Thread: A Search for Truth and Tonality, Part 2 ...

[Soundproof]

I have to thank Amir for the link to this audio podcast of an interview with Floyd Toole. at about 7:30-11:00 into it he claims you cannot duplicate the sound of an original performance, it can't be done says he.

http://twit.tv/show/home-theater-geeks/14

John Atkinson editor of Silly-o-phile Magazine says the same. Listen to what he says from 28:00-32:00. In his live versus recorded demo the sound from the recording didn't have the "bigness" of the pianos. Of course it didn't. One look at the speaker design and judging the sound fields they propagate would have told him why. I think I'd go into cardiac arrest if an electrical engineer (Toole is, Atkinson isn't) ever demonstrated to my satisfaction that he knew what a vector is.

http://www.youtube.com/watch?v=_mEsuKqj5wA&feature=relmfu

Oh what an admission to concede that no matter how much money is spent, even this relatively simple aspect of the problem of high fidelity sound recording and reproduction has beaten them. And what arrogance to think that it cannot be solved by people far more clever than the two of them combined. Hahahahahaha. That problem is the tip of the iceberg. For not only can they not recreate the tone of the piano in the same room as they are, they could hardly appreciate the much more difficult problem of recreating the tone at a large live venue where it isn't the tone of the instrument as it would be heard in your home but in a concert hall where its tone is entirely different. Why is it different? Because in the typical concert hall sounds at 8 khz will die out at about twice the rate as they will at 1 khz and since the reverberant field represents the overwhelming preponderence of the energy you hear in every seat in the audience, that change affects the perceived tone. Without recreating the reverberation you can't recreate the tone, it is a dynamic event, they are part and parcel of the same phenomenon. In short, if you need to recreate the tonalities heard at a large venue, whatever the frequency response of your sound system is.....it's wrong.

Here's another tougher problem than F&A can't solve. Where a sound system will make a grand piano sound like a large source in your listening room, will it make a recording of a human voice sound as big as a piano too? The answer is no, it's still the size of a human voice. After 4 years of experimenting with it, I haven't figured out why yet, all I've got are some hunches.

Good post indeed, it goes to a point I have been enthralled by. Why do mono recordings with a low frequency cut-off sound as good as they do?
Are we, in fact, going down the wrong path with higher frequency extension, damping of first reflections and a negation of the natural reverberant field in dominant recording techniques?

My answers to that three-part question are Yes, Yes and Yes, in case anyone is curious.

 

[NorthStar]

Sorry, Bob, didn't respond to this before: see http://en.wikipedia.org/wiki/Being_There -- Peter Sellers as "Chauncey Gardiner".

Frank

Gotta check that one; right out of my mind at this instant I cannot remember if I saw it or not.
...But I think I did. ...Or perhaps not.

 

[Soundminded]

My first good speakers, years ago, were Goodmans Axiom 301s, did my own box for them: here you have 98dB sensitivity, 20W power handling. This translates to peak SPL of 116dB for a pair at a metre, and these units were manufactured in the '60s -- how much progress has been made to date? They were magnificent beasts, with a huge magnet structure: makes a Peerless driver look wimpy!!


That's interesting ... Have you ever tried recording and analysing the frequency spectrums to compare why there may be such a difference between those two positons?

Frank

I don't have to. The spectrum will be the same. That's why their tonality is the same. It's been carefully designed and adjusted to produce that result. What's different is the way the energy is propagated and the resulting distribution in time and space. That makes all the difference in the world. It's why they sound different. That's what the math model I created 38 years ago explained, exactly what's happening in regard to that. Understaning why they are perceived differently is something I've been pondering for 36 years. I have some interesting tentative conclusions I might publish one day.

Atkinson was right about the brain modeling sound but he didn't even scratch the surface and his instinctive conclusions are wrong. BTW, radically altering the model can take up to a second or two, even more for the brain to process and integrate, not a tenth of a second as he said. Here's an experiment you can try yourself. See what results you get and how it compares to mine. Play a loudspeaker as you normally would. Play program material that contains all frequencies including high frequencies. Allow yourself to become accostomed to the sound. Now play just the tweeter. Listen to it for a minute or two. Now cut the woofer back in suddenly so that it's a full range system again. If your own reaction to it is like mine you will find that for the first second or two you will hear two distinct sounds, the tweeter and the muffled woofer (and midrange if there is one) before your brain re-integrates them as a single source. Now try it the other way around without the tweeter. Then cut that in. The sound integrates immediately. If you try it, see if you get the same reaction I did.

Toole talks about using headphones. Anyone with a technical education who has not come to understand why the sound from headphones is radically different from the sound of live music and sound from loudspeakers and is therefore almost entirely useless for research in this area should tear up his college diploma and get a refund of his tuition, his professors failed to properly educate him. Also, any system which tries to use multi-binaural recording through headphones to simulate hearing to try to externalize the source will have to arrive at the corrected field continuously as ones head is turning not in milliseconds but according to data I've seen within 2 to 5 microseconds to produce the desired result. And it will have to be in all planes. Anything worse will likely produce a very confused distorted sound that will be nothing like what is desired.

 

[NorthStar]

-----Another excellent post Soundminded.

 

[fas42]

I don't have to. The spectrum will be the same. That's why their tonality is the same. It's been carefully designed and adjusted to produce that result. What's different is the way the energy is propagated and the resulting distribution in time and space.

So you're positive that the spectrum will be same, firstly midway between the two sound sources, and secondly when measuring outside the room -- even though the reproduced piano outside the room does not have the same volume? When you say the volumes match when midway between the piano and the speakers, is this done totally subjectively, or using a sound level meter? Why I say that, is because the behaviour of the ear/brain mechanism means that the subjective impression of volume can vary greatly, even though the measured intensity remains the same.

Here's an experiment you can try yourself. See what results you get and how it compares to mine. Play a loudspeaker as you normally would. Play program material that contains all frequencies including high frequencies. Allow yourself to become accostomed to the sound. Now play just the tweeter. Listen to it for a minute or two. Now cut the woofer back in suddenly so that it's a full range system again. If your own reaction to it is like mine you will find that for the first second or two you will hear two distinct sounds, the tweeter and the muffled woofer (and midrange if there is one) before your brain re-integrates them as a single source. Now try it the other way around without the tweeter. Then cut that in. The sound integrates immediately. If you try it, see if you get the same reaction I did.

Unfortunately, not easy to do! My "main" system is essentially full range, with supertweeters, and single subwoofer. I've run it many times without subwoofer, and the sound is still complete. The subwoofer (about 180Hz down) integrates very nicely with the rest, I don't know whether such a test would be relevant for such low frequencies.

Toole talks about using headphones. Anyone with a technical education who has not come to understand why the sound from headphones is radically different from the sound of live music and sound from loudspeakers and is therefore almost entirely useless for research in this area should tear up his college diploma and get a refund of his tuition, his professors failed to properly educate him. Also, any system which tries to use multi-binaural recording through headphones to simulate hearing to try to externalize the source will have to arrive at the corrected field continuously as ones head is turning not in milliseconds but according to data I've seen within 2 to 5 microseconds to produce the desired result. And it will have to be in all planes. Anything worse will likely produce a very confused distorted sound that will be nothing like what is desired.

In one sense I would have to beg to differ: perhaps my hearing is different from most others but the sound through headphones, even very high quality ones, matches the speakers in the areas of quality that I'm interested in. The experience of using headphones I find not particularly involving, in fact it irritates quite rapidly, perhaps because I enjoy the extra qualities brought to the soundscape by the reverberant sound in the listening room.

Frank

 

[fas42]

From http://www.whatsbestforum.com/showt...-the-bandwagon&p=111319&viewfull=1#post111319 ...

None of those limitations have to apply to digital recording, although to realize its potential there needs to be both some interest and some more resources applied to its improvement.

Both parties in this "argument" are missing the real issue: it's not recording or formats that will "fix" digital, it's the playback environment. DSD sorta fixes the problem because it goes around some of the issues: the way the electronics are doing their job when playing DSD from a "good" decoder makes the sound better, not the fact that it's DSD per se.

As a simple solution to getting top notch digital: convert all PCM music material to DSD using high quality software translation. Then playback through a high quality DSD device -- problem solved ...

Frank

 

[fas42]

From http://www.whatsbestforum.com/showthread.php?6491-It%E2%80%99s-All-a-Preference&p=111308&viewfull=1#post111308 ...

When we get to the playback end which will and does introduce distortions of their own, we users have a tiny bit of control. Fidelity is an ideal. It is not a reality at this point in time.

Sorry, Jack -- wrong!! We have a huge level of control: a so-called lousy recording can sound excrutiatingly unpleasant, impossible to listen to, which in my experience has always been on mega-expensive equipment, to quite magnificent, totally "believable; and that is purely a function on how well the system has been set up. At times I have been amazed at what a total mess so-called high end audio rigs have made of some of my difficult recordings, firstly by slicing off huge chunks of the information that's there, and then adding layers of either honey or obnoxious, hyper exaggerated detail to what's left!

Fidelity is here and now, and has been for years; the big trick is to carefully enlist the good bits of ear/brain psychoacoustics behaviour to make it happen ..

Frank

 

[Soundminded]

Hi Frank. I judged the sound levels to be the same halfway between by ear and outside the room also by ear. I didn't make any measurements. That would not be easy since the loudness is constantly changing. It would have to be the same passage played again and again and recorded on a storage scope or data logger. Anyway it wasn't necessary.

In principle Toole was right when he said in the interview that the piano radiates sound in all directions but not to the same degree. In actual fact though he was quite wrong. The sides of the case is an inch thick and it isn't made of particle borard or plywood. It's made of something much harder and denser, in my case it might be mahogany. little if any sound gets through it. What does that mean. With the top lid closed but folded back on its hinge and me seated I'm not in direct line of sight or direct earshot of the strings, the hammers, the harp, the sounding board. In short ALL of the sound I hear from the piano is reflected off the walls, the floor, and ceiling first before it reaches me. This is why it sounds so big and powerful, it literally fills the entire half of the room its in with sound. So do other instruments radiate indirectly mostly. For string instruments it's easy to see. The string itself vibrates in many modes simultaneously like a piece of limp spaghetti but it makes very little of the sound directly. You can experience this at a music store. Pluck or bow the string of an electric violin with the amplifier off. That's the sound the string sends directly to your ears, you can hardly hear it. Most of the sound comes from the vibrating front and back piece and from air coupled through the f holes. So it's also multidirectional. What about horns (brass and reed instruments.) They are highly directional but they are rarely if ever pointed at you. some are pointed up like a tuba, sideways like a French horn, or straight down like a clarinet, oboe, or bassoon. Some are diagonally downard like a trombone or a trumpet. Also in a symphony orchestra the horns sit behind two or three violinists, violas, or cellos so the direct sound doesn't reach you. What about for AR9s? As designed all of the sound above 200 hz is directed at you. There will be some sound that reaches you through reflections. That sound will not have flat FR unless you make it that way. That's what the other 11 tweeters that are aimed at the walls and ceiling do above about 6khz.. But they don't send their sound in precisely the same directions as the other speakers. That means that while the total energy transfer will be flat, it will not have the same direct/reflected ratio at all frequencies and the reflections will not arrive from the same directions and at the same times at all frequencies. This beaming of most of its sound between 200 hz and 6 khz is why it cannot sound "big." This is one problem of every direct radiator box speaker. It's a flaw in the concept that only allows its performance to go so far and no further. While panel speakers are less distorted in this respect they are not designed with the intention of solving it and they do only to a relatively small degree. They also have a lot of other problems. Even knowing all this and how to fix it, it still took two tries and four years. It's not easy until you figure it out. Then looking back someone would probably say, is that all there is to it? Yeah, after about a few hundred other things that didn't work.

 

[fas42]

Frpm http://www.whatsbestforum.com/showt...-the-bandwagon&p=111360&viewfull=1#post111360 ...

there is simply not enough information in the PCM digital formats to work with. and that is where analog has the advantage, much more information. so your theory of better dacs or adc's fails since they don't add data.

Again, wrong, 100% wrong ... digital clobbers analogue, both theoretically, and practically. People say, " ... but when I listen, it just doesn't sound up to scratch!". Yes, because they're listening to an implementation of digital playback that's not performing correctly, and when digital misbehaves it always sounds worse than any sort of reasonable analogue playback ...

Frank

 

[fas42]

From http://www.whatsbestforum.com/showt...Physical-Media&p=111455&viewfull=1#post111455 ...

What I take from this is this, aside from using the computer for most of my own pleasure listening, is that something in playback from disc is changing the audio. Clearly the bits are the same. What's different? The first thing that comes to mind is that playback from disc involves tracking the spiral pattern made by the pits in order to read the disc, operation of a tracking servo to keep the read laser in the correct position and properly focused, redundant reads when necessary, decoding the 8:14 modulation back into binary code (there are no "ones and zeros" on a CD but instead, 9 different length "pits") and performing error correction when necessary. All of these must be performed in "real time" as the disc plays. With playback from a computer, all of these functions have already been performed (when the file has been extracted from a disc) or are not necessary (when the music has been delivered as a computer file).

Yet again, the "evidence" that digital replay is fragile. For all the reasons that I've repeated over and over and over again. Of course, computer replay of "different" CDs will always sound the same, or very, very close to it -- because the file, the audio information is identical every time. But, different CDs, from different pressing plants, vary as physical objects from each other, they are "mechanical" devices in the same way as vinyl pressings are, and because the actual process of reading the CD is a totally analogue process. Yes, analogue: only in the circuitry at a certain point in the reading circuitry is the electronic waveform interpreted as being digital information.

And so in the same way as different LPs of a recording on the same turntable will sound different, so will different CDs "sound different". But why? Surely once it's considered as digital information we're safe ...!! No. Because of the same old gremlins that bedevil audio replay in every area, that is, everything matters!! The process of reading the disc is mechanical, analogue, requires analogue bursts of current to flow in the circuitry which interacts with other circuitry, through the air as RF, back down the mains cable, and finally, disturbs the actual audio replay quality.

Saying that this is too minor, shouldn't happen, that all the best "experts" are certain that this is irrelevant, will not change one iota the fact that this is exactly what does happen! Esoteric theories of subtleties of error correction are only a small part of the "puzzle": the fact that a CD player finds one CD much "harder" to read, for multiple reasons, than another of the same music almost guarantees that the end sound will differ ...

Frank

 

[fas42]

From http://www.whatsbestforum.com/showt...quot-real-quot&p=111277&viewfull=1#post111277 ...

I am in clubs or other venues on average 2-3 times every week to hear live music. There is something about a kick drum that is very difficult to reproduce over a hi-fi system. It is a combination of that instant 'thwack' that moves air, plus a certain tonality of the drum head being hit that is quite distinctive when you hear it live. It is deep, fast and startling. And, it doesn't have to be miked through a PA to create that sound. A big dynamic speaker can give you the depth of the bass, but it sounds a little artifical, almost like you are hearing it in a well, and doesn't sound as quick and turning it up loud doesn't make it startle, it is just louder. I'm not sure about a horn being capable of that kind of depth, only because I'm not sure I've ever heard a big enough horn woofer in a home environment. Granted, large PA systems in big venues can do it, presumably because they are using pretty massive horn loaded woofer systems, and the rooms are large. But, that 'jump' or startle factor is one of the things that I find, at least on kick drum, separates reality from reproduction. Maybe some big dynamic systems with tons of power behind them are capable of it, but I haven't heard it done convincingly-

Worth commenting on this ...

Why a lot of systems don't do the kick drum properly, is because there's a lacking in the treble: get a system that does do it properly, kill its tweeter, and listen to the miserable drop in quality of the impact factor ...

Where a lot of systems go wrong is that the big bass notes suck the power reservoirs dry, the smoothing cap's are struggling to keep the voltage up to scratch, and some intricate, or intense treble detail comes along: the amp makes a vain attempt to reproduce that detail and fails, quite often badly. This is why orchestral climaxes frequently degenerate into a mess at realistic levels on normal systems ...

And this is exactly the sort of behaviour that is never, ever tested with conventional measurement techniques ...

It doesn't need super powerful amps to get it right, although it helps -- if a vehicle has square wheels, then adding enough V12 engines will eventually force it to move -- just properly executed engineering of what one has.

Frank

 

[fas42]

From http://www.whatsbestforum.com/showt...t-upgrade-quot&p=112063&viewfull=1#post112063 ...

Why upgrade if you dont know whats wrong?

Exactly ...

But, usually, I could imagine it's because an album just doesn't come up to the mark, in terms of how you feel it could sound -- so you try something else to perhaps get you closer to the best rendition of that piece of music.

Me, I work on the principle of getting the "worst" recordings to give of their best. Which automatically ensures that all the "better" stuff will in turn sound so much more rewarding ... it's really all very logical ...

And I determined eons ago that changing elements at the macro level never solves everything, just bits and pieces of the puzzle; it merely reorganises, reorders the problem areas stopping you getting good sound under all conditions. And that the only true solution is to resolve all the "unfixed" weaknesses in whatever gear you have on hand at the time ...

Frank

 

[fas42]

From http://www.whatsbestforum.com/showthread.php?6758-Do-we-use-our-ears&p=112314&viewfull=1#post112314 ...

I know one guy who assembles a system, tweaks the hell out of it until it sounds fantastic then sells everything off and starts over. That guy really loves the process.

... way to go !! I envy his energy -- need a few years peeled off my brain cells to make it that easy, but he's certainly got the right idea ... :b:b

Apart from that, this thread, on using your ears, for the 1000'th time identifies and separates the people who have experienced "good" sound, and those who haven't. And those who haven't constantly point the finger at formats, 2 channel vs. 3 channel say, or any other of a decent selection of handy "offenders" to implicate as reasons for why one it is "impossible" to achieve high quality sound: deliberately manipulative sound engineering, as another culprit, is very convenient of course.

Now, if the only sound you've ever experienced has been the relatively tininess of conventional stereo replay this all makes sense. But it is sad that these people then insist that their view of the universe is the only correct one ...

Frank

 

[fas42]

From http://www.whatsbestforum.com/showt...re-information&p=112394&viewfull=1#post112394 ...

Tape is more liquid and flows effortlessly out to the edges. Maybe someday digital will be able to be as 3 dimensional as tape, but I doubt it.

Roger, you're talking about different styles of distortion. As Tom(elex) will point out, R2R adds quite solid gobs of distortion to the sound, but it is of a type that is relatively pleasant to the ears. Digital, on the other hand, if not totally optimised injects really, really unpleasant varieties, they may be literally 100's of times lower in level than that of tape, but they are particularly offputting to listen to; hence why so many people struggle to get long term pleasure listening to CDs, etc.

With digital it is indeed all or nothing: if the sound is on song then it is massive, encompassing, totally believable; if just a tiny bit awry then exactly the same material can sound bloody awful! There is very, very little room between those 2 extremes with a high performance system -- why the journey is many ways is so much harder, going this way ...

Frank

 

[NorthStar]

Frank, Digital ain't 'liquid'?

* By the way, you should visit me (Visitor Messages box); got few surprises. :b

 

[fas42]

Frank, Digital ain't 'liquid'?

* By the way, you should visit me (Visitor Messages box); got few surprises. :b

Bob, "liquid" sound is just short hand for sound lacking distortion of a certain type: when vinyl, digital, or R2R replay drops below containing a certain level of the particular audible artifacts then it all becomes "liquid": this is sound that the ear/brain finds easy to digest, and in many ways is very close to "natural", or "real" sound -- it is sound, assuming the setup has got the effective grunt to do it, where the volume can be wound up to maximum levels without discomfort or problems, allows you to fully groove with the music making. Quite easy for R2R to do; very arduous to achieve, and then fragile for digital in many systems ...

Pretty snazzy links there, I'm impressed!

Frank

 

[fas42]

From http://www.whatsbestforum.com/showt...re-information&p=112451&viewfull=1#post112451 ...

But the question that bears asking is again, with all of its theoretical advantages, why doesn't a high rez digital file sound like the original tape?

Because, because, people still suffer under the delusion that audio components are "magic" black boxes that operate with absolute purity in performance within themselves, irrespective of what other electrical components are plugged in, powered up, connected to the particular device being considered. Digital sound is fragile, it is affected by, it affects all the other electrical apparatus that is part of the audio setup, and those beyond the audio environment. Just plugging in some supposed high performance box and expecting it perform magical audio tricks, without tweaking the environment, is almost certainly guaranteed to yield somewhat disappointing results ...

You only have to hear how atrocious studio monitors sound when surrounded by a sea of electronic mush and mud to be aware of this ...

Frank

 

[NorthStar]

Bob, "liquid" sound is just short hand for sound lacking distortion of a certain type: when vinyl, digital, or R2R replay drops below containing a certain level of the particular audible artifacts then it all becomes "liquid": this is sound that the ear/brain finds easy to digest, and in many ways is very close to "natural", or "real" sound -- it is sound, assuming the setup has got the effective grunt to do it, where the volume can be wound up to maximum levels without discomfort or problems, allows you to fully groove with the music making. Quite easy for R2R to do; very arduous to achieve, and then fragile for digital in many systems ...

Pretty snazzy links there, I'm impressed!

Frank

That is a pretty good description of "liquid sound" Frank.
And Analog stands a much better chance than Digital in that concept (realistic theory), I believe.

...I hear and I see you, from 'Avatar'. :b

 

[fas42]

From http://www.whatsbestforum.com/showt...re-information&p=112572&viewfull=1#post112572 ...

just listen to a digital cymble, then on high level vinyl or tape. overtones and decay is missing on the digital. the metalic quality of the tone is missing on the digital. the fine details are blurred on the digital. the life and energy are missing comparitively.

From someone who has a very ambitious system, and for others not to challenge his assertion, this is very telling. These are the usual symptoms of digital not working right, distortion in simple terms, and I'm not referring to the format itself. Correctly working digital replay has no problems doing this sort of thing, in fact it can do it to a much higher level than analogue if the recording is encoded with the right material.

So the only conclusion I can come to is that most people on this forum have relatively non optimal digital sound, irrespective of how much money they've spent on it -- a "suspicion" :b I've had for a very long time ...

Frank

 

[fas42]

From http://www.whatsbestforum.com/showt...re-information&p=112660&viewfull=1#post112660 ...

Low-level detail seems to be where the"magic" is & yet digital, natively by design, has a weakness at rendering low level detail- quantization error. It requires the addition noise in the form of dither to overcome this error.

Wrong, wrong, wrooong!!! What you hearing is introduced distortion, note this word, introduced, distortion occurring in the process of playback. I've spent 25 years plus working on this "problem", and I know it by heart, even though I haven't got all the technical answers for why it occurs. A dead, lifeless quality in low level detail when playing CDs is something I've always heard when listening to other people's systems, but one doesn't have to live with it, it can be fixed -- if I hadn't been able to "solve" it, I would have chucked the hifi setup out the window years ago, and I in fact did so, effectively, for about 10 or more years, because it got so frustrating ...

But if people keep banging on, prattling, about needing more bits, less jitter, higher sampling rates, better dithering, you'll never, ever solve the problem in a wider context ! Never, ever ... it will just remain another, Never Ending Story ...

Frank