[Felipe]

Hi people,

A lot has been talked and done around this so called "super upgrade" that some high end audio companys seem now to be fans. The growing research for drivers units that go on and on through the frequency range up to 100Khz seems to be overtaking the audio comunity being in tweeter or "super tweeters".
Also amplifiers designers claim to have linear response to 30Khz or 40Khz. I must ask...is this really worth it ? Has anyone really heard a clear improvement on overall sound when adding a super tweeter that only works in the ranges from 30-100Khz ??
I assume this can have a subjective intrepertation or analysis as one could compare it to the other end of the spectrum where we have  been having sub-woofers claiming to reach 5Hz or 10Hz. Yes, still not hearing those frequencys i believe we can "feel" them as low frequencies tend to vibrate ones body ! What do you reckon ??

[Tinker]

A problem with questions like this is the immense difficulty involved in testing them empirically.  As readers of this forum know, there are a LOT of variables to be considered in the recording an playback chain (not to mention the original source and the acoustic environment).

At the risk of starting a frank exchange of views, the answer is, it might.  Let's conjecture based on some relevant empirical data:

Physiologically, the ear/brain does not resolve frequency components above about 5kHz as a definite pitches and there does not seem to be any hair cells that respond to frequencies above 20kHz, so the physiologists will argue that you can't hear that high.  Is there a "feeling" analogous to felt extreme sub-bass? Who knows? None has been documented that I know of.

What I do know, is that several years ago I was involved in evaluating 96kHz audio playback devices and a range of tests found that 96k did sound better.  In the end we concluded this was mostly due to the fact that the wider bandwith of the high sampling rate allowed better filters to be designed: namely anti-imaging filters being able to place digital artifacts at frequencies outside our hearing range, and reduction in in-band phase shifitng, shorter ringing and/or better attenuation etc.  When a 44.1kHz filter was implemented in a 96kHz playback chain the subjective choice between the 44 and 96 k versions was at chance.

In short, extended bandwith seemed to confer an advantage by moving artifacts out of our hearing range.  Rupert Neve has been building recording desks with 50k+ bandwith for years, and many claim this is part of the magic.

The same may be true for tweeters: wider bandwidth could mean better performance in the range we CAN hear. 

It is unlikely that anything happening at those high frequncies is heard directly, and any conventional transducer producing frequencies that high would be very directional.

T.

[Tinker]

I'll add as an adendum that my various speakers use
Scan Speak D2905 tweeters, which are flat on-axis to about 30k
Morel MDT40's which are flat to about 15K then begin falling off
Heil AMTs which go out to about 22k

I don't plan to upgrade any of these in the near future.

[Tliner]

Hi All,

A few years ago I was involved in constructing composite panells for use in high speed aircraft. These panells were destruction tested. All panells had a "natural" resonance and when this resonance was amplified the panel ultimately deformed and in some cases shattered like a sheet of glass. IE the singer and the wine glass. Without going in to technical details it was not the panells "natural" resonance frequency that did the damage but rather the build up of harmonics that combined to form another frequency, that was the trouble.

From that it may be extrapolated that it is the unheard harmonics (if a transducer is capable of producing sound waves, say over 30Khz) well outside the human hearing range that combine, adding "colour/warmth etc" within the hearing range. I don't know really if this is the case with music etc. But harmonics can play strange games in composite panells. It can be said that when a rock is thrown into the proverbial mill pond waves are radiated from the impact point evenly spaced. But this is not the case as there are many ripples at differing magnitude and spacing - harmonics associated with the original impact ?.

This is now encroaching on the subjective area of discussion when the extemded bandwith of a transducer influneces the music we hear. Makes the music more real ?

Food for thought and discussion!.

Tliner.

[Tinker]

It seems the combined points here offer something of a conundrum:

1. Out of band material, faithfully reproduced, may contribute to realism
2. Out of band material, unintended, may be destructive

Point (1) argues for wide bandwith throughout the signal chain, point (2) argues for limiting it.

In either case, CD as a source is already deliberately band limited, vinyl is bandlimited by analogue processses, so it is worth noting that there is nothing above 20k or so that the recording engineer INTENDED to be on the track. 24/96k is a slightly differnt matter, but only if the engineer used a wide-bandwidth master (e.g., Chesky, Telarc, and other audiophile labels).

T.

[Tliner]

Hi Tinker,

A you point out there can be two valid ingrediants in the conundrum. However, in the case of audio reproduction there is little likleyhood of physical destruction unless there is gross operator error. In the process of recording I have recently experienced very good local (Australian) recordings. I asked the recording guy I know to master a totally unclipped recording of a band. To record as low and as high as his recording equipment allows, then not to alter anything in mastering. My hotted up old CD palay goes from 2hz to supposedly over 35hz at the top end. The amp and speakers have a fairly flat response from 50hz ( being 6db down at 30hz) to over 30khz.

The listening tests (subjective) by members of the band and others who have been exposed to many live performences all say that the listening session was as close to live, as if the band was in the room. So I suppose that the extra recorded information does add to the reality of the music.

It is interesting to note that the recording engineer has received a few complaints that the CD is distorted. The complainants all have cheap mini fi type systems which apparently can't deliver the information faithfully as possible.


TL.


 

[Tinker]

Hi TLiner,
        By "destructive" I meant out of band sounds causing audible artifacts (such as IMD), not physically damaging equipment. i.e., "destructive" top your listening experience.  Although, on the physical side, I have seen the result of very high frequency square waves from test equipment accidentally fed into speakers.  The predicatable result is very fast destruction of tweeters - I guess most manufacturers aren't in the habit of putting in inductors to save tweeters from records designed for dogs.

You have a CD player that plays out to 35kHz? Is it an SACD player?  I never owned one of these, but these supposedly have usuable response out to 50k.  I heared two at AKSAfest and thought they sounded great in every respect but the bass - which I put down to the speaker and room combination, not the player.

I have one of those cheap players that won't play some stuff without weird artifacts.  I've never bothered to open it up and figure out why. I stopped using when I bought my first Rotel.  I guess despite the realisable range of CD many pop recordings must do quite a bit of mangling to allow consistent playback on a wide variety of playback devices.  It's not like it HAS to be done, not like the "elliptical" eq on the bass of vinyl.  I did work in a recording studio for some years, and know some mastering engineers who supposedly have their own "secret sounds good on FM recipe."

I know what you mean about realism - I've heard 24/96 masters from blumlein pairs that almost sound like you're in the studio...  I have had relatively little listening experiences with DSD masters which should raise the effective bandwidth even further. Something to look forward to.


Cheers,
T.

[BobRex]

It seems the combined points here offer something of a conundrum:

1. Out of band material, faithfully reproduced, may contribute to realism
2. Out of band material, unintended, may be destructive

Point (1) argues for wide bandwith throughout the signal chain, point (2) argues for limiting it.

In either case, CD as a source is already deliberately band limited, vinyl is bandlimited by analogue processses, so it is worth noting that there is nothing above 20k or so that the recording engineer INTENDED to be on the track. 24/96k is a slightly differnt matter, but only if the engineer used a wide-bandwidth master (e.g., Chesky, Telarc, and other audiophile labels).

T.

Analog is not as bandlimited as you might think.  There have been tests, published in Stereophile and I beleive presented to the AES, that showed signal on lps well beyond 20kHz.  And this wasn't the old quad subcarrier, it was consistent with the music.  True, the signal was rolled off, but it was present. 

[mgalusha]

A few years ago I was present when a friend was demoing some add on super tweeters that were supposed to play only above 20K. His is an all analog system and so does not have the brick wall filter of red book CD. I was surprised to notice a difference and I know for a fact I can't hear that high but something was there. Mostly noticeable when the devices were removed. He has since moved to plasma tweeters which go out to 40kHz or perhaps more. I can't postulate on what is happening but I do know that he system is awfully special. 

[Tliner]

Hi Tinker,

My CD player must be one of Sony's origionals, a very robust CDP-70 of the 1984 vintage. I bought it new but it only lasted a few days, it refused to co-operate. The dealer had it for months as spares were not available substitute components were used to get it going. Finally it was re-clocked (re-chipped) with a Burr-Brown unit of now unknown spec that is supposed to go over 35Khz. It is not SACD by any stretch of the imagination but appears to extract the last ounce of information from a CD. The sad story is that it is becoming mechanically close to its use-by date.

To augment the bass (if I feel that bass is deficient, CD room or whatever) I run a dedicated transmissionline woofer with its own variable volume amp. It is designed to play music faithfully down to the lowest notes that instruments produce. You can feel the bottom end music and feel it as well.

It is interesting that all agree that if music reproduced above say 20Khz does offer some measure of "realism" to the overall experience.

Tomorrow night a member of the Melbourne Audio Club will be presenting music played via Blu-Ray at the monthly meeting. Technically the bar should be raised even higher than SACD offers. There will be a cheaper V more expensive Blu-Ray machines as well as a CD player demonstration. It will be a very interesting meeting.  The way of the future?

Cheers,

TL.   

[Tinker]

Analog is not as bandlimited as you might think.  There have been tests, published in Stereophile and I beleive presented to the AES, that showed signal on lps well beyond 20kHz.  And this wasn't the old quad subcarrier, it was consistent with the music.  True, the signal was rolled off, but it was present.

You are absolutely right - many old masters have material out beyond 20k. In fact 25k is often published as the practical upper limit (the -3dB point) which means there is quite a lot more out there.  However, not *much* comes off the average (note the word average) turntable. It can depend on how the turntable designer has implemented the RIAA eq, but mostly it is limited by the fact that the mass of the stylus and the radius of the inner portion of the disc radically limits the maximum amplitude possible for high frequencies.   Recent (post 80s?) advances in balanced tonearm design, ultirlight stylus and cartirdge units, isolation etc make LP much more responsive.  The old RIAA spec had something about acceptable HF loss of -3dB at 12k after a dozen or so playbacks due to wear alone (someone will have to check that figure for me). A modern audiophile would abhor that kind of loss.

T.

[andyr]

Hi people,

A lot has been talked and done around this so called "super upgrade" that some high end audio companys seem now to be fans. The growing research for drivers units that go on and on through the frequency range up to 100Khz seems to be overtaking the audio comunity being in tweeter or "super tweeters".

The long ribbon used in the bigger Maggies (IIIa/3.X/20.X) is supposed to go up to 40Khz.  I certainly have never felt my IIIas were lacking in presentation of, say, cymbals but there is a dealer in SoCal who specialises in adding "supertweeters" to these top-end Maggies!    So I guess there must be something in it! 

Also amplifiers designers claim to have linear response to 30Khz or 40Khz. I must ask...is this really worth it ?

Hugh can confirm but I believe the original AKSA power amps had a -3dB frequency of 35Khz.  We all thought that was pretty nice then ... but then Hugh brought out the Lifeforce.  This, I think, has a -3dB point of 80Khz ... and this extra bandwidth gives an amp which is much more revealing.    Then he bought out the Soraya - which I think goes up to 150Khz!     And this sounds even better! 

So ... bandwidth is good, IMO! 

Regards,

Andy

[BobRex]

Analog is not as bandlimited as you might think.  There have been tests, published in Stereophile and I beleive presented to the AES, that showed signal on lps well beyond 20kHz.  And this wasn't the old quad subcarrier, it was consistent with the music.  True, the signal was rolled off, but it was present.

You are absolutely right - many old masters have material out beyond 20k. In fact 25k is often published as the practical upper limit (the -3dB point) which means there is quite a lot more out there.  However, not *much* comes off the average (note the word average) turntable. It can depend on how the turntable designer has implemented the RIAA eq, but mostly it is limited by the fact that the mass of the stylus and the radius of the inner portion of the disc radically limits the maximum amplitude possible for high frequencies.   Recent (post 80s?) advances in balanced tonearm design, ultirlight stylus and cartirdge units, isolation etc make LP much more responsive.  The old RIAA spec had something about acceptable HF loss of -3dB at 12k after a dozen or so playbacks due to wear alone (someone will have to check that figure for me). A modern audiophile would abhor that kind of loss.
T.

First, turntable designers DON'T implement RIAA curves, neither do cartridge designers.  That is the province if electronics designers, specifically those who design phono stages.  Your "average turntable - whatever the hell that term means" if provided with a decent cartridge can extract the information.  Keep in mind that many of us own such tables.  Also keep in mind that your original statement regarding analog bandwidth limitations wasn't qualified concerning "average turntables", it was stated as a fact that "vinyl is bandlimited by analog processes" - a statement that is blatantly wrong.  The idea that this information was only "available" since the 1980's is also erroneous - it's in the vinyl, how it can be extracted isn't the point.  Earlier cartridges, specifically moving coils and moving irons, could easily extract the information, there were also well isolated decks with good tonearms available back then - you probably just aren't aware of them.

Second, you are correct that any music recorded digitally using redbook standards won't have anything above 20kHz, that's a function of the brick wall filter on the recording deck.  But, any garden variety lp recorded pre-digital has the capacity to have extended highs.

Third, take a look at this article: http://stereophile.com/features/282/  This is an article titled "What's Going On Up There" in which John Atkinson runs spectral tests on live music, recorded music, and then garden variety lps (he even measures at the inner diameter of the record).  Please pay attention to pages 2 and 3 of the article where Atkinson measures signal content out at 40kHz - well beyond the 25kHz you believe to be the limit.

Oh, and the RIAA curve is an electrical spec, it does not take groove wear into consideration. 

[Tinker]

First, turntable designers DON'T implement RIAA curves, neither do cartridge designers.  That is the province if electronics designers, specifically those who design phono stages.  Your "average turntable - whatever the hell that term means" if provided with a decent cartridge can extract the information.  Keep in mind that many of us own such tables.  Also keep in mind that your original statement regarding analog bandwidth limitations wasn't qualified concerning "average turntables", it was stated as a fact that "vinyl is bandlimited by analog processes" - a statement that is blatantly wrong.  The idea that this information was only "available" since the 1980's is also erroneous - it's in the vinyl, how it can be extracted isn't the point.  Earlier cartridges, specifically moving coils and moving irons, could easily extract the information, there were also well isolated decks with good tonearms available back then - you probably just aren't aware of them.

Second, you are correct that any music recorded digitally using redbook standards won't have anything above 20kHz, that's a function of the brick wall filter on the recording deck.  But, any garden variety lp recorded pre-digital has the capacity to have extended highs.

Third, take a look at this article: http://stereophile.com/features/282/  This is an article titled "What's Going On Up There" in which John Atkinson runs spectral tests on live music, recorded music, and then garden variety lps (he even measures at the inner diameter of the record).  Please pay attention to pages 2 and 3 of the article where Atkinson measures signal content out at 40kHz - well beyond the 25kHz you believe to be the limit.

Oh, and the RIAA curve is an electrical spec, it does not take groove wear into consideration.

Thanks for your reply BobRex. Interesting stuff!

Sorry for my misattribution about whom does what. RIAA is only an EQ spec which happens at the cutting and preamp stage.  I think my poor prose has distracted us from the bigger issue about what comes off MOST turntables and what CAN comes off ideal turntables.  I do know many AKSAphiles have top turntable and they are certainly capable of great feats.

The few (mostly unreferenced) historical reviews of LP I have read don't cite specific technologies and years, but seem to speak in terms of trends.  My reading of these is that it is mostly through technologies that how to do really good tonearm balancing, resonance damping, top shelf tracking, optimised stylus shape, etc did not have analytic solutions and the best units were empirically built. Sine the widespread use techniques like finite element modelling (I very briefly worked with a group that did this for tonearms) have become widespread enough to be accessesd by mere mortals, playback closer to the theoretically possible maximum has become far more acecssable and cheaper.  I am willing to be disabused of these beliefs.

You know a lot about vinyl so you might be able to tell me which standard (again unreferenced) does speak to acceptable groove wear. Anyone with data or a reference on this please chime in!

I am not contesting that information is there in masters which come from high speed tape or direct disk, I know this stuff is there form my days as a recording engineer. Stuff at 30k+ was clearly visible when calibrating 2" tape units.  I was merely stating that apart from very high end playback equioment, my guesses based on some back of the envelope calculations suggest that catridges which I could buy at reasonable cost have severely attenuated HF simply on the basis published coil compliance and inductance data.

Your comment about Redbook brings us back to Tliner's comment - he has a modded CD player which "goes out the 35K" from which I gather some kind of oversampling and possibly a rebuild of the analogue filter has been done.  This places the nastiness of of antiimaging filters out where they are less obtrusive - consistent with the 44 vs 96 shoutout from a few years back.

Interesting article you pointed us to.  Many thanks.  I wish the pics were big enough to see clearly.

Anyhow, we now have two issues: the issue about what you can get off an LP.  And the original question of whether extended playback bandwidth improves the audio experience.  If we confine the latter question to CD then this may clarify the discussion somewhat. 

Cheers,
 T.

[Tliner]

Hi All,

BobRex, you have stirred up the gray matter! The guy who has recordded the CD's recently uses old "vintage" recording gear which puts down everything?? that the mikes are capable of recording. He can clip the sound later if desired and usually does for commercial non specalist CD's.

T, From my recollections from about 25 years ago I believe what you have described is about what was done to my CD player. I think that the oversampling and the properties of the analogue filter were altered substantially. A few years ago I got hold of a CD player in near new condition of the same make and model. It was not as "musical" original player. Unfortunately I sold it, my spare parts supply gone!

However, I heard the shootout between CD, SACD and two BLU-Ray players last evening. The guy who demonstrated the various players had CD's, Blu-Ray recordings etc of the same original recording. IMHO the Blu-Ray machines were superior even when a CD was played through them. The more expensive ($670) Panasonic Blu-Ray player and even the ($270) Sonique Blu-Ray player beat the relativly expensive SACD machine by a more than decent margin. What is the future? If there are more Blu-Ray music discs I think that there will be a cat amongst the other formats for a very moderate price. Then there will be the hibrid Blu-Ray music machines no doubt!.

Cheers,

Tl.
 

[Tinker]

However, I heard the shootout between CD, SACD and two BLU-Ray players last evening. The guy who demonstrated the various players had CD's, Blu-Ray recordings etc of the same original recording. IMHO the Blu-Ray machines were superior even when a CD was played through them. The more expensive ($670) Panasonic Blu-Ray player and even the ($270) Sonique Blu-Ray player beat the relativly expensive SACD machine by a more than decent margin. What is the future? If there are more Blu-Ray music discs I think that there will be a cat amongst the other formats for a very moderate price. Then there will be the hibrid Blu-Ray music machines no doubt!.

Thanks for reporting back to us TL.  So the Blu-Ray boxes were "stock"? No outboard DAC?  If I understand you rightly there were "native" blu-ray audio titles on show, but even standard CD titles sounded better through the blu-ray player. Promising!

T.

[AKSA]

This whole topic is difficult and contentious.  More bandwidth is trickier to control, particularly where stability might be marginal and EMI from cell phones, radio and TV is all around.  Furthermore, there are artefacts in digital sources which can seriously upset some amps and preamps.  Band limiting becomes a necessity, but the indications are that more bandwidth, to a point, improves the sense of air and ambience.  This is certainly the impression I gained last night and on previous occasions, at the very same Melbourne Audio Club meet as Laurie, aka TLiner.

Cheers,

Hugh 

[Felipe]

However, I heard the shootout between CD, SACD and two BLU-Ray players last evening. The guy who demonstrated the various players had CD's, Blu-Ray recordings etc of the same original recording. IMHO the Blu-Ray machines were superior even when a CD was played through them. The more expensive ($670) Panasonic Blu-Ray player and even the ($270) Sonique Blu-Ray player beat the relativly expensive SACD machine by a more than decent margin. What is the future? If there are more Blu-Ray music discs I think that there will be a cat amongst the other formats for a very moderate price. Then there will be the hibrid Blu-Ray music machines no doubt!.

Cheers,

Tl.

TL,

I appreciate your insight in this, yours is one opinion i like to consider when debating these Human audio perceptions. Unfortunatly i do not share the same experience, and i do believe that the results you heard may be influenced by bad choice of equipment.
In the past when SACD came along, there was just the same debate and despite i agree on SACD superior sound, dued to mainly in my opinion for its better mastering and NOT for the wide bandwith it throws at us.
I can say i have heard CD players that sounded way better than respectable SACD players, on the same system.

What unit in fact were demo'ed ? Was there any recent CD player from a known good brand ? I recently auditioned NAD's latest CD player and found it to be very good. I would like to see it compared to a BD soon. If a recent $700 CD player is trashed by a $700 Blueray player...i am sold.

If the BlueRay player on the show, improved the sound of a CD, that is just proving the flaws are not on the CD/BD/SACD but on the pure hardware of the machine itself, not the technology of the discs. Better DACS or just better transport is enough to evidence that.

We have seen MAJOR audio companys like KRELL, Audiolab, Theta, etc, adapt its Home Cinema Processors to High Def Audio, making them capable of reading HD signals, but NONE of these companies launched a BlueRay player. Still they are selling DVD players for thousands and thousands. I believe they had to adapt their processors so the general public could play BD in their expensive systems. The reason they do not produce an "audiophile" BlueRay player is because maybe they just dont believe it is necessary to increase the audio bandwith much more !!

[Seano]

The reason they do not produce an "audiophile" BlueRay player is because maybe they just dont believe it is necessary to increase the audio bandwith much more !!

The more obvious reason is that they don't believe that there's money to be made from licencing, designing, building and selling a Blu-Ray source....at the moment.

Nothing to do with band width or other technical malarky.....first and foremost, it's all about the lucre.

[Carlman]

(disclaimer) I'm ignorant about any studies on real life sound reproduction and measurements of them.  I'm just trying to wrap my head around why I'd want more bandwidth...

What bandwidth is real music creating and over what range?  e.g. Is the range on a tuba -20db at 5Hz, 0 at 40Hz and -5db at 100Hz for a given note?   If real things produce sounds in the nether-regions of bandwidth then, yes, I feel certain it's worth pursuing recording and playback technology that captures it and has the ability to reproduce it.

-C