Fas42’s Stereo ‘Magic’

[fas42]

7 hours ago, John Dyson said:

A more simple and straightforward explanation why speakers make a more profound difference -- more significant degrees of design freedom and more significant impact of environment (and probably a few other things I haven't thought of.)

An amplifier might have a -0.25dB error at 50kHz, and totally flat below that, even at full power.   A speaker with that kind of quality (probably impossible) still does not create a 'totally flat' frequency response WRT the listener.   A very very good speaker might provide +-5dB at the listener because of the transducer and environment.   The averaged response might be better -- but I am speaking of the narrowband frequency response.

 

Why does a speaker have to have, flat response? ... Think about the real world - anything that makes a noise, or musical sounds, can have widely varying acoustics around it, which changes the FR enormously, in terms of how the signature of that sound reaches our ears - but our hearing instantly groks what the sound is all about- "it sounds normal!". But the loudspeaker is some sort of mystical object, that needs this great degree of precision about it, to achieve the same effect - in the world of the audiophile ... why?

 

My experiences answer the why ... when the sound reproduction is "good enough", then the same mechanism that deals with the real world sounds kicks in - and a wildly erratic FR also sounds normal ... we mentally adjust for the 'wrong' curve, completely unconsciously - and the replay sounds, well, natural. OTOH, when the reproduction of such sounds is sub-par, then every tiny detail of any misdemeanours screams at us - and, we're not happy, 😉.

 

Quote

 

An amplifier might have 0.001% of harmonic distortion over an entire power range, with any expected load, and 0.002% IMD over an entire power range with any expected load, but NO speakers nowadays can do that (I mean, at normal listening levels, even super low levels.)  A speaker that does less than 1% distortion over the entire frequency/power range is still very unlikely, even with feedback methods.

 

A more realistic figure is 0.1% distortion - I remember loudspeaker measurements in audio magazines 30 years ago, that actually did measure things, like harmonics generated - and it wasn't that hard to get that 0.1% figure over where it really matters, the midrange and treble. Bass is always the nightmare - it takes a lot of money, and effort, to get good figures there ... which is why people use speakers that do lots of bass wally to season their sound tonality - plenty of incorrect harmonics to add one's favourite signature ... 😜.

 

Quote

 

There are so many ways that speakers and their environment (setup, etc) are such a challenge.   Even microphones are easier than speakers.  Use of microphones, maybe difficult -- but even then -- speakers are a mega challenge.   That is why, for precision work, I have to use headphones.  It isn't for accuracy alone, but also coupling and relative lack of variations.

 

Like any large transducer like a speaker, it will have characteristics that produce a 'sound', even a perfect speaker will be dependent on the environment.

 

John

 

 

My experiences have constantly shown me, that as a playback chain improves in the region prior to the speakers that the subjective experience approaches only hearing the characteristics of the particular recording - and that the speaker only plays a very minor role in varying this. If a rig, no matter how expensive, imparts a strong, consistent flavour to the proceedings - then it's wrong; and this is made even more obvious if changing the speaker completely alters the presentation - "transparency" is miles from being in the picture.

[fas42]

Just read this,

 

As good an example as any of what making the right moves achieves ... there's a telling comment there, which underlines how I see audio people always approaching this business "the wrong way". Doug says,

 

Quote

I made several significant changes to my stereo over a three-month period, and it is hard to tell which items did what.

 

Ummm, the need to hang on to the Adding Goodness stance is embedded deep in the audiophile's psyche - which means that everything has a specific role in contributing to what you hear ... the "did what?" question is uppermost in the mind, 😝. If one can drop that thinking, and just see everything as a journey in Subtracting Badness, then it all becomes very obvious ... the SQ snaps into shape when the last, crucial, anomaly in the presented sound is sufficiently attenuated - there is zero point in ascribing qualities in what you hear to the Goodness of some added bit ... all you want is the absence of any disturbing artifacts in the replay, which may be achieved in myriads of ways.

[John Dyson]

2 hours ago, fas42 said:

 

Why does a speaker have to have, flat response? ... Think about the real world - anything that makes a noise, or musical sounds, can have widely varying acoustics around it, which changes the FR enormously, in terms of how the signature of that sound reaches our ears - but our hearing instantly groks what the sound is all about- "it sounds normal!". But the loudspeaker is some sort of mystical object, that needs this great degree of precision about it, to achieve the same effect - in the world of the audiophile ... why?

 

My experiences answer the why ... when the sound reproduction is "good enough", then the same mechanism that deals with the real world sounds kicks in - and a wildly erratic FR also sounds normal ... we mentally adjust for the 'wrong' curve, completely unconsciously - and the replay sounds, well, natural. OTOH, when the reproduction of such sounds is sub-par, then every tiny detail of any misdemeanours screams at us - and, we're not happy, 😉.

 

 

A more realistic figure is 0.1% distortion - I remember loudspeaker measurements in audio magazines 30 years ago, that actually did measure things, like harmonics generated - and it wasn't that hard to get that 0.1% figure over where it really matters, the midrange and treble. Bass is always the nightmare - it takes a lot of money, and effort, to get good figures there ... which is why people use speakers that do lots of bass wally to season their sound tonality - plenty of incorrect harmonics to add one's favourite signature ... 😜.

 

 

My experiences have constantly shown me, that as a playback chain improves in the region prior to the speakers that the subjective experience approaches only hearing the characteristics of the particular recording - and that the speaker only plays a very minor role in varying this. If a rig, no matter how expensive, imparts a strong, consistent flavour to the proceedings - then it's wrong; and this is made even more obvious if changing the speaker completely alters the presentation - "transparency" is miles from being in the picture.

Generally answering the question -- if the response to the head isn't accurate, then the presentation isn't as accurate as it can be.   Why even bother with frequency response on amplifiers at all?   Also I also do NOT believe that very many speakers have an actual 0.1% total distortion at  reasonable/real listening levels and very seldom there is an actual flat response without coloring.  The physics are just too complex. 

 

I don't believe in 'tweaking' design without careful measurement -- man, I have gotten a big dose of that problem.  I do believe in disciplined understanding and resolving of issues, or it becomes a continual game of 'whack a mole'.   Tweaking itself can be a hobby along with listening to recordings -- and I suspect that is the hobby of some people reading this now.   If 'tweaking' is the hobby, then 'whack a mole' comes along with it -- and if being honest about it, then it is certainly okay.

 

The physics distortion problem is more of an issue at lower frequencies because of the huge excursions creating significant FM/PM, and it is important that the LF and MF/HF don't mix together.  The multiple causes for distortion is the close secondary reason for multiple transducers.   Then, as soon as you have multiple transducers, then you have all kinds of phase cancellation effects, beyond just the separation of the front and back on the transducer.  (Ever notice that a lot of microphones have long tubes in them?   The purpose is to keep the LF in the front and back from cancelling.  Phase issues are even a problem in microphones.)   This isn't just a problem in the past, but we do have more clean/accurate modeling tools today to tweak the designs better.  Some response EQ can also be done (both phase/timing/freq), but still don't really solve -- just help mitigate.

 

I am not claiming that all hope is lost, but designing a reasonable speaker (and that is about as good that can be done) is tricky, and 0.1% distortion (really) is not likely in the freqs where hearing is most sensitive, except for significantly low listening levels.   There is just too much opportunity for nearby LF to modulate the diaphragms for higher freqs.   The higher frequencies IMD each other also, but become less and less a problem as freq increases.   This is because of the nature of diaphragms.  (I'd suspect that feedback schemes can help the situation a little.)

 

Now, LF distortion by itself is a reasonable place to say 'who cares', because probably EVERY recording in most peoples collection have lots of LF/lower MF distortion from complex gain control.

 

All I know is that as soon as one frequency causes variation in another frequency (including doppler changes), distortion will ensue, but NOT necessarily in the form of the gritty/grainy stuff, but instead a softening of detail by stealthy FM.   I'd suspect that most people dont' truly understand that intermod occurs just as a matter of physics, not easily designed out at all -- and is a matter of fact about moving air with a diaphragm.    It is almost like 'speed of light' wrt travel speed limits -- it just IS.

 

(Avoiding diaphragms can probably mitigate the fuzzing/modulation problem.)

 

So, I have a very disciplined view of any part of a system design -- speakers are the most tricky and actual *technical* quality similar to other parts of the system are IMPOSSSIBLE because of all of the effects due to the sheer energy/signal levels/mass/flexing/doppler,etc.   Other transducers are also somewhat tricky wrt reproduction quality, but NOWHERE near speakers.

 

Bottom line:  speakers do color the sound -- it is all about the coloration that you like.   More expensive speakers might do better, or might color differently.  It is all about what YOU like, which can be different from what others like.

 

John

 

[fas42]

40 minutes ago, John Dyson said:

Generally answering the question -- if the response to the head isn't accurate, then the presentation isn't as accurate as it can be.   Why even bother with frequency response on amplifiers at all?   Also I also do NOT believe that very many speakers have an actual 0.1% total distortion at  reasonable/real listening levels and very seldom there is an actual flat response without coloring.  The physics are just too complex. 

 

In the early days, Hi-Fi News did testing like this, and currently this mob have organised an ongoing testing regime ... https://www.soundstagenetwork.com/measurements/speakers/thiel_scs4/. Note the distortion is being measured at 90dB, which is well above what anyone would listen to a pure tone at.

 

Quote

 

I don't believe in 'tweaking' design without careful measurement -- man, I have gotten a big dose of that problem.  I do believe in disciplined understanding and resolving of issues, or it becomes a continual game of 'whack a mole'.   Tweaking itself can be a hobby along with listening to recordings -- and I suspect that is the hobby of some people reading this now.   If 'tweaking' is the hobby, then 'whack a mole' comes along with it -- and if being honest about it, then it is certainly okay.

 

Many people may try a "whack a mole" approach, but that won't lead to improved playback of all the recordings one has. Measurements are not necessary if one has learnt how to listen for flaws in the playback - you are not trying to make it "sound better"; rather, you recognise a defect in the chain, confirm what's going with some simple experiments - and then do what's necessary to resolve it. For example, most audio chains are far too sensitive to noise on the mains supply - something which is trivially easy to check: you deliberately introduce far greater noise on the supply, and then/or disconnect every possible source of noise in the area - virtually all systems will demonstrate quite significant audible changes when doing this ... you now know you have an 'issue' - up to the individual to decide whether they're going to do something about that.

 

Quote

So, I have a very disciplined view of any part of a system design -- speakers are the most tricky and actual *technical* quality similar to other parts of the system are IMPOSSSIBLE because of all of the effects due to the sheer energy/signal levels/mass/flexing/doppler,etc.   Other transducers are also somewhat tricky wrt reproduction quality, but NOWHERE near speakers.

 

Bottom line:  speakers do color the sound -- it is all about the coloration that you like.   More expensive speakers might do better, or might color differently.  It is all about what YOU like, which can be different from what others like.

 

John

 

 

People's hearing will always differ, as to what they're sensitive to. My experience, as has been the reaction of others who listen, is that it's possible to for a normal playback system to deliver the same, subjective listening experience as when hearing live music - I go with what is achievable, rather than attempt to prove that such is impossible, by use of technical arguments, 🙃.

[fas42]

Noting that Amir at ASR has looked at a very low cost Edifier - and liked it, 😉. Measurements numbers are very ordinary, but, it gets what counts right ... and isn't that what it's all about? 😁 Looks like the company has a good handle on making the best of the raw ingredients - which often goes missing in action, with components in the audio world 😜.

 

What I really wanted to point to was this, by Amir,

 

Quote

Not really. The imaging is apparent with a single speaker just as well. Try it.

 

He was ridiculed in the superbestaudiofriends forum for saying this - but, this is indeed the case. What matters is not that you are getting signals from two different locations, but, the clarity of the sound. When low level detail is clearly presented, with minimal crud injected from the playback chain, then the acoustics of what's happening in the captured event make sense to the listening mind, and 'imaging' occurs.

 

Getting back to excellent value Edifier speakers, that sort of detail in the resolution of the music in the recording is there in spades for my set of that brand's speakers - even being directly in front of the speaker on one side, the sound is completely behind the drivers, with the sense of each sound element in the mix being fully distinct ... not invisible speakers, so far; but damn good considering how little tweaking has been done to them, 🤪.

[fas42]

Noting this new thread,

 

 

Don't want to lead a new member astray, 😁, but will just mention here that the concept of systems suiting certain music is intrinsically absurd ... if a playback setup is getting the sound correct, that is, being true to what is on the recording, then the content is irrelevant. Provided the rig has the capability of playing the track at the volume that the listener wants without any limits being reached - an obvious example would something with very high level low bass frequencies, which exceed the ability of the driver to reproduce them, as far as the physical travel of the cone is concerned - then there shouldn't be issues.

 

Of course, if the listener wants a highly distorted version of the track; to match how such music comes across on a concert PA with all the meters pegged in the red, then all bets are off. For this, just get a domestic version of a PA speaker monster which can tolerate being hammered to death; which will still smile and ask for more, after being abused for hours on end, 😉.

 

For myself, I want heavy metal style music to be 100% true to what's on the album - which especially for older material - I can't speak for recent, heavily compressed stuff 🙂 - will be just as enticing to listen to as the most ladylike classical items. A system should be able to go from from a full throttle, guitars driving to the max, track to a plaintive, solo violin piece without missing a beat ... this is the sort of 'test' I use on my setups all the time - to be able to be true to any sequence of musical styles I toss its way; if something falls short, that means the system is not, ahem, fully sorted 😉.

 

That said, if in a shop checking out gear, I would use a heavy metal track at high SPLs to check if the basics are there - if it couldn't do such without collapsing in a heap then I would immediately pass on it; too much is wrong to warrant the work needed to lift its game.

[fas42]

Dud 2nd post ...

[fas42]

Noting Peter has upgraded his USB cable,

 

Of course, this is a commercial product, so Peter has to be enthusiastic about its abilities, above what he was selling before - but can just a USB cable miraculously turn recordings into special listening moments ... elevate the most mundane album into a truly wondrous presentation - of course not!! Will Lush^4 make it 10 times better than this one, and then Lush^5 10 times better again - well, what do you think ... ?

 

So, what's going on? Well, it's a means for getting closer to the true sound of the recording - which means the law of diminishing returns has to kick in at some point ... once you have only the qualities of the recording apparent, then anything further is seasoning ... à la, tube rolling fun, 😉.

 

Meaning, the properties of the cable are actually the properties of the recording - which is getting closer and closer to having its intrinsic nature 'revealed'. ... And they are,

 

Enormously spatial  ... tick

Robustness ... tick

Realism, as in convincing cymbals ... tick

'Blackness' ... tick

Natural ... tick

 

The difficult bit is that often the closer one is to the very best in one 'quality' aspect of the replay, that the slight misdemeanours in other senses are now "too visible" - as per Peter's example, when space is there in spades, that there is a hardness at some points. This is not a backwards step; it just means that one has to dig even deeper to locate the cause of the hardness - which may originate from an entirely different area ... the progress to optimum replay just continues, an almost endless journey ...

 

 

 

 

 

 

[fas42]

A short comment, about transients ...

 

Getting the punch of impulsive sounds coming across correctly has absolutely zero to do with the digital processing side of things; playing with sampling rates, and all sorts of cute filtering doodahs, has nothing whatsoever to give, in helping getting the subjective qualities right in this area.

 

The best example is our early 90's Yamaha synthesizer; from cold the acoustic piano sound of this is atrocious - transients, what transients??!! Yet, leaving it running for 3 days solid conditions the hybrid and analogue circuitry - and, it sounds like the real thing. In that period not one iota of the digital processing altered; but, the transient impact of piano keys being struck hard is fully in place - quite impressive.

 

IOW, getting the transients rights is an analogue thing ... not, a digital thing !

[fas42]

I'm mighty pleased to have come across this ... the first, serious research paper that I've come across that delves into the capability of 'simple' two channel stereo to throw up convincing sound images, http://boson.physics.sc.edu/~kunchur//papers/Stereo-height--Kunchur.pdf. Interestingly, it investigates whether vertical positioning can be conveyed, to a listener, and it was confirmed that indeed this was possible ... finally, that one can be put to bed, 😁.

 

Nice to see something like

 

Quote

There is a category of realistic sounding systems—sometimes referred to as “high‐end
audio” (which I will abbreviate as HEA)—that lies far beyond mainstream‐consumer audio.
Carefully setup and tuned HEA systems are extremely rare, and most people interested in
music are not aware that this level of audio reproduction exists.

 

actually put down in print, in a "proper paper", 🤪.

 

[fas42]

Nice discovery ... chap over in the Russian regions who was getting non-constipated SQ back in 2012 - showing that rock can be delivered in the home, with good impact, and without mangling the clarity of the instruments ... note how the cymbals come across,

 

 

 

[fas42]

Just had to share it ... this, gentlemen, is how a full bore audio sound enthusiast goes about it, 🤣

 

https://www.whatsbestforum.com/threads/a-visit-to-mik-bhavnani-unique-audio-uk.29414/post-678926

 

Reading the other posts, the owner of this menagerie does indeed get special sound - and has no hesitation whacking a bit of weight somewhere using Blu Tack ... a man after my own heart, 🙂.

 

[fas42]

Just noting this post, elsewhere, https://www.whatsbestforum.com/threads/taiko-audio-sgm-extreme-the-crème-de-la-crème.27433/page-58#post-596794. Key sentence,

 

Quote

The ... continue to give me the best midrange density and least high frequency glare

 

"Glare" is code for hearing disturbing distortion - competent replay removes this entirely from the equation, for all recordings - unless such was intended, as part of recording and mastering decisions 🙂. The process of debugging a playback chain has as a key element finding those recordings which provoke "glare", and then using that as a tool for finding the problem ... that unpleasantness is always caused by a problem, and the chances of fixing it by throwing 'decrappifiers', and expensive hardware in other areas, at the situation is not particularly good ... smartest, and usually cheapest move is finding the real cause of the less than optimum SQ, and properly resolving that.

[fas42]

Note Peter's latest fiddling of USB cables,

 

This exemplifies what the tricky area in digital playback is about - being able to deliver very intense levels of sound from instruments which if poorly reproduced sound obnoxious, or very dull and uninteresting. Cymbals are as good as any for this purpose, and anyone who hasn't heard what's possible in audio playback will never really understand what's being discussed, 😉.

 

At high SPLs, all the distortion artifacts which can be swept under the carpet in subdued listening situations have nowhere to hide, and the true accuracy of the rig is fully on show - get it right, and all the subjective intensity of live music is in the room; if it is even slightly off colour, it just doesn't jell - and the volume has to be dropped, to make it palatable ...

[fas42]

Good post in the WBF forum, about reaching a point where "the improvements can be truly staggering",

 

https://www.whatsbestforum.com/threads/what-is-the-benefit-of-very-expensive-dac´s.12037/post-679706

 

Indeed, "there is a tipping point in the evolution of a system where it begins to give back more in terms of performance than expected with a given upgrade" ... this is the 'land of magic', and is a wondrous place to visit. What's even more rarely understood, 🙃, is that big bucks are not necessary to go here - awareness that it's within reach of most mere mortals is the starting point, in the thinking, if one wishes to consider such a journey ... 😉.

 

[AnotherSpin]

2 hours ago, fas42 said:

Good post in the WBF forum, about reaching a point where "the improvements can be truly staggering",

 

https://www.whatsbestforum.com/threads/what-is-the-benefit-of-very-expensive-dac´s.12037/post-679706

 

Indeed, "there is a tipping point in the evolution of a system where it begins to give back more in terms of performance than expected with a given upgrade" ... this is the 'land of magic', and is a wondrous place to visit. What's even more rarely understood, 🙃, is that big bucks are not necessary to go here - awareness that it's within reach of most mere mortals is the starting point, in the thinking, if one wishes to consider such a journey ... 😉.

 

 

Certainly, we are talking here about the balance of efforts+resources vs. result. It is desirable when the result exceeds expectations. The right way to do this is to reduce effort+resources to a minimum and give up high expectations. The relationship to the result will definitely change in favour of the second. 🙂

[Blackmorec]

Frankly (s’cuse the pun Frank) I don’t know why certain people want this thread closed. IMO there’s still some excellent information. I recognise the dichotomy between Frank’s low cost system (note I didn’t say low-fi) and his lectures about better sound quality. On the other hand, his experiences don’t contradict what I’ve found and am still finding. Close this thread and you’re definitely throwing the baby out with the bath water. 

[Rexp]

1 hour ago, Blackmorec said:

Frankly (s’cuse the pun Frank) I don’t know why certain people want this thread closed. IMO there’s still some excellent information. I recognise the dichotomy between Frank’s low cost system (note I didn’t say low-fi) and his lectures about better sound quality. On the other hand, his experiences don’t contradict what I’ve found and am still finding. Close this thread and you’re definitely throwing the baby out with the bath water. 

Agreed. I'd take Franks audio advice over yours any day. 

[Blackmorec]

10 minutes ago, Rexp said:

Agreed. I'd take Franks audio advice over yours any day. 

Well there’s some progress! Well done 

[Rexp]

8 minutes ago, Blackmorec said:

Well there’s some progress! Well done 

Thanks to you too for starting this thread, you should be proud. 

[Blackmorec]

16 minutes ago, Rexp said:

Thanks to you too for starting this thread, you should be proud. 

Thank you. Appreciated.  

 

Frank’s advice is valuable to anyone who doesn’t want to invest too heavily, whereas my advice and experience is quite a bit errrr..less economical. 😈

[Blackmorec]

Duplicate again....🤡

[fas42]

A good point here,

 

There is,

 

a) The urge to turn it up - the listening mind recognises that the experience now much more closely mimics how sound in the real world comes across, and if this is something pleasurable then the instinct is to immerse oneself in it, more fully ... but what defines, "much more closely"?

 

Answer is,

 

b) A lot less distortion - irrespective of how many thousands of measurements can be extracted to "prove!!" that any distortion is inaudible, the ear/brain knows a lot better ... 😉. Irritating artifacts in the sound, disturbing giveaways that the SQ is not up to scratch do all the damage - and you have no desire to increase the volume to more realistic levels.

 

This is something that the industry still hasn't come to grips with, even though the components are getting better and better - meaning, that lower and lower cost items get enough right so that only a small amount of tweaking is needed to get the quality over the hump - and "magic" happens in the room. The recognition that the overall integrity of the system is paramount, and must be of a necessary standard for the system playback chain to 'disappear' is something that eventually will happen - but will it be in my lifetime? 🤪, 🤣 ...

[fas42]

"Are we there yet?" ... as regards just sticking in an exceptional measuring unit in a chain, and thus delivering exceptional sound ... umm, no.

 

This post says where it's at,

 

I was hoping that the Purifi Eigentakt modules would be well enough engineered to not need further tweaking, but this says otherwise,

 

Quote

The Purifi amplifier, for example, has an initial "wow" factor, but in spite of the stellar measurements, just does not provide a lifelike reproduction. I tested it with various preamps as well as without a preamp on its low-gain setting. 

 

Alas, this means that a rig has to be cajoled into higher performance with significant tweaking by the owner, if using 'ordinary' implementations of the Purifi circuits; or, wait for highly optimised components using these modules to be developed by other companies.

[fas42]

I found this clip of the rig that Chris listened to working,

 

 

and this demonstrates how the qualities of a system in good shape can easily be captured in a simple recording. The sense of space, the sound elements in the mix, and connection to the musical event come across easily - the first test of whether a setup is working well.