Relative importance of differences in stereo systems

[John Dyson]

6 minutes ago, ralphfcooke said:

Given the significant differences in the sound  between even very high end speakers;

which swamp any differences in the intervening electronics, why aren't we more concerned about

getting this right?, rather than trying to decide 'how many angels can dance on the head of a pin'

wrt different DACs, cables, etc?

Even though electronics nowadays is mostly 'perfect' when compared to speakers -- there are two issues (off the top of my head).  One issue is the use of legacy equipment -- for example, I used to have a Nakamichi preamp and Hitachi amplifier (one of the first FET amplifiers.)  Both distorted fairly noticeably -- and I didn't have golden ears -- just good perception.

The other thing is that sometimes people don't want 'perfect', but would prefer a certain 'sound'.  For example, few vacuum tube amplifiers are as perfect as well designed semiconductor amplifiers -- however some people might (rightfully so) prefer the even order distortion, lower damping or other artifacts of the 'tube/valve sound.'

Summing up:  electronics differences include using older -- less well designed equipment can make a difference, and also boutique equipment with a certain coloring.  There are probably other aspects that might make the HW different -- using 30 gauge wire might make a difference on the speakers themselves, or maybe using high capacitance boutique wires of any kind anywhere can make a rather 'interesting' difference from perfection.

[John Dyson]

 

I promise you that a major historical problem with listening  pre-recorded material has been leaked DolbyA.  Amazingly, all of the 'golden ears' didn't even realize that they were often (not always) listening to 10dB of compression at 3-9k and 15dB at 9k-20k, thinking that they were getting 'high quality' audio (that is, DolbyA encoding.)  (BTW, DolbyA compression is not active in the same way as a normal compressor -- so it doesn't sound as bad as the compression figures imply.)   This leakage is ONE REASON why people had earlier on noticed the 'harsh digital sound'  Even the HDtracks Carpenter's album is (or was when I purchased it about 1-2yrs ago) DolbyA encoded.

There ARE always complications in getting good material and lots of things like speaker positioning/etc.  I quit the audiophile hobby back in the very early 1990s, because the CDs never sounded good, and the vinyl was... well, vinyl.  It is only recently with computers being very fast, and the ability to emulate DolbyA decoding (not so easily) on computers that I found out that not only a few titles were DolbyA encoded, but MANY titles were.  I'd consider any material created between late 1960's though early 1990's to be vulnerable, and I have even heard more recent material than early 1990's as DolbyA encoded.

So -- quality source material is probably the biggest bugaboo, and it has been so since listening to prerecorded music in the home has been possible.  Digital has not solved the problem, but it could -- someday.

[John Dyson]

1 hour ago, Blackmorec said:

Because, generally speaking, what goes into my speakers is what comes out.  The better the quality of the signal going in, the better the quality of the music coming out. If you’re listening to the sound of speakers, you’re probably also listening to the wrong speakers. But if any speaker is set up incorrectly relative to the room, that better get looked at before bothering with anything else. That’s true. 

It seems like you might have good speakers -- not criticising that, however speakers are NEVER accurate.  They can create a sense of accuracy, beauty or sound good -- but they aren't accurate.  I see that you wrote 'generally speaking' and that is a good thing to say.

Poor speakers will hide the differnces in your source materials, good speakers will show some difference.

However, even the tightest coupling to your hearing system will not be totally accurate -- and speakers with all of their resonances (yea, you can tune part of that out, but that is ONLY frequency responses), will only give an approximate representation.  (With things like speakers and almost any physical device, there are REAL mathematical limitations as to how well the timing (phase) and frequency responses can BOTH -- BOTH be corrected.)  That, of course, assumes that the speakers are linear -- and low frequencies really cause problems in those areas.  Even coupling between the LF speakers and the MF/HF speakers cause phase distortion -- usually not alot, but just one example.  Any time you have to move a diaphragm a long distance very quickly -- it will cause odd phase shifts.

Speakers can never be accurate, but they  might sound good -- and even might be more accurate than usual...

 

My guess is that a big part of the hobby for some people is trying to chase the ideal...  Have fun at it -- it gets more and more tricky as your stuff gets better and better.

 

John

[John Dyson]

1 hour ago, Blackmorec said:

Thanks for the discussion but you know what? I don’t honestly care if my speakers are accurate or not. Since I don’t know what the original sounded like I don’t care if the facsimile is 100% accurate. I care that it sounds great, natural, musical, rhythmical and dynamic and blows me away with music. As long as they’re 100% involving and 100% convincing, it matters not a jot whether they’re accurate or not. 

Add the fact that I’m playing them in a fairly small,  reflective, lossy but reverberantly well balanced room and the best I can hope for  is a highly musical, natural and transparent rendition of the signal they’re fed, with no identifiable coloration. If every track I listen to has its own acoustic, tonal and dynamic character rather than something the speaker or room imposes on the music, I’m good. 

Its worth bearing in mind that what we’re listening to is a recording of the original event, if there even was an original event

I agree that accuracy isn't important, but it is also impossible (when compared with the rest of the system).  The best thing is to try to enjoy and play.  "Sounds good' or "Sounds better" is all that is important.

[John Dyson]

3 minutes ago, The_K-Man said:

 

Actually, album mastering is more important than DSP, unless that DSP is concerning time alignment issues among speakers in larger setups.

I am working a big part of the mastering problem (I mean, for older material) just this second.  My project is being developed with a recording engineer (and help from several others -- including grammy winner involved also.)  A lot of old material is poorly mastered, or even just played out from a tape -- they call that 'mastering'. ('They' isn't the recording engineers, the 'They' are the bean-counters.)

Previously, dealing with the old DolbyA recordings (long since on digital) has been expensive (having to run the material real time through ancient hardware.)  Even if it IS on analog and quickly available (no trips to Iron Mountain are necessary), then dealing with the HW is a pain.   My project is to eliminate that step, and do it better -- less IMD and other kinds of distortion.  (NR systems usually splat IMD all over the place -- only the slow attack systems avoid that, and then they have other troubles.)  My DolbyA decoder does all kinds of tricks with distortion cancellation and moving distortion sidebands into hiding places (kind of.)  Also, it shapes the attack/decay curves ever so slightly to minimize the amplitude of distortion sidebands (IMD.)

MASTERING IS/HAS BEEN A REAL PROBLEM.  The sad thing is that it isn't the frustrated engineers fault -- it is all that 'bean counting' going on.

John

 

[John Dyson]

19 minutes ago, The_K-Man said:

 

Just be sure to notice my profile avatar, and remember that just cranking it up and heavily limiting it is not mastering, or remastering for that matter.  Preserve those peaks! Use peak limiting only for the few outlier peaks.

 

Resist demands by the client(the artist, the producer, or label) for LOUDNESSABOVEALL.  If you lose those customers because they went somewhere else to have their mixes intentionally distorted or otherwise over-processed for final release, they weren't worth your time and talent as a post-engineer in the first place.  If you value great sound, you will attract like-minded clientele.

The guy that I work most closely with really dislikes any extra processing.  He is into producing real quality stuff.  I don't do mastering -- that is a specialty area.  I do software/analog hardware/DSP/etc type stuff.

My DolbyA compatible project is meant to retrieve as much of the signal as possible -- without causing artifacts.  There are modes that I can set up in the decoder, and it can pull out more detail -- but sounds 'not quite right.'

[John Dyson]

2 hours ago, fas42 said:

High end speakers don't "swamp any differences in the intervening electronics" - what they do is highlight, exaggerate with a ruthless attitude everything that is wrong in the preceding chain - if I want to know how awful audio playback can sound, how unappetising it is to be in the presence of sound reproduction, I just have to visit any hifi show ... .

 

Typical audiophile systems make one intensely aware of the tiniest anomalies - they are a hair shirt experience of the worst kind ... a good rule of thumb is that the greater the potential of the rig, the more one has to thoroughly 'debug' it - otherwise, they make listening to anything but the "right recordings" quite unpleasant.

 

The problem with having a system with 'personality' is that nuance IN GENERAL can be lost -- but certain kinds of nuance will be made intolerable.

Speaking of frequency response -- when I am working on my projects -- even 0.5dB error can be incredibly irritating.  If you know how something is 'supposed' to sound, it is pretty easy to detect the smallest errors.  If someone has gotten used to an eccentric kind of sound characteristic, than equipment that is neutral can sound 'inferior', weird or maybe even bad.  Vice versa is also true.

It is so very easy to actually believe that something with 'personality' (I am being kind) is better -- because that is what one might be used to.

This is why it is important to just try to enjoy the music, or enjoy tweaking ones system or both.  Trying to make a system superior to some other system -- well -- beauty is in the eye of the beholder.

 

John

 

 

[John Dyson]

1 minute ago, Kal Rubinson said:

Not in my experience as a proselytizer for multichannel.  The biggest objections are that there isn't enough room for more amps/speakers and that spreading the available budget over 5.1 or more channels would mean that the cost per channel would decrease as would quality.  They offer this as an argument against MCH without realizing that its advantage over stereo trump any such concerns.

I believe that some people who are hooked on the audiophile hobby will keep on tweaking & enhancing until they are finished.  People can be 'finished' by lack of funds, lack of imagination or finally learning to just enjoy the music.  Nothing wrong with the tweaking&enhancing phase, just as being satisfied by the best practical system is also okay.  Some people dont' stop at the practical and technically near perfect, and overshoot into hyperspace.  How far they end up in hyperspace depends on funds, persistence, getting married and/or maintaining interest.  Loss of acute, 20yr old hearing is not an impediment to the hobby -- practical constraints are the major outside influence to the hobby..   The persistent audiophile is not limited by his/her hearing ability -- up to a point. :-)

[John Dyson]

10 minutes ago, Teresa said:

 

I thought only very rich people had separate rooms for video and audio systems. For the rest of us a single system is used for both audio and video. And many of the few I've been in with multichannel systems have the rear speakers setting on top of the front speakers because they don't want wires running along the walls. If one rents an apartment they cannot put wires in the walls. I suppose one could get wireless speakers for the rear. Oh, well. 

Even though I used to have a rather significantly high income (into early 7figures), I wouldn't/couldn't dedicate a room to multimedia.  (nowadays, I fell into a rather frustrating illness -- not quite as 'fast track' anymore)  Frankly, as I have more user interfaces, I become more and more confused.  In fact, I almost never log in to my windows boxes directly -- it is easier to pull up a window on Linux (hate to use that name -- used to be a horrid competitor of mine.)

If anyone is into video/audio -- I am, and have been since the 1980's, but I did have an outlet at work -- maybe that is a partial excuse.

My own situation (even as a forever single guy), it just doesnt' make sense to compllicate my life with more 'things'.  I like to have 'fewer' but more useful things.  "things' tend to be a pain to deal with for me :-).  Maintenance is a frustrating distraction, etc....

 

John

[John Dyson]

These discussions are bit confusing to me -- because I have been around the best quality -- and it often doesn't have polished wood cabinets and/or nice foam carrying cases.   I like have worked with super high quality audio -- yet so many people who are audiophiles seem to go for an ambience that isn't important to me.   For example (wrt my project) -- lots of audiophiles have likely listened to material that has been mis-mastered, yet feel like maybe a lower jitter (or some other mostly specious attribute) has helped the sound.  (Jitter CAN distort audio, but one doesn't need the kind of jitter specs that are often looked for (or, the actual 'jitter' is usually more of an analog noise situation.) Here is what I mean:  people who are desperate to improve their equipment might be looking in the wrong direction for good solutions.  (This is where the snake-oil salespeople misdirect/mislead for their own financial benefit.)  A proper and complete analog engineering degree (which can be gotten in 4 yrs), is a very good first step (along with additional study/materials) which will prepare the 'audiophile' to solve their own problems.  How many people have a hobby that has lasted 20yrs, and they are in the same rut of polished speakers, questionable home-brew and snake-oil designs, and using the same defective recordings (loving the nice, bright highend or some other defect?)  A small part of the 20yrs can be used to actually learn what is going on -- then being immune to the snake-oil, but maybe even create something (gratis) for others to use...   So, BEING MORE PRACTICAL -- instead of the full education -- a bona-fide attempt to bypass the snake-oil would be helpful -- don't just look in the audiophile popculture information -- and watch out for snake-oil people trying to look like they are contributing to legit imprrovements.

(Some new, esoteric ideas ARE good ones, e.g.  the very interesting  time base correction scheme for audio tape -- that IS a good thing.)

 

Good audio equipment isn't 'elite', but boutique equipment certainly is 'elite.'

 

Point being -- if someone likes the sound of their polished mahogany speakers  (or amplifying devices which double as space heaters and night lights) -- that is cool, but such ambience isn't needed for good sound.  It MIGHT be needed to make a given audiophile happy -- and that is okay also, that is -- unless someone who 'really doesn't know'  might  judge against my simple, effective and very good soundign equipment.   My equipment is like that used in an EE lab, being used to TRULY benefit the quality of recordings -- and it is just as good as any other 'semi-industrial' equipment.  (Geesh -- DolbySR, which was used along with DolbyA to encode/decode old recordings used pathetic OpAmps (many of them in the path), that would make the 'elite' audiophile cringe.)  Some of their 'music' has been tainted by that equipment which 'cannot sound good.' :-).

 

Status isn't important to me -- been there, done that.  I found that simplicity is my own key to happiness.

 

John

 

 

[John Dyson]

28 minutes ago, fas42 said:

There is a circle of confusion - but it's not centred on the speaker ... .

 

Rule of thumb: the more the subjective signature of a rig varies as different loudspeakers are tried, the more the electronics of the chain need to be looked at more closely - the end result of 'correct' optimising is that you hear only the recording, and not the "characteristics of the speaker" - if it is truly a high end, transparent chain, how can it not be that way ... ?

 

You are implying the futility of total technical perfection -- it is best to have the equipment that makes the owner perfectly happy instead :-).

 

The best thing that an ideal speaker can normally do is to be clean enough to detect differences more clearly (and not over-enhance the differences as a freq response peak or distortion might do.)  Even if you DO compensate the freq response of a speaker system there are mathmatical limitations as to how accurate you can make the timing & freq response fixes.  Just because of all of the variables it is difficult to make an audibly perfect accuate transducer -- and speaker systems have even more variables and larger scale problems (e.g. mass, varying environment.)  Equalizers don't fix things as well as freq response might imply -- EQs might help, but even a perfect equalizer cannot fix all of the problems.)  DSP equalization can help -- generally having more degrees of freedom than an analog EQ, but there are practical (and some mathematical) limits.

 

(Here is a simple example of all of the complexity that real engineering has vs even a somewhat knowedgeable tech might not know -- did you know that the components of impedance -- resistance and reactance are not normally mathematically independent?  For normal circuitry, there is a simple (but sometimes difficult to calculate) relationship -- most people don't know it. There are so many subtle facts in engineering & physics that makes the world harder to deal with than it might seem.) 

 

So, simplify the environment, minimize the depths of the resonances, decrease the mass, etc.  Even the most expensive OR the very best peaker systems are all going to sound different.

The best that can normally be done -- find a clean sounding speaker system that the user/owner likes best.  Headphones are easier to make accurate, but only decrease the scale of the problem.  Even with headphones, there are geometry issues which make perfection a subjective notion.  (A flat response transducer will not necessary produce an accurate representation in human hearing.)

I am not claiming that 'all is lost', but the only perfection in a practical large mass transducer is going to be that the owner is perfectly happy.

The listener is best to find equipment that makes themselves happy and listen to it -- or enjoy the hobby of continual attempts of getting closer to the desired sense of perfection.  (JUST MY OPINION.)

 

[John Dyson]

27 minutes ago, fas42 said:

 

 

 

Ummm, they will sound the same - the best system I came across recently, that was not my own, used top notch Dynaudio speakers, the best Bryston amplifiers - a long way from what I use ... and this system nailed the recordings. The sense of the event, the tonality was a big tick - 'natural', effortless; it was all about the music, nothing about the equipment.

You are making my point -- you are happy with what you heard -- I guarantee you that someone else will not think so.  I spend hours and hours working on audio stuff -- using my finest ability (which isn't good enough) to discern the smallest imperfections...  I doubt that there is ANYTHING which will be 'perfect' to me -- unless it is wired directly into me, or I am there at the performance.

My project sucks -- because the training does odd things to how audio is perceived.

All I know is that the math, physics/etc along with any normal speaker design isn't going to 'nail' the general reproduction of music.

It can sound 'good', but 'nailing it' is an impossibility.  (Headphones come closer, but still don't nail it.)

 

I don't need to use experience -- just knowing what goes into the design of the equipment.

 

John

[John Dyson]

1 hour ago, fas42 said:

 

You are not right, at least for part of the population - there is a standard of SQ which triggers the brain into experiencing a convincing auditory illusion; if you haven't personally come across this, it may be because A) your brain is not wired for allowing it, or B) the standard of playback has never been good enough.

I have done/designed SQ/QS decoders,  and hear the effect.  But it isn't audible perfection either.  (in fact, SQ/QS aren't really that good performers of the quad illusion.)  I know all about the math/etc involved -- in fact my DHDA DolbyA compaitlbe decoder uses similar math (yes ,processing in hybrid/90deg terms -- similar to some of the QS/SQ math.) (If you know much about the math for gain control -- only primitive gain control or slow gain slew schems do the simple 'gain*signal' type operation.  Simple changing gain causes IMD splats all over the place -- probably one reason why digital gain control things somtimes don't sound very good.)

 

All I can tell you -- is that human hearing is subject to brain-processing, and it is amazing how much a person can be fooled into something sounding similar, but it really isn't.  A good example is the perception of spoken language sounds for non-native speakers.

 

John

[John Dyson]

11 hours ago, fas42 said:

 

You are not right, at least for part of the population - there is a standard of SQ which triggers the brain into experiencing a convincing auditory illusion; if you haven't personally come across this, it may be because A) your brain is not wired for allowing it, or B) the standard of playback has never been good enough.

I still don't think that you understand -- you are missing details if you think that SQ works very well.  It works adequately well for some people who haven't learned to discern.  SQ is a fairly cheap replacement for accuracy...  That is okay...  There huge messes in the differnece between SQ/QS and friends vs. the actual signal & human hearing.

If someone's hearing cannot discern -- then the limitation is in their hearing.

 

John

[John Dyson]

On 1/19/2019 at 9:32 AM, STC said:

 

It depends on which parameter's accuracy you are referring to. 

I am speaking both of the listening experience being an inaccurate representation of true quad, and also electronically it is very very far off from the painstakingly created multichannel material.

If the artist/engineer PLANS for the limitations of SQ, it might be convincing, but it still wont be accurate WRT the quad recorded for mastering.

Don't get me wrong -- SQ can sound nice -- I am not talking about 'sounding nice.'   The big problem with 'sounding nice' is that it is very subjective...  For example, there are some people who really like compressed sound -- and considering the artists intentions when using compression, then that is their choice.  However, certainly much much of the time, a compressor (esp limiter) will create very large amounts of IMD when compared with the best of linear/non-processing equipment.  Much IMD is not due to the quality of the gain control elements or the quality of 'capacitors' -- it is minimally due to the math/physics of gain control (esp fast/hard gain control.)

Point being -- 'beauty is in the eye/ear of the beholder.'   Much of the time, I can hear the IMD from compression  (DBX is a little better, but a lot of FET compressors splat all over the place, it is just that people have become used to it.  Optos help the problem in a similar fashion as DBX), so even though compression usually distorts the sound, and distracts me -- alot of people do like the  fullness (sometimes intensity) of the sound.

Point being -- if you say that something sounds accurate -- well, to you it might -- just as someone whose native language is different than mine, I might not notice sounds that he/she does and vice versa.

Absolutes in the subjectiive/perception area are seldom possible to accurately claim, however absurd statements can be filtered out as 'misguided opinion.'   Note that I seldom advocate, by my position is to observe and apply some science and physics to the situation.)  (Science & physics include perception issues.)

I am all for an audiophile hobby -- but I disagree with making strong subjective claims as fact unless it is clearly implied as 'IMO.'  Even IMNSHO is okay.  A lot of people are NH esp in their field of interest/love of hobby.

[John Dyson]

23 minutes ago, STC said:

@John Dyson sorry for the misunderstanding. I was referring to sound quality not stereo quadraphonics. Although, IMO Quad can sound better than stereo when implemented right. 

Well -- sometimes (many times) some kind of multi-channel thing can sound impressive.  Sometimes (one in a million) might come close to spatially accurate (at one location in the listening venue as recorded in the studio or hall.)  IMO, this is another case of 'sounding good.'

Maybe a scheme where a head with the same geometry as the listener, microphones in each ear, then headphones on the playback -- that might come close to realism.  However, there are still issues with the frequency response variations of the microphone (pretty well controlled) and the coupling of the headphones to the listener (usually not so accurate, but sounds good.)

Then, then 'head' can be moved around to emulate the location in the studio/hall. :-).  Creating a listening environment where moving around is identical to the recording studio/concert hall isn't likely (but might come close for some people and/or limited movement.)

 

Still, that is 'sounds good'.  My interest tends towards the electronics part of the whole thing, and except for NR damage & tape noise that the NR tries to remedy (of course, on old material), that stuff is close to perfect.  Maybe in the early 1980s, there were still problems, amplfiers  but nowadays with power transistors with fT in the tens of MHz, almost all amplification issues have now been resolved.

The ubiquitous U47/U87 and other colorful microphones, and using a DolbyA for vocal enhancemnt -- those were bad also, but an artistic choice.  Of course, back in the '70s, I was using Condenser/FET microphones totally flat in the audio range, so perfection in some areas back then was nearly possible -- but not always used or desirable.  Using 'colorful' microphones doesn't bode well for realism either.

 

Point being -- perfection (or nearly so) has been available when it comes to everything but transducers and listening environment-- it is just a matter of practicality and taste.

[John Dyson]

16 minutes ago, Axial said:

NR damage, as in tape decks with A, B, C, S Dolby. ...Filters for the music, adding a veil to disguise the impurities, adding coloring to mask the rain falling down the sky (wow, flutter and jitter). 

 

I think it was in that meaning direction ...

I just uploaded some decoding examples using the DHNRDS decoder.  Please refer to the 'Lies about vinyl vs. digital' forum about what to expect/etc.  The DolbyA encoded material might not sound like what people think -- so read the section that I wrote if you are interested -- it will set-up the thinking process.  PS: this is meant to be informational only -- I am only talking about the decoder software -- and I am certainly willing to talk about most of the internal technology if anyone is interested.  A few sections of the technology are proprietary, and likely never been done before (the gain control is not just gain*signal, but takes a big part of a 4core CPU at realtime, being multithreaded (12 threads, 10 are active DSP calculations) with super complex math.)  (Technical paper will likely be written.)

 

Repo:

https://spaces.hightail.com/space/yDG3L339Rn

 

[John Dyson]

34 minutes ago, esldude said:

I've listened to the Eric Clapton "Wonderful Tonight".

 

The undecoded version has some obvious, pumping, well maybe not pumping, but dynamic inconsistencies.  The decoded version sounds very much like the Mobile Fidelity CD of this I have.  So I suppose they were using a master tape or something that was done properly.  

 

Thanks for these examples.  It is very interesting. 

There are some more ABBA examples at the site listed below.  Unforutnately, some of the DolbyA copies were EQed, and I forgot to do so.  There is only half of what I planned right now, but enough to see that DolbyA doesn't sound HORRID, but is a serious problem with the reproduced sound quality.  (The worst thing as I noted before -- it flattens the spatial image, and just sounds lifeless with HF dynamics problems.)  Right now, there are a few examples -- more to come at this ADDITIONAL site:

 

repo: https://spaces.hightail.com/space/zqrnA5nqnV