like this:

This now seems a bit more realistic.

Midrange Driver: Handles frequencies from 350 Hz to 3500 Hz. This is a critical range for both music and voices, containing most of the fundamental frequencies of orchestral instruments and the human voice. A 4" or 5" midrange driver would be ideal for this range. Cutoff is above the highest note of the tenor voice and the highest note of the soprano voice. It's also within the range of frequencies where the ear is particularly sensitive (around 2000-5000 Hz). Here will also ideally be a 3db dip switch for around 3-4khz, to account for HRTF directionality, now with the crossover being where it is, this may be impossible, and I will have to find another way to mitigate this problem.

Not a bad idea.

There are a number of 4" drivers that offer wide bandwidth.

Normally it is considered that the "formant range" of music (what makes the "tone" of instruments) covers ~ 200Hz to ~ 5kHz.

A good 4" driver, perhaps used in multiples, is capable of covering this range.
A favourite Sound Reinforcement speaker of mine used something like that. A vertical array of 4 pcs 4" drivers with a HF driver in the center. Plus a 15" Woofer all arranged semi-coaxial.

The Geithain RL901k (one of the absolute best studio monitors bar non) uses a single 5" and Tweeter:

musikelectronic geithain gmbh - RL 901K

This uses crossover points of 550Hz & 2.8kHz. I'd like it a bit wider on the "midrange, but we are all getting into a 200 - 550Hz and 2.8 - 5kHz range of bandwidth.

It should be noted that the active crossover in the MEG RL901k is time-compensated, the system is equalised and it is all done in the analogue domain.

Tweeter (High Frequency Driver): Handles frequencies from 3500 Hz up to around 40,000 Hz. This range includes the highest notes of the piccolo and violin, as well as the harmonics of lower-pitched instruments and voices. A 1" dome tweeter would be a liekly choice for this range.

I would suggest that if you are able to push up the crossover point to the HF driver, using a 20mm ring radiator HF driver is a better choice. In fact, I think all ring radiators are better than domes.

Keep the distance between Mid & HF minimal, 7cm is the wavelength of a 5kHz sound wave.

Consider making a so called MTM (or d'Appolito) array, this gives symmetrical vertical sound radiation and provides some vertical pattern control.

This CANNOT be successfully be used with an even order crossover - you need a 3rd order ACOUSTIC or even better, a 3rd order acoustic highpass on the HF unit which incorporates the natural 2nd order highpass behaviour of the driver and a first order lowpass on the MF.

You may find the results preferable to a 4th order LR...

Another thing to consider is to get a coaxial Mid/Hf system. Seas have an option:

H1699-08/06 MR18REX/XF

A coaxial system gets around a lot of problems in the crossover region but often adds more problems.

A semi-coaxial system with the HF unit suspended before the midrange can reduce the issues from the interactions between cone and HF driver.

Everything is a compromise.

Woofer (Low-Mid Frequency Driver): Handles frequencies from 84.9 Hz to 350 Hz. This range includes the fundamental frequencies of many musical instruments and the lower range of male voices. A 6.5" or 8" woofer would be suitable for this range. It's above the highest note of the bass voice and below the lowest note of the soprano voice. It's also below the "sweet spot" for orchestral music (around 500-600 Hz).

I would go with dual 8" or even as large as dual 12", making a WMTMW Array (turn vertical). This extends the control of the off-axis control to even lower frequencies.

Now, for my "next trick", let us for a moment consider that the "box" is a big problem.

What if we drop the box, make a plan baffle of a suitable width (swat analysis -> 60cm) and use an "acoustic sump" type device (say a pipe of diaphragm diameter filled with basotect foam) on the back of the drivers to and delay attenuate the rear radiation of bass & midrange.

This now gives an adjustable cardioid response from HF to LF and happens to reduce significantly any room interactions.

Another way to do a cardioid midrange is shown by Amphion, it uses a more "conventional" box shape.

1. Subwoofer (Low Frequency Driver): Handles frequencies up to around 84.9 Hz. Given the long wavelengths of these low frequencies, a larger driver is needed to move the larger volumes of air. A 12" or 15" subwoofer would be good choices. The cutoff ensures that the entire bass vocal range is handled by the same driver, the woofer.

Now we can start considering the Sub. Our big panel, at 60cm width if using the right driver will be flat to ~ 120Hz (there is a bit more to it) and will then roll off very gently at less than 6dB/octave if our woofers have a Qt of 0.7 to the woofers open air or build in resonance.

This means a subwoofer will really only need to cover very low frequencies and can share a lot of load with or cardioid woofer, only taking over for very low frequencies.

I will still be using 4th-order-linkwitz riley, l-pad for the tweeter, zobel networks for impedance correction and some form of equalization for the bass. The enclosure for this will be more advanced than just a box, after having read more incl. what was available at linkwitzlab.

In your case it is impossible to get a 4th order acoustic response. Remember, with real drivers, the real acoustic slope matters, not the electrical crossover response.

BTW, There are other ways to make such a system as I described. One done for a customer used a pair of "Kilowatt" 12" drivers that had one driver placed in a sealed enclosure made from a concrete pipe standing up, with the driver on top and an open baffle shaped a bit like the Star Trek TNG communicator badge frontpart above that:



This baffle housed a 12" driver in normal orientation placed thus at < 7" distance between sealed and dipole driver and a so-called Manger transducer, an interesting driver in it's own right, that covers the audio range above 200Hz all the way to supersonic frequencies, around 8" diameter, with an acoustic sump behind it.



The wideband driver operated with a factory recommended passive highass and was driven by the customer "special" amplifier.

The bass section used DSP and four convection cooled (no fan's) Class AB Pro Audio amplifiers, bridged to give 1.6kW per driver.

The DSP was used to create a perfect cardioid response with 20dB rear output attenuation throughout the room's modal range and rolled out the dipole section below the modal frequency region of the room with the sealed boxed equalised to continue flat below 16Hz (8Hz cutoff) and the input signal was taken from the speaker level feed to the Wideband driver.

Compared to any conventional speaker - this system managed to transport the listener into the performance space and open a deep and wide window into the performance, in a way no conventional speaker I ever heard managed, excluding the RL901k that pulls a lot of the trick.

Anyway, take your idea, download Vituixcad and simulate it.

Your current idea is a lot more likely to be able to be made to work, you only need to get over the fixation on electric 4th order LR filters and you will have something interesting.

You could make something even more unusual and interesting, maybe time to go out and listen to some of the more interesting unusual speakers starting with the Geithain RL901k.

After you hear a wide mix of technologies from DSP/Digital Amp systems to passive and all inbetween, you might appreciate both what is possible with such systems and where the problems are.

Seriously, get out and listen and run some sim's before you spend money and start cutting wood.

Thor

 

[JohnRoberts]

Hi Adam,
dont feel overloaded by all the comments in this thread. All participants want to help you. And a lot of them have great experience in acoustics and electronics..
30% sound quality is done by the speaker itself, 70% depend on room issues.

For even more TMI, the iconic Bose 901 speakers only had one driver radiating directly toward the listeners, and 8 drivers reflecting off rear walls (11%/89%).

In theory this was similar to the ratio of direct/reflecting sound typical of concert halls.

Of course opinions vary, but Bose sold a lot of them.

[edit - for the rest of the story, the home brew loudspeakers I made back in the 70s were a variation on the 901 but using 16 drivers (2 facing the listener, 14 bouncing off the rear walls.) They required Active EQ similar to the Bose. Even with the EQ boost the top octave was lacking so I ended up adding piezo tweeters (without their horns). Not pretty but played loud and sounded OK -edit]

JR

Optimizing room acoustics is therefore more effective than optimizing the speaker only. As you explained your living room is 5x5 m. I suppose the room height is around 2,6 m. The lowest standing wave frequency is therefore around 34 Hz and gives a boost of at least 6 dB. In conjunction with a closed design with HP2-corner at approx. 55 Hz you can get a considerably flat response down to 30 Hz. For this construction a 6,5" or 8" bass chassis would be fully sufficient..
I strongly recommend startin with a "small" speaker design to avoid cost and frustration

 

[Adam Frandsen]

A lot to take in haha - will try and check out the speakers mentioned. Thanks for all the advice!

 

[thor.zmt]

For even more TMI, the iconic Bose 901 speakers only had one driver radiating directly toward the listeners, and 8 drivers reflecting off rear walls (11%/89%).

Indeed, ignoring for now the whole raft of B*se Jokes (No highs, no lows, it must be B...."), what Bose and many "omnidirectional designs achieve, in terms of spatial rendering they make it sound like the band or orchestra are transported (suitably miniaturised) into your front room.

I call that a "they are here" speaker. They do tend have excellent tonal balance in the midrange (more with EQ) both off and on axis, they fill a room with sound.

For contrast, let's assume a correctly made dummy head resording being listened to on a correctly equalised headphone, this delivers "I am there". You are transported into the recording venue.

Now let's take a speaker that does the opposite of B*se, namely producing 89% direct sound and 11% indirect sound.

The actual acoustics of the performance space are already embedded in the recording and just need to be "decoded". I call such speakers the "I am there" speaker.

One of the first choices to be made in speaker design is if we consciously want to maximise the room contribution and make a "they are here, but pretty small" speaker, or if want to minimise the room contribution and make a "I am there and how did my 200sqft front room end up sounding like a huge concert hall?" or something in-between and exactly where in-between.

After that we can work on achieving the goal.

Another caveat is that not all recordings are made equal. A minimalist miked acoustic recording can be a revaluation and out of body experience im "I am there" speakers, but a recording artificially created in and escaped from a lab cobbled together from different random instruments recorded in random spaces playing to a click track, is revealed as discombobulated mess.

So the ideal speaker in my view has near flat on axis response and freely adjustable DI (directivity index) and if required frequency dependent slope of DI, to adjust to both environment and recording and/or desired effect.

Thor

 

[JohnRoberts]

I'm not ready to venture down the rabbit hole of "stereophony".

This is a very successful audio illusion that we have enjoyed for many decades, let's not ruin it with over-inspection.

JR

 

[thor.zmt]

I'm not ready to venture down the rabbit hole of "stereophony".

This is a very successful audio illusion that we have enjoyed for many decades, let's not ruin it with over-inspection.

JR

Actually it is that "over inspection" by people like Blümlein that made it possible.

I'm very interested in this, as back in the 80's I was involved with research on this at the RFZ In East Berlin, where we learned a lot about this illusion, from recording to playback.

Even now, where popular music seems randomly assembled in a laboratory by mad scientist from random leftover body parts, before escaping and wreaking havoc on the charts, much of it remains valid and useful.

Thor

 

[mhelin]

About "stereophony" - is that something they call "immersive audio" these days?

 

[thor.zmt]

About "stereophony" - is that something they call "immersive audio" these days?

No.

You might call it an earlier version.

Proper stereo recordings played back properly can give a realistic illusion of a sound stage within which individual instruments can be located individually left to right.

Most of the research was done by Alan Dower Blumlein before WW2, he patented the long tailed pair and the ultra-linear amplifier circuit, various patents on Television as well as what he called 'Improvements in and relating to Sound-transmission, Sound-recording and Sound-reproducing Systems', laying the foundation for the Stereo LP and 2-Channel audio:



In WW2 Blumlein was central to the development of the H2S airborne radar system. He was killed in the crash of an H2S-equipped Handley Page Halifax test aircraft while making a test flight for the Telecommunications Research Establishment (TRE) on 7 June 1942.

Thor

 

[JohnRoberts]

About "stereophony" - is that something they call "immersive audio" these days?

nah, immersive sound might be talking about surround sound, anywhere from 4 to a dozen loudspeakers spaced around the listening room. Back in the day I sold a delay based surround sound decoder kit (phoenix systems). The old Dolby surround algorithm was basically subtracting L-R, delaying that difference signal a few tens of mS and playing back through one or two rear speakers. Modern surround sound is far more sophisticated using multiple discrete signals.
===
By "stereophony" I was referring to the use of two loudspeakers for playback, being driven by different signals, aka stereo. Pretty much anything not mono.

The human brain works overtime to make sense of sounds that it hears. The dominant difference with stereo playback is that sound sources can appear to be localized (coming from) somewhere other than directly from the two loudspeakers (the brain uses loudness and arrival time cues to localize sound sources). Pan pots in mixers generally proportion signal level between L and R to create the sense of localization. Some more sophisticated consoles experimented with adding delay too, but those are not common.

As I was trying to say, IMO it may be better to not overthink this, many have done exactly that for decades. If interested in going down that rabbit hole, maybe search "psycho acoustics". Lots of interesting papers published in early AES journals.

JR

 

[thor.zmt]

nah, immersive sound might be talking about surround sound, anywhere from 4 to a dozen loudspeakers spaced around the listening room. Back in the day I sold a delay based surround sound decoder kit (phoenix systems). The old Dolby surround algorithm was basically subtracting L-R, delaying that difference signal a few tens of mS and playing back through one or two rear speakers. Modern surround sound is far more sophisticated using multiple discrete signals.

Actually, the game "immersive sound" system is build Michael Gerzon's extension from binaural to ambisonics.

It literally use the Ambisonics multichannel format.

By "stereophony" I was referring to the use of two loudspeakers for playback, being driven by different signals, aka stereo. Pretty much anything not mono.

Hmmm, that is a bit too generous.

The human brain works overtime to make sense of sounds that it hears. The dominant difference with stereo playback is that sound sources can appear to be localized (coming from) somewhere other than directly from the two loudspeakers (the brain uses loudness and arrival time cues to localize sound sources). Pan pots in mixers generally proportion signal level between L and R to create the sense of localization. Some more sophisticated consoles experimented with adding delay too, but those are not common.

Again, this is well known and understood.

A phantom sound source that is centered between the two speakers (1) as starting point can be shifted towards the left or right speaker by making one channel louder or by having a time delay between the channels.

Both terms are frequency dependent. Below around 700Hz time delay differences are relatively more important than amplitude differences, above 700Hz amplitude differences become dominant and time delay begins to be perceived as delay/reverb rather than directional cues.

It also follows that panpot "stereo" or single point minimally miking (blumlein crossed 8, ORTF Array etc) that relies only on level leads to distortion of the perceived space/location (the wandering instrument), unless at frequencies below ~ 700Hz the level difference is progressively increased (see Blumlein "stereosonic shuffler"), which of course makes it the opposite of "mono compatible" in mixes.

One big problem with many recordings these days is that they give a hilarious dummy spacitial presentation, because what I just compressed into three paragraphs is not understood by many "sound engineers".

They do "glue" in mastering by clipping, compressing and adding harmonics, but fail to "glue" the space and to ensure their mixes are spatially stable.

As I was trying to say, IMO it may be better to not overthink this, many have done exactly that for decades. If interested in going down that rabbit hole, maybe search "psycho acoustics". Lots of interesting papers published in early AES journals.

The spatial presentation of a recording unless Mono is an important aspect.

Thor

1) A trick to see if your speaker system is good at "stereo" is to play white noise equal in both channels ("mono") with both channels in phase - the virtual sound source of the noise should be a minimal size point midway between the loudspeakers.

Then invert one channel, the sound should not have any perceivable source of sound, OR in rare cases (speakers with especially well controlled dispersion) appear to come from behind your head as tightly focused spot.