AudioMasters

http://theaudiocritic.com/blog/index.php?op=ViewArticle&articleId=41&blogId=1

Discuss!

And if this subject has caught your eye, David Griesinger's home page http://world.std.com/~griesngr/ should interest you as well, if you're not already very familiar with it.  Several weeks of reading to be had there on all sorts of audio matters.

Reading further I've now gone through the powerpoint document at http://world.std.com/~griesngr/intermod.ppt - and it makes really intriguing reading (and I think that the soundcard manufacturer beginning with "C" referred to in the document is well known here!).  The author concludes at the end of his presentation -

"Conclusions for “High Definition” Audio:

Adding ultrasonics to a recording technique does NOT improve time resolution of typical signals – either for imaging or precision of tempo.  The presumption that it does is based on a misunderstanding of both information theory and human physiology.

Karou and Shogo have shown that ultrasonic harmonics of a 2kHz signal are NOT audible in the absence of external (non-human) intermodulation distortion.

Their experiments put a limit on the possibility that a physiological non-linearity can make ultrasonic harmonics perceptible.  They find that such a non-linearity does not exist at ultrasonic sound pressure levels below 80dB.

All commercial recordings tested by the author as of 6/1/03 contained either no ultrasonic information, or ultrasonic harmonics at levels more than 40dB below the fundamentals. 

Our experiments suggest that the most important source of audible intermodulation for ultrasonics is the electronics, not in the transducers.
Some consumer grade equipment makes a tacit admission of the inaudibility of frequencies above 22kHz by simply not reproducing them.  Yet the advertising for these products claims the benefits of  “higher resolution.”

Even assuming ultrasonics are audible, loudspeaker directivity creates an unusually tiny sweet spot, both horizontally and vertically".

The author notes earlier that sample rates of 48kHz and above were largely justified by the limitations of early AD DA converters, but that even inexpensive converters were now able fully to reproduce all frequencies which can be proven to be audible in musical signals at 44.1kHz (I hope that summary is correct).

I suspect those wedded to 96kHz etc recording standards will not accept a word of this, however.  I've always been totally wedded to 44.1kHz as representing all I can hear (and more), and therefore of course I accept it with glee!

I'm with you 100%, Peter.

A couple of years ago I was listening back to a live concert recording of some chamber music which had been recorded multichannel at 24-bit with a back-up direct to regular audio CD-R. The replay was on a good hi-fi set up in a pretty poor listening room and the performers were mulling over their performance. Also present was a prat from the company which had supplied the recording equipment - he'd flown over from London to fix (again) the Tascam multitrack HDD recorder. Whilst we were trying to listen to the recording he kept blethering on about how obviously inferior the 16-bit CD was to his 24-bit recording, and couldn't we hear the 'graininess' in the sound from the CD.

I resolved never to consider buying audio equipment from his company...

So it ain't my ears and gear making me not hear any difference...

The only advantage so far I found is that "they" seem to take more care about those sacd/dvd-a recordings.

In the real world, the punters don't get to compare and contrast anyway.  For instance, over here the local radio stations each get to record the major touring classical performers in their city's major hall.  Some of these guys use pretty high-end equipment at 24/96, others use similar equipment at 16/44.1, others perhaps use less exotic equipment such as good old minidisc.

Occasionally we compare and contrast the results we've got, and there are inevitably significant differences - but those differences plainly arise from different mic techniques, different mics, different halls, different post-pro methods and tastes, etc etc.   Discerning amongst those variables the 24/96 recordings from the rest, even if you admit there are audible differences, would be a hopeless task.  And that's what the record-buying public are supposed somehow to be able to do, even to the extent of funding the costs associated with the more "advanced" ways of recording - whereas it's increasingly clear that in practice the bulk demand is for mp3.  The commercial case seems to me to be very hard to make.

Yep, I knew it: more snake oil, green markers, and welder's cable for mains.  Oh, what? We have some sensible science on the Internet?

The commercial case has been hard to swallow even for the small part of the public that cares.

So the Journal article isn't on the 'net.  That would be a good read.

But I like the idea that 1) higher bit depth gives more slop room and 2) we're talking current-standard converters here, or the bets are off. 

This is a nice confirmation of what most of us have agreed on long ago: there are greatly diminishing returns past a certain point.  What's bold here is placing that point where audio was set 25 years ago!  Great fun.

I'll still record 24bit whenever I can, to leave "sloppy" room, and surround sound is a different animal altogether.

He's not the only one...


I've heard the odd horror story about the Tascam multitrackers as well - rather overpriced and overhyped - especially compared to the Alesis, that just keeps on going, is easy to use and sounds at least as good most of the time, and a lot better when the Tascams aren't working!

I'm well pleased with the HD24XR - best multitrack machine I've ever bought.



When you consider just the mics and the halls, the differences will completely outweigh any fractional differences in performance, even between only moderately competent equipment.

I have, on one occasion, been very naughty. I've played back a recording from a CD, where I have processed the HF from about 7kHz upwards to have a very slight upward tilt - it's got to no more than about 3 or 4dB @ 20kHz,  to somebody (who couldn't see what I was doing) in direct comparison to the original material, played from the Alesis - so that's 24bit 44.1k. It won't take a genius to work out which one the punter swore blind was a 96k recording, would it?

I have a few mics that will produce output above 20kHz when provoked by trumpets, etc. and on the occasions that I've recorded the output for any reason, I've always rolled them off above about 18kHz, probably got cleaner recordings, and never heard a word of complaint from anybody about the result.

The other reasons for rolling off everything above 20kHz, even if you are recording at 96k (which I don't, except for experiments) are primarily two-fold. I'd roll it all off anyway, just to stick a finger up to all these people who insist that it's necessary, although none of them can actually tell whether it's present or not... and the more practical reason, when doing the experiments, was ultimately because of the one creature who can hear it - and doesn't like it - and that's the cat. Oh yes, I can reproduce content very comfortably up to 35kHz or so; that's where the Adam P11 ribbon tweeters start to roll off. And if you play anything like that through the system, with all of the LF removed, he gets pretty cross. Everybody thinks that dogs are the creatures with the extended hearing response - but cats can hear a lot higher up than any dog can.

This discussion reminded me that we have a (supposed) pest control device which emits ultrasonic audio, and which makes a handy test of the upper reaches of audio reproduction.  I've just heaved it out from the back of a cupboard and plugged it in, and recorded it from about a foot away with the little Zoom H2 running at 96kHz, 16 bits.

Looking at the result in Audition. one can see the result in the Frequency Analysis display very clearly.  With the noise floor knocking around -108dB, there's a sudden increase in level at 35kHz up to -72dB, which disappears back into the noise floor at about 45kHz.  The spectral display shows that the ultrasonic device emits a sweep between those frequencies with a period of about 1.3 seconds, and in fact it probably goes up to much higher frequencies still, judging by the shape of it.

Needless to say I can't hear it.  Nor can the dogs (they seem unconcerned when the device is turned on).  Attempting to record it playing back on Tannoy 607 speakers results in nothing significant showing, apart from a curious spike at 45kHz which I suspect is coming from something else (what?!). 

Conclusion - the Zoom can, remarkably, make ultrasonic recordings, but it looks like my speakers can't reproduce them.  I guess I could create a test tone in Audition to see, via the Zoom, what they can do at the top end.

These days my hearing gives out at about 12kHz at best.

I did once work with a producer in her late 40's who registered very high frequencies as pain.  We had a squeak in a piano which for the life of me I could not hear, but she ripped off her headphones exclaiming "ouch"!  Slowing down the recording later on, indeed there was a squeak at about 18Khz which I could not hear, but she found most unpleasant.  It would have been interesting to have had her on the SACD investigation listening panel.  If anyone could have heard the SACD difference it would have been her.

A little more on topic - see http://www.bostonaudiosociety.org/explanation.htm for details of the actual test gear and locations.  Apparently people in other forums are doing their best to rubbish this study on the ground that "other people using other test setups would be able to tell the difference" but I don't think they've realised the comprehensiveness nor the high standard of the test.

When a student, Michael Gerzon had a measured response up to 23kHz.  I have heard, curiously, that an extended response is somewhat more common in severe asthmatics, as he was.  With a few people like that around, it is not surprising that the best previous study I know of indicated that 44.1kHz and 48kHz could sometimes be distinguished, but not any higher sampling rates.

Paul

The problem with all of those tests is that you are never sure exactly what you are listening to, and there's a good chance that it's not the sample rate change per se  that causes the difference in what is percieved anyway. Even if the anti-aliasing filters are perfect (haha!), just running the same converter at a different rate will alter things like the monotonacity, and what happens on the leading edges of every step change, which will almost certainly make minute differences to the summed output - within  the audible range, never mind outside it. The real culprits, in earlier tests, were the anti-alias filters though - had a sound all of their own!

What I don't know though is whether anybody has repeated any of these tests seriously with modern over-sampled D-A systems, which behave far better than their predecessors. I suspect that even if the tests could be run without serious experimental problems, they could only be done with headphones to be in any way reliable. And where do you get enough sensible children to test this on? I find it very hard to believe that there would be very many adults at all who could reliably distinguish any difference.

It would be interesting to try this with Apogees - a lot of people think that their converters running at 16/44.1 sound better than some of the 24/96 kit around.

The link between asthma and hearing is interesting, and I have no idea what a causal mechanism might be. I was mildly (not severely) asthmatic for many years, and my measured hearing response held up pretty well into my 20's, at least - although it never extended to 23kHz, I'm pretty sure. Well, it might have done when I was about 3, I suppose, but nobody tested it then...

Funny - as I recall it the producer with the extended hearing ability that I mentioned early had an asthmatic tendency.  Interesting.  And many years ago I read that there was some connection between asthma and the tonal analysis of a mother's voice.  Kind of circular!

Heck, there's some unbelievable crap being posted in other forums concerning this paper.  One guy suggested that only people who had spend a long time acclimatising in a studio should have been included in the listening panel.  Or, presumably, be allowed to buy "high res" players.

I've just seen this result attributed to "bias"!  Someone has no concept of what testing means, I think.

Paul

If anyone is interested enough to do a search, there was a thread not more that a few weeks ago on this paper on the Hydrogen Audio Forum
http://www.hydrogenaudio.org/forums/index.php?act=portal
that quoted a few relevant parts of the paper. Towards the end of the thread is a post from one of the paper’s authors and also some links to where he wrote pointed replies to various specific criticisms in some of the audiophile oriented forums.

We have a similar device. When I bought it I plugged it in and waited for the dog to doze. Then I turned it on. He lifted his head and looked around, then got up and quietly left the room.

Whilst the cat's not too bothered about my ultrasonic cleaner (the energy pretty much stays in the liquid), he gets extremely  distressed if I dare to switch on the ultrasonic leak detector that I use for door seals, dodgy walls, etc. This puts a lot of energy into the air at 40kHz, and he can hear it from several rooms away - enough of this energy gets around pretty much any normal house door, and it's well within his hearing range - which according to some sources goes up to about 64kHz - way beyond a dog's, which on a good day will extend to about 40kHz tops. Porpoises and dolphins have probably the best range though - it extends up to about 150kHz.

Thanks, Andy.  I've just spent a happy hour going through http://db.audioasylum.com/cgi/search.mpl?author=EBradMeyer&user_id=45230&forum=ALL&sortRank=None&sort=date&sortOrder=DESC which is one of the author's responses to criticisms of the paper.  It makes the most delicious reading.  I should probably have been doing other things but I just couldn't stop reading the repeated nailing of "audiophiles" to the wall.

So I'm at SBES the other day - Wildduck and ryclark have departed, and I have to hang around for my train, so I went over to the Tannoy stand and had a look at their current monitor range. Tannoy's MD is there, and I asked him to explain about the supertweeters they now fit - response up to 50kHz... You have to bear in mind that I have a badge on that says Acoustic Consultant, and he'd obviously clocked this. My opening question about them is "so you're marketing to bats these days?"

The next five minutes was fascinating. Apparently, although we can't hear anything coming from these supertweeters, they improve the apparent performance we get from their monitors. This information was imparted to me (I didn't interrupt - it was wonderful!) with so many caveats, exceptions, you name it, that I wouldn't have had a pair of them if he was giving them away!

But he's not giving them away at all - they cost a bloody fortune if you buy the retrofits. He knew that he wasn't about to sell me anything, because I'd already told him that I had a pair of Tannoy Devons (flat to 20kHz with no supertweeter) that I was quite happy with as location monitors. But apparently I almost hadn't lived until I'd heard this extended response... (perhaps)

Eventually I told him that I had a pair of Adam P11's as well, and that I couldn't hear any important HF differences if I monitored on those. And that's with mics/pres that he's certainly not going to argue with, so that's okay... He asked what the claimed response of the P11 system is, and I pointed out that the ribbon tweeters in them measure inherently flat up to about 35kHz.

He didn't actually go as far as to accuse me of being deaf, but I'm sure he wanted to. I reckon that my hearing's at least as good as his - and probably better. But I can't hear up to 20kHz - or even get that close, come to that. But I never get complaints about HF levels at all. There is some serious BS going on here, I think!

For once I'm currently recording a CD of piano works at 24/96.  I've just looked at one long file of takes, scanning selection in AA frequency analysis, and basically there's nothing but noise above say 16kHz.  At least with this material, there seems to be no point whatsoever in the higher sampling rate.  I was sufficiently amazed by what I saw that I've just set the equipment up in the kitchen with the anti-rat HF generator, and it is actually recording in the 20kHz+ range - so it's not an equipment defect.

Same with organs, often noted for having a somewhat 'shrill' sound - which means that there's often a lot of HF content up to around 17kHz or so, but nothing of any significance at all above that - just noise. Even with brass instruments, where there are sometimes local artefacts above 20kHz, you wouldn't want them there anyway - they don't enhance the resultant sound one bit!

And it's not that I'm failing to capture what's there - all the mics I currently use have a response that's essentially flat up to 20kHz, although the DPA 4006's can be altered acoustically to reduce or extend this somewhat, with different polar characteristics.

With real recordings of real instruments in real rooms, radiation above 20kHz reaching either microphones or audience to any significant degree is a myth.

Just to be sure, before this morning's session properly started (ugh! Sunday morning sessions are not my favourite!) I walked round the piano jangling my keys while recording - which generates a surprising amount of sound above 20kHz - and later I'll verify that it shows up, and that therefore there's no doubt that the lack of HF elsewhere in the recording is not my fault.  The instrument does sound perfectly bright - possibly a bit over-bright, if anything.

Just to add that I've now looked at a frequency analysis of a keyshake test conducted at the start of the most recent session, as compared with the analysis of the subsequent piano recording - this confirms that the equipment was recording well past 40kHz but that the piano recording only has noise above 16kHz or so.  To my mind this proves that for solo piano, anything more than 44.1Khz is a complete waste of effort.  (It would be interesting to analyse any SACD solo piano recordings there may be).  Other instrumental forces may well produce quite different results, though.

Even if these frequencies accidentally reach the mics, I still don't see how they could be significant - I can enjoy music just as much without them, and I always have done.

What has always intrigued me about this is that all recordings and reproductions are in themsleves very poor reflections of what it was like to actually be there anyway - so why is there so much fuss made about a part of what's perhaps present near the performers only anyway? It's certainly not the impact of ultrasonic frequencies that makes recordings a pale shadow of the real thing, is it?

In other words, if you made a recording such that when it was played in your living room, it effectively transported you to the location (in audible terms), I think that the shock of that much realism would rapidly make you forget all about the possibilities of ultrasonic frequencies  - they wouldn't even be the least of your listening experience, because you wouldn't 'experience' them anyway unless you sat amongst the performers - perhaps - and who would want to sit there for a listening experience?

The producers of not a few "surround" (i.e. 5.1) recordings, I'm afraid.  I have certainly read one arguing that it must be preferable to sit in the middle of a string quartet!

And on the realism bit, that is why I have long been a supporter of developments like ambisonics - even though this is still doomed to incomplete success because of the remaining influence of the listening room (which may also be why ambisonics is most widely used for large-scale sound projection).

Paul

Yes... I'm aware of them. Fortunately for common sense, they don't count. If they persist in trying to sell gimmiky sound, then as far as potential purchasers are concerned, they are going to fail. And who's that doing any favours for? Pretty much nobody except themselves, as far as I can see.

I have no objections in the slightest to ambisonics, to the limited extent that it works. But even to get that far, it's staggeringly impractical - which is why it will never catch on. I've tried a few ambisonic experiments with my Soundfield MkV, but in practical terms, I've only ever used it as a steerable stereo mic with some rather useful post-production characteristics. And I've heard a few of these larger-scale ambisonics demos too. Every single one of them has not caused the suspension of acoustic disbelief on my part... although I'll freely admit that they could easily have been improved in acoustic terms; the venues weren't really ideal for any sort of reproduction, never mind surround.

Since when was "practical" the main consideration?  Stereo was initially seen as domestically impractical; quad was seen as impractical; but none the less, quite a lot of people now have a grotty 5.1 system - and I believe it is possible to coax a better impression of space out of those systems using ambisonic techniques than by using the weird and wonderful collections of ironmongery with microphones sticking out of them that I have come across.

Paul

I think that practicality is a consideration for a lot of people, actually, and that 2 was a good number for a lot of practically-minded people to stop at, both in terms of being able to manufacture the apparatus to record and play on, and in terms of domestic acceptability. Heaven knows, it's been hard enough here to get even a couple of decent loudspeakers into our living room (we are currently in a failed state from that POV), never mind 5.1 - *SWMBO's objected completely and utterly to this, even though it would be mostly for her benefit, because I'm not the one watching the films. And I know of a lot of similar households...

I've personally yet to see a well set up domestic 5.1 rig, although I've been told that there is one somewhere... and all of the attempts that I've seen have been flawed to the extent of being no more than effects systems, and not very good ones at that. But since that was all that the format was really intended for, perhaps that isn't so surprising.

I agree entirely about the strange hardware concoctions though - no good will come from any of them, and technically they make no sense. AFAIK the only strange hardware multi-mic concoction that's ever found any degree of long-term acceptability is the Decca Tree, and that's only really necessary to compensate for the appalling state of most listening situations. But those strange five-arm constructions with the Neumann mics on look for all the world like gold-plated turds, and probably sound that way too. Have you ever heard anything that's knowingly come from one? I don't think that I have.

*If you get the impression that SWMBO doesn't appreciate good sound, you'd be right - she has other stirling qualities, though, and much better visual sense. That said, I am getting a bit fed up with all organ music being referred to as 'hooting'....

I've never seen a domestic installation of a 5.1 system.  Plenty of stereo systems with one speaker on top of the bookcase and the other behind the sofa, though.

Debates like this are always fascinating - apart from anything else, it forces us to confront the prejudices and experiences of others.

I have seen the results of my colleague ozpeter's experiments with the WMD (Weapon of Marsupial Determent) which are perfectly valid.  The results on the piano are as one expects for such a large scale high inertia source.  However the key jangle tests might be more representative of cymbals, bell trees, and sitars plucked or struck with hard objects.  Now whether he found the key jangles more realistic and musically satisfying, we didn't get round to ...  (And I'm not sure that I would want to listen to these sources at the distance the mics are usually placed.)

The AESJ paper at the heart of this is interesting but I would have some qualms with details of the methodology and configuration.    But publication by a learned society is merely an invitation to others to attempt to replicate the test and achieve the same results.  Wait a year or so and see what follows (Malcolm Hawksworth, over to you ...)

There is an inherent bias in some of the comments here that by the time (through diligence and experience) we achieve the status and respect of audio luminary, the usual process of decline has set it, and the top octave of our hearing capability is a distant memory (like me - my left ear is 60 years old, my right ear is 40 years old, stereo monitoring on headphones is most unreliable).  But those of inferior years who have not reached luminary status but have their hearing by and large intact may have a different view of that top octave.

I am much more inclined to look at differences in the amplitude/time response of different configurations rather the usual amplitude/frequency.  And I want to know if the results can be duplicated in different populations (European, Asian, South American).  Maybe somebody like Prism could produce a version of their converters optimised (doesn't that beg the question) for 16/44K1 and compare the results with other configurations including say 88K2/24 native, and likewise subsequently reduced to 16/44K1 all played back on the same configuration.

But I'm also a broadcast engineer (pragmatist) and I know that even 16/44K1 exceeds by a wide margin the capability of our broadcast chain so recording straight to CD is highly satisfactory for our primary purpose.  And I recall my dealings with a Sydney based audio guru from whom I purchased many years ago my long neglected wideband AM tuner.   This is someone who would 'upgrade' your favourite bit of audio gear to remove all those nasty commercial impediments to good sound.  He remarked that the best sound available to him was direct broadcast from the Sydney Opera House via ABC-FM.  For this particular situation, it was physical program lines from the OH to Broadcast House, through broadcast standard analog gear (and some switching equipment that might have been a bit passe) and out again by physical program lines to the transmitter.   The amount of transformer iron in this circuit would endear it to some of the analog fraternity, but for me the  killer was knowing that everything went through a rolloff above 15 khz and significant (>50dB) attentuation at 19kHZ.   So promoting his gear as having response extending way beyond 20kHz was maybe a bit of hypocracy on his part ...

Jeremiah (crying in the wilderness) Ch 17 v 9 (King James) says:  "The heart is deceitful above all things, and desperately wicked: who can know it?"  Remove the first and last letter of the second word, (error in translation?) and the result is equally valid.

That's been discussed by others - mainly people who will not believe the outcome, even if the rest of the world accepts it. Certainly Greisinger's researcher's credentials, and his research methodology are beyond dispute as far as I'm concerned - speaking as a researcher. The right thing was done, in the right way, as he always would. But I'm slightly more concerned about some aspects of some of the Brad Meyer/Moran setups, I must admit...


Well, I might be getting older, but I have a perfectly fine memory. I can recall being able to hear dog whistles well into my 20's (which with hindsight proves, I suppose, that my hearing range may have been a bit more extended than I thought, because most of my peers couldn't do this). The other thing I can remember well is what real instruments sounded like then, in comparison to good recordings. And with individual instruments known to exhibit ultrasonic sound (mainly brass - they are actually the worst offenders, I think), you could achieve results which, if listened to on good monitors, sounded timbrally pretty much identical to the real thing - without the ultrasonic part being recorded at all, because the analogue equipment response stopped a little short of 20kHz.

I was lucky - I actually had the opportunity to experience this as a part of somebody's research. But I do recall, even then, realising that anything that actually had a flat response up to 20kHz was going to do the business, regardless. So if I'm biassed - which I might well be - it's because of an experience I had at an age when if there was a difference to be heard, I probably would have heard it.

So regardless of whether one could actually hear the ultrasonic content or not, I still don't think that it makes a scrap of difference worth having to the percieved timbre of an instrument.


Without getting into a huge discussion about it (and it would be huge), I have to say that the converters themselves are nowhere in this (excluding anti-alias filters of course), because the single largest distorter of transients is the transducer that gets them back into the air - by a long, long way. Even the best normal monitors you can get make quite a mess of transients (for anybody else wondering, that's the initial amplitude/time response of a loudspeaker/headphone). And it's for this reason that the only room system that is least likely to create massive transient smear is the 4th one in Brad Meyer/Moran - the system with the ELS's in it. That would have an exemplary step function response, and the lowest transient smear you can achieve. The only other things that get close are small full-range drivers like Auratones, but they're not a patch on an ELS.

The listening test I did was using original Quad ESL57's, incidentally, and the rest of the chain was state of the art (back then) analogue.

You made my day... 8-)


The day that will be better recognised a lot of companies will go out of business. I'm out of the loop these days, but is there still any research going on to improve this end of the chain? It looks as flapping cardboard is going to stay forever.

While I completely concur with Steve's conclusion (even when you can hear above 20Khz, the timbre of an instrument is unaffected), I think this is slightly overstated.  Lots of editing in spectral view (with B&K 4006, AKG 414, and Neumann KM-84 mics as the mic sources) has shown me what goes on above 20K is measurable, though rarely significant, and probably never audible.  Percussion, of course, gives the strongest upper energy, and loud trumpets can be pretty interesting too.  But the one that really piques my curiosity is... violins.  You can have a whole orchestra nearly drowning out the strings but still see the unique fingerprint of violin vibrato up in the 20-24Khz range -- the only thing prominent in that range!  Can you hear this?  Of course not; it's masked by the substantially louder amplitudes of the rest of the spectrum.

So pick your myth, I suppose... to the mic and Spectral View one could make the case that something significant is there.  But not to the ear.

I'll have to try this on my young kids and find out what they can hear.

By the way, Peter, if you can get piano overtones to 16Khz, you're doing well.  I usually don't see much above 11 or 12.  I suppose if you miked it a few mm from the hammers, you'd get a much more visually pleasing result (while torturing the ears to a very unmusical one!).

I don't think it is - if it's 'probably never audible' and doesn't affect timbre, then to reach the audience to a significant degree it would have to be loud enough to cause damage - because otherwise it's not significant, as we've already established - because we can't detect the signifier. So I'd say that it was correctly stated - and although contentious on the face of it, it's not really when you examine it a bit.

And the other thing about this, of course, is that you can pick it up locally to the instruments, but nowhere near as much gets to an audience positioned somewhat further away. Most microphones tend to end up at least slighty closer than most of the audience are, and spot mics a lot closer. So from the perspective of the people listening to the recording, you've probably overbalanced the ultrasonics anyway, if you record and reproduce at a sufficient bit rate without attenuating them. As I've said before, I roll this stuff off completely - and I've never had a single complaint about it.

I think that all that violin ultrasonic detritus (on instruments that can be pretty excruciating in the wrong hands, and sometimes in the right ones...*) is probably enhanced by using too much rosin. Best thing to do with that is roll it all off regardless, and do the dogs and cats a favour. Also, on reproduction equipment with less than stellar intermodulation distortion performance, it's probably going to sound better like that anyway, because it won't be exciting problems.

I think that there's quite enough to worry about when making recordings at the best of times without worrying about the stuff that you can't hear as well. I think that recordists should concentrate on getting the best out of performers rather than trying to confuse the domestic wildlife, myself.

*One of my sisters played a good violin extremely badly as a teenager, and it regularly set my teeth on edge. I was mighty relieved when she gave it up, and concentrated on singing instead. Fortunately my own children have all veered away from anything that's bowed and even my other sister, who played the cello for ages, now only plays the double bass - so I'm spared these days!

As an aside, this Cymbal (http://paxel1677.republika.pl/forqueray/Jedrzejow%20-%20www%20foto-Pages/Image8.html) stop is apparently a solely Polish invention - lots of tiny untuned pipes on a single foot for each note.  I guess that close up there's lots of ultrasonics there, and even at a distance it's hardly likely to be described as "hooting"!

Paul

I doubt it very much - and that's based on some recent experience with a load of misbehaving pipes that small in an organ with a lot of short-footage mutation stops. The timbre you get from pipes that length is quite pure in harmonic terms, and doesn't extend beyond the normal hearing range at all. The reason that those cymbal stops are built like that is precisely because organ pipes of this nature don't produce a lot of extended harmonics - but if in that case, it's an artificial HF 'chiff' noise they wanted to add, and that seems like a sensible way to save a lot of extra action-building that you don't need.  In principle, there's no reason why you couldn't use the same approach with the top pipes of a cornet, only with tuning slides on the pipes, so that the result was harmonically related. But with either case, you need the small pipes, because it's a lack of harmonics that you are trying to overcome.

You are more likely to get ultrasonic output from some much longer reed stops on some organs, because of the way that they resonate - in a similar way to what happens with a trumpet. In a trumpet, it's the body itself resonating that causes the ultrasonic output, not the excitation as such. But generally this doesn't happen much with organ reeds either, because the material that most pipes and resonators are made of is relatively well damped, and supported. Both of these factors reduce the potential for ultrasonic emissions quite significantly. In an organ with decorative en chamade trumpets, it might be a different matter when it comes to ultrasonic emissions altogether, though.

So, in my recording of the organ I had the trouble with, there is no output above about 17kHz at all. But, I still think that in many ways, it sounds over-bright. That wouldn't matter so much if this was balanced out a bit better, but it still has some pedal deficiencies, as far as I'm concerned. And if they can sell enough CDs, then perhaps the church's music trust can afford to rectify this. And yes it has to be the music trust - it's the only part of the organisation that's truly viable these days in financial terms, because it's the only bit that draws the crowds.

For once I actually measured the distance from the piano to the mic as I have to recreate the same sound for the last session in a couple of weeks time (well, from the tail leg of the piano to the mic stand) and it was 2.6 metres - FWIW.

Meanwhile a quick check of the recent organ/choir recording I mentioned the other day - which was only 16/44.1 - reveals stuff all the way to 20kHz and more, but watching the waveform, most of the HF content simply relates to esses.  Looking at an organ solo portion only, there seems to be nothing describable as significant about 18kHz at the most.

Interesting stuff!

That all makes a lot of sense. Just in the interests of investigating, I'll record my next session at 88.2k, which the HD24 has no trouble with, and see what, if anything, occurs that I may have missed - apart from echo-locating bats, that is...

The topic discussed here is a highly interesting one and I hope not to offend any of the established masters by plugging in lately. To state it at the very beginning: I consider the distribution of stereophonic material in a higher resolution than the red book standard as a waste that is aimed at the psychoaudiofidelists that consider gold plated AC power lines as an indispensable prerequisite.

Despite that global remark I would like to expose some more detailed experience and opinions. I try to structure this a bit as otherwise, not using my native language, I am prone to get struggled.

1. Ears: When I was about 14, I had the chance to test my ears at an audio fair and people and myself were astonished that I could perceive 22 kHz. At that time I also remember that blowing a dog whistle close to me caused heavy pain. BTW: I was never asthmatic but suffered from many middle ear inflammations. At 20 to 24, when I was working as a part time sound assistant in early TV to earn money for my engineering studies I also recall that all the 15.625 kHz singing from the monitors was annoying. That has gone as my ears are 70 (both !) and have a cut off at 13 kHz, which is reasonable for my age but bad for Hi-Fi. Nevertheless I assume that I still can distinguish between good and bad sound reproduction. Question: Should we provide frequency response to beyond 40 kHz just to repeat the pains of a dog whistle to some eccentric people as I was (and still am) ? NO !

2. Ultrasonics: The battle is still open about how much spectral power is produced by different sources of sound. I tend to assume there is much more ultrasonic content as commonly believed. This holds certainly for many (nasty) types of noises such as brakes of trains, rattling chains or pressurised hammers. With respect to piano, I have some controversial opinion as to the limits stated as 11 – 12 kHz or 16 kHz. I have recently recorded piano recital, experimentally in 88.2/24 bit format. In AA, I found some spectral traces up to 24 kHz. You may argue this as deficiencies of my recording equipment and not a real acoustical content. I can neither prove the one nor disprove the other. Listening to the original 88.2/24 file and the reduced one to 44.1/16 did not reveal the least difference. Thus, my initial statement stays, hopefully not only due to my personal lowpass in the ears. Unfortunately, a recent Jazz session I had taken only in 44.1/24 format – yet there are components up to 20 kHz. Question: Does the possibility of ultrasonic presence influence the quality of recording and reproduction ? NO! (Except the dog whistle for borderliners)

3. Microphones/Loudspeakers: The majority even of well reputed microphones are not suited for ultrasonic recording. The exception known to me (apart from some types built for measurement ) is the one of Sennheiser. The commonly used small diaphragm types may extend slightly beyond 20 kHz, depending on size. The family of large diaphragm microphones are physically limited to an edge frequency below 20 kHz. Among these is also the admired Neumann U87 which is down 10 dB at 20 kHz in cardiod position and even worse in figure of 8. Despite that I would vividly welcome to get a pair as I cannot afford them for personal use. For loudspeakers there is much more need to get an overall smooth response and a rigid low fundament. The latter is often lost in the modern trend to small cabinets. I admit the ribbon high frequency unit of the PA11 as a very brilliant (sometimes a bit overly) reproducer.  For the overall impression I like to stick to the old 15” Tannoy coaxials in a Westminster-like cabinet, especially when refurbished with a new rubber suspension. My Abessins, unfortunately they passed away 4 years ago, indicated a sine close to 20 kHz by a very strong look at me. They could only be stressed when playing back the purr of an unknown cat. It is a shame that Tannoy corrupted their own elegant coaxial principle by the 50 kHz units, that are neither coaxial nor make otherwise sense. In my eyes it is a step down from the reputation as for at least 2 or 3 decades, the Tannoys were the only monitors found in the studios of Swiss National Broadcast and TV. Question: Is it a primary sign of quality that microphones and Speakers go into ultrasonics ? NO !

4. Recording Equipment: Steve pointed out that different quality in reproduction from 96/88.2 kHz and 48/44.1 kHz sampling may result from the non ideal antialiasing filters. Those familiar with DSP will certainly agree that filtering to 20 kHz is easier (and therefore usually better) from 96/88.2 kHz sampling than from the lower frequencies. This may be a reason to master at the higher sample rates. The pay is longer processing time for editing and mastering and the outcome barely audible. Question: Is it worth that effort ? I do not know ! Better to invest more care into the recording situation such as microphone placement

I remember when both I and television were quite young, I could stand outside a friend's front door and tell from the high frequency whistle whether their TV was on or not - if it was on I would leave without asking if my friend could come out to play!  As I grew older, they seemed to improve TVs so that the noise could no longer be hear.  But of course, it was my ears already going downhill...

Oh, I second that.  I never hear this anymore, but as a kid I also remember being annoyed.  That frequency is still perceivable to me -- if it's loud enough.

By the way, welcome Bert, and interesting to hear your perspective on the matter. 


I always assumed it was like the human voice: a spectrally-rich puff of air is produced at the source (almost like a triangle wave) of a rather unpleasant tone that is filtered and enhanced along the acoustic path of the sound.  But now that I think of it, listening to a brass player warming up on a mouthpiece alone isn't likely to induce stress from ultrasonic frequency overload -- it's not a bright sound at all.

I hate to remind the English-origin members about hearing line whistle from TV sets, but when ozPeter was very young, UK TV was predominantly the 405 line system with a line frequency of 10,125 Hz, which I also found very audible when working there in the late '60s  It improved with age because the UK migrated to the PAL 625 line system where the line frequency was 15,625Hz, which I could also hear at the time, depending on how well the EHT transformer was made, and how much it exhibited magneto-strictive behaviour.

And at the risk of repeating myself, I think perception of the effects of higher frequency cut-off is a time based thing, ie., transient behaviour rather than steady state.   While theoretically an appropriately configured FFT analysis could show this, wavefront behaviour is not something that will show up in a quasi-steady-state analysis such as FFT analysis used in most DAW programs.  Can't suggest how this might be studied effectively at the moment.  One could do tests with artificially generated pulses, but relating this to a music response may be difficult (except perhaps in the case of certain techno-music idioms ...).

Heh, that's a very good point!  They moved the goalposts!

Not just the Line Output Transformer (aka LOPT - the actual cause of the whistle - EHT generation was/is its additional function), but the line coils on the tube neck could emit sound as well in some cases - modulating 50Hz. I don't think that most coils do this - they are supposed to be pretty solid constructions. But some of the early colour sets with scan correction coils and magnets beside the main coil assemblies weren't too wonderful at all from this POV.

I could hear that horrible whistle until relatively recently, but I can't now for two reasons. Firstly because my steady-state hearing (you may well have a good point about transients) doesn't extend that far, and secondly because we've recently got rid of all the CRT TV's that are in the house anywhere near where I might be able to hear them!

This is one of the most interesting and unusual discussions I've come across in a long time.

I do recall reading early on in my (self-trained) audio "carreer" of some pretty complex mathematical constructs that basically "deconstructed" HF components into (very low-level) modulation artefacts of lower frequencies. It was mostly out of my league, but the gist was that there WAS (past tense?) some argument that HF components could modify low frequency components (mostly low-level lower frequency) as harmonics. If I can find the paper (I got it during my juniour membership in the IEEE), I'll see if I can make a reference.

I know that reading the original blog report left me with a bunch of questions about the methodology, the characteristics of the sound source(s) (identical isn't necessarily NOT a variable, if the characteristics were limiting all the audio experiments - a cardboard woofer and a $2 tweeter would definitely make someone think that there was no difference between a well-mastered HD recording and a 22k downsample of the same thing), but I have to assume from other comments here that no part of any of the equipment used was limiting or colouring the results in any way.

Maybe they should have used Monster (TM) Oxygen-Free (R) Coaxial 240V cable... :mrgreen:

I don't recall any papers ever where anybody actually tested this rigorously and came up with a statistically meaningful result in terms of being able to percieve any differences - I think largely because the ear can deconstruct whatever's there to a large degree, and if an additional component was added, the ear would effectively remove it again if it wasn't recieved (by mechanical means) in the first place. There is, however, a body of research that indicates that extreme HF content makes no difference whatsoever to pitch determination, or the perception of timbre...


One of the reasons for testing this on a variety of systems is to eliminate some of the issues that go with the concept of 'identical' - by testing on a variety of systems that have been determined to be 'good', and not relying on one system in particular. But as far as I'm concerned, if you can't tell the difference when you use exemplary equipment with Quad ESLs (which have staggeringly good/accurate transient and phase response) then it simply ain't there - which it isn't, of course.


We have a thread about that somewhere - there are circumstances where that appalling monster stuff can actually make things worse, not better!

I'm sitting here listening to an Earthworks demo CD (www.earthworksaudio.com) for their "HDM High Definition Microphones" which tout impressive impulse response (i.e. reduced ringing following an impulse) and flat frequency response to as high as 50Khz.  They're suggesting that this accuracy improves the detail and believability regardless of the audio chain (44/16 CD valid, so are cheap mixer preamps) because the accuracy affects the entire frequency spectrum.

I don't immediately hear more detail; I hear a more aggressive mid-upper midrange, which flatters some things and sounds unnatural on others; and I hear more room ambiance (translate: short reflections/slaps).

So if any of you want to go after a technical writeup about this (Steve, is this a bunch of stuff not strewn together very well, or is that just the state of my flu-recovery brain?),  here's an article  (http://www.earthworksaudio.com/tech/world_beyond_20khz.pdf)that serves the purpose.

By the way, the author of that article is David Blackmer, founder and engineer of Earthworks, and rather notably also the founder of DBX.  No slouch engineer.

Yes, but it's still self-serving, isn't it? And on that basis, I'm afraid no credence can be attached to it.

How many independent ABX tests has David Blackmer taken part in where the results have been published? Can you find any at all? I'm afraid that in the real world, none of his claims have been shown to be repeatably demonstrable.

And as for his reputation - well, it's very easy to trade on your 'reputation', and in Blackmer's case it helps to recall that this was the man who foisted the DBX noise reduction system (with some apalling aretefacts) onto us with no apologies at all. Personally, I can't take anything he says about 'quality' too seriously.

DB also introduced a delta-modulation digital recording system in the early '80s (I probably have the AES pre-print hidden somewhere)  along the lines common at the time.  Processor which outputs data on a video pedestal, recorded on a one-inch helical.  One majpr issue at the time was variable quality of world-wide distribution of cutting masters.  Put the digital recorder across the output of the cutter feedback coil as usual , but then clone it as many times as needed and send them off. Each one identical to the original.  Better than the 4th or 5th generation dub that often turned up.

The exact specifications (clock frequency, effective S/N) elude me, but there was some degree of integration in the coding to deliver appropriate signal to noise, possibly some analog compansion as well.  However, I clearly remember that the perception of the AES group listening to the results, was the improved (it's that word again) "musicality" of the result.   Possibly the better standard of conversion, possibly the lack of sharp cutoff anti-alias filters, maybe the higher tolerance of the ear to slope overload rather than peak overload, and the suitably of delta modulation for a basically triangular spectrum.  Of course all the production issues that plague DSD applied equally then.  Now confined to the garbage bin of history ....

No disagreement here...  I don't have the time and motivation to do this justice, and the foregone conclusion would of course be NO!  Yes, it is self-serving by definition -- it's company propaganda masquerading as a white paper.  I wish an independant someone would do the work, though.

And like Dolby B tapes, almost none of which are correctly trackable anymore, was the Holy Grail?  This is a pick-your-poison issue -- using DBX to record piano on cassette was a bad idea.  But the way that system worked allowed me to record things in the pre-digital days that were impossible any other way.  And those tapes still "track."* 

Whether obscure early-digital recording, industry standard compressor-limiter devices, or noise reduction, DB's achievements warrant that this be given at least a thorough look-over, in my opinion.
Which, perhaps, we have just done.



*Actually, its success was its downfall.  Using DBX in a quality system preserved so much extra source detail that the inevitable noise artifacts seemed particularly out-of-place. 

I remember watching an Imax movie (Chronos) where the audio, played back on a 35mm interlocked 6-track mag dubber, used DBX I rather than Dolby A, and in the theater noticed as the audio crept up from silence that hiss was pumping in...    No doubt you'd hear this on a DVD copy too!  But killer dynamic range?  Well, you know.

Yes... and even applying NR to the results doesn't end up giving you an acceptable result either - I tried.

To be fair, the system worked well enough if applied per track to multitrack systems where you only got one instrument per track - as long as that instrument wasn't a piano! I still preferred Ray Dolby's more advanced systems, even though they were harder to set up properly. At least he'd thought about the pumping problem...

Snake Oil alert!  If money is no object, you can have THIS (http://www.kipnis-studios.com/The_Kipnis_Studio_Standard/Cinema_Beta.html) for your home theater, with speaker response from 1Hz to 102.5Khz.  Sounds great to me -- especially if someone else is paying for it!

I read somewhere in the blurb that the response of the system includes DC. Since air is an elastic medium, I suspect that this might be a tad unnecessary...


Sapele isn't exactly ideal, or idealogically sound either, for speaker cabinets, and anyway, aren't you supposed to view your cinema in the dark? What's the point of the (might look exotic, but...) sapele, then? And a loudspeaker is a musical instrument now? How much skill does it require to learn to play it?

What?  Get real!  If you paid that much for your sound system and a video projector that would put out daylight brightness, you'd leave the lights on too so you can sit in constant admiration of your extraordinary achievement, and hope that everyone else is drooling too.  Hee hee...
 
Surely this is merely a measurement of an over-achieving HVAC system!

Ah, but you're going to need a good HVAC to keep the room cool because of all those toobs!

(If you're in the UK, that's valves...)

This just in with the April, 2008 issue of Mix Magazine, in the "Insider Audio" feature by Paul Lehrman:

The Emporer's New Sampling Rate (http://mixonline.com/recording/mixing/audio_emperors_new_sampling/)

Enjoy.  Smugly.

(Yes, this references the same study that was the subject of the article linked by Peter in the original post, many moons ago.)

Very nicely written. :-D

Hey Folks

I missed to track this item for a while. Being back now I found this to be a gross error as the newly announced audio system reproducing from DC to above 100 kHz is really fantastic. Yet, for my personal use it is kind of an overkill, as there is no one left that can profit from the upper 3 octaves since the death of my 2 Abessyn cats, and, on the other side, elephants communicating by infrasonics are very rare in my region. Nevertheless I sight a very particular application for this system I would like to promote here. I think that the inventors, intending the system for Audio reproduction are rather on the wrong steamboat as we use to say in german.

In the alpine regions, especially in Switzerland and Bavaria, singular weather conditions exist dominated by a particular warm wind coming from south. It is called "FOEN" (unfortunately i know no translation for it). This badly reputed situation causes a number of problems to part of the population such as headaches, dizziness, and forces many individuals to stay away from work. The mechanism causing this is recognized as strong variations in barometric pressure bouncing up and down within fractions of an hour. Beyond the incompetence of UBS top banking managers, this is a non negligible factor influencing national income in my country.

Similar to compensation of ambient noise already realised by some companies, a low pressure compensation would be a great relief for these those suffering from FOEN. No doubt that many might be willing to pay considerable amounts for such equipment. It might be a giant step for the health system as medical treatment has shown to be rather problematic so far. I am willing to act as general contractor due to my experience gained in refurbishing old houses and to hire competend people able to seal windows and doors hermetically to keep pressure. I could even abandon my audio activities for that and might also be willing to remove the apple from my scull. Comments are NOT welcome.
 

Just on a technical point, you can't transmit a DC excursion of a loudspeaker cone through air, because air is an elastic medium. Accordingly it tends not to stay where you put it for very long...

Yes, but with the .1 channel fed to the HVAC system in a sealed room, you'd have some potential to change the barometric pressure.

Well, since the 'V' in HVAC stands for 'Ventilation', I think that you might have a bit of a problem with this...   :lol: :-D :-D :roll:

Now Steve, don't throw cold air on his warm ideas!  :-D

If you search the inner recesses of your brain (or maybe someone elses brain through subliminal techniques),
you might recall that there IS a subwoofer available that DOES go down to DC, assuming it is mounted at
the room interface and the room is well sealed.  If I can find the link I will post it, but I think it was discussed
here before and works like a fan with blades that can change their tilt dynamically (or statically for that matter).

Now if you fill the room with bats who had to take the day off from work due to FOEN you could make full
use of the dramatic range of Bert's brilliant but somewhat atmospheric conception.

 :roll:

Doesn't really work though - all reasonable loudspeakers respond to DC - if you apply +ve DC to the positive-marked speaker terminal, the speaker moves forward and stays there until you remove it. Also, if you compare the volume of air moved forward by even a large speaker cone, and its effect on the elastic medium it's compressing (which averages out over the room), you'll find that it's well less than the volume of air you are breathing in and out. And so, your mere presence within the space invalidates any DC air compression effects by the loudspeaker completely!

.

Depends upon what the definition of reasonable is.

And unlesss I am entirely mistaken, I do not that Bert was suggesting anything reasonable!

Here is the subwoofer if which I spoke, which if mounted at the room interface (read through
the optional installation instructions), can raise the atmospheric pressure in a fairly well sealed
(perfect seal not needed) room.  Since "sound" is a pressure variation this would represent
response to DC.

It is not a sub-woofer I would choose to use, but I do have a problem with FOEN or cats or
bats either!

That they follow the Rice/Kellogg design... in other words, that they have a moving cone with a surround, at a bare minimum. And the voice coil has to be sealed from the rear of the cone as well - which not a lot of loudspeakers achieve particularly well.


A perfect seal is needed. If the air can return to its previous pressure state, then you have an AC response, albeit a very low frequency one. And at VLF, and even more so at DC, it's the pressure of the entire vessel that you have to consider, because we are not talking about a wave propagation effect at all.

I feel deeply depressed about what my post has caused. I intended it for being taken as a joke, but obviously my language skills in English are bad enough that my words had not been taken ironically. Assume a heavy cone loudspeaker with an effective diameter of 38 cm perfectly sealed experiencing an excursion of 1 cm. This will move approx. 1 litre  (0.001 m3) of air. A common small living room of 5 x 5 x 2.5 m has a volume of 62.5 m3 or 62500 litres. Thus for a change of 1 mbar in a medium barometric pressure of 1000 mbar there is need for 62.5 such speakers.  I hope that someone living in the imperial world can translate this into imperial units. Although the speakers might be placed on one wall only (it is the same number even in imperial units),  it must be noticed that the changes caused by FOEN usually are several mbars. Furthermore - based on 8 Ohm units connected in series, they would present a load of 500 Ohms which is rather uncommon for most power amplifiers. BTW: Fortunately I personally do not suffer by FOEN.

You could connect them in series/parallel configuration. That way you could get the impedance nearer to 8 ohms.

I am deeply depressed by the fact that after I go to all the trouble of re-sealing and re-caulking my room, then find enough amplifier power to run them all, the formidable challenge remains of where to find a half speaker that still works.

No, it's not quite that bad... If you use a series/parallel arrangement, you can present a much more sensible impedance to the amp - but you are still going to need a lot of power.

Think 2 x 8ohm speakers in series (16 ohms). Then get another two, also wired in series, in parallel with the first pair. You then have 8 ohm speakers still presenting 8 ohms to the load. The snag so far is that for the same displacement, you'll need more volts and more current - you don't get owt for nowt. But by extending the principle, you should be able to achieve a drivable load, I think - if you have a DC coupled amp that's up for it, of course... but like I said, you are going to need a lot of power.

I've just seen this result attributed to "bias"!  Someone has no concept of what testing means, I think.

Why artificially limit the universe of consideration?!  :roll:

A new type of Loudspeaker

(http://www.rotarywoofer.com/howitworks_files/trwanimate.gif)


Several of these and you can very substantially increase the DC pressure in a room!

(even with leaks in the room) 

There might be a small problem with a whirring noise on quiet passages (the MFG, of course, does not comment on this).

Here is the Freq response measurement:

(http://www.rotarywoofer.com/index_files/image003.jpg)

You will need another Woofer to take over above 40Hz, (i.e - a regular muli-element speaker system) and then some pretty good tweeters to handle the above 40Khz (for your 192Khz sample rate), otherwise you are set to go.

What you have effectively done is to cause a local DC offset in the air, and modulated it. What you can't do with this system is create a relative offset in the other direction without stopping the fan and rotating it the other way. The transmission linearity of this system is going to be simply appalling! At the heart of it, what you've created is a system where the air velocity is controlled, not the pressure.  In fact, if you ran this system in a relatively small enclosed space, you couldn't possibly cause any pressure shift across the entire space at all - think about it!

So you have whirring, and wind noise, and to top it all, the enormously bad smell of a right load of BS......  :-D

Wrong.

The blades when aligned to neutral generate no pressure,  when tilted "full clockwise" generate positive air pressure, when tilted "full counterclockwise" generate negative pressure.  (the fan does not need to reverse because the blades will tilt BOTH ways). And it IS pressure we are talking about here since the fan (Just like many rotary impeller devices including most water pumps) is acting as a pump.  The transfer function can be made to be fairly linear too.  This thing has been demonstrated at a number of Audio Conventions and t really works down to DC as advertised. If you baffle it sufficiently, it can even be made to be "fairly" quiet, but not quiet enough for me  :wink:, besides, I am not particularly interested in any sound under say 20Hz or so.

But then, ultimate practicality is NOT what is called for here because we are fighting the SCOURGE of FOEN!!

(but then some people do not seem to have any sense of humor whatsoever)

It's true that you could make it flow both ways, but you are wrong about this being a pressure change.There is NO WAY that this is a pressure change - because that fan is suspended freely in an elastic medium that is not contained. It is a velocity change - it has to be. It's the same as with weather changes - what appears to be an increase in pressure in one place has to be caused by a flow in another - usually a long way away. But it's not a static effect - it's only the rate of change passing though zero at the centre. And because air is an elastic medium, that change will not be constant, and will be represented throughout the atmosphere as a gradient - just as it is on any weather map. Any locally generated effect that isn't contained can only be achieved by molecular movement of the air, not compression of it.

Just because they told you it was a pressure change, it doesn't mean that they know what they are talking about - because if they said this, then they don't. I had a look at the website, and the only accurate statement about it was the last one in the explanation, which says


That's move it, not compress it. Whoever wrote their advertising copy should learn the difference. In fact, even a trivial analysis of this device reveals that it is effectively behaving as a very long-throw unsealed cone - which it would have to be to partially overcome the impedance mismatch at low frequencies. The impedance is the force divided by the velocity - there's not a lot they can do about the velocity in free air (it's pretty much fixed at about 330m/s), but by increasing the force they can reduce the effect of the mismatch, but not the mismatch itself. The idea that they can force more air over the blades by rotating them  simply compensates for the fact that it is not sealed. In all other respects it behaves as a cone driver would - albeit with more force. And if you really want to analyse how it works, you have to look at mass-spring systems, because that's exactly what this is.

Pumps inherently cause flow, not pressure. And this thing is a reversing pump, so it wouldn't cause a constant pressure anyway. To cause a build-up of pressure, you have to have containment. Any changes in pressure will only happen as compressions and subsequent rarefactions of the wavefront - they won't occur as the equivalent of DC at all.

If you put all those amplifiers in a sealed room, the pressure will rise... pV=nRT.

Put green-marker-loving audiophiles in the same room with Steve or a few others around here, and the pressure will rise too!

Maybe we could record that jam session in SACD format?

You could use one of these for that!

http://www.korg.com/gear/info.asp?A_PROD_NO=MR1&category_id=3

And anyone wanting to know how users have fared with the Korg machines could grind through the posts at http://taperssection.com/index.php/topic,80529.0.html and http://taperssection.com/index.php/topic,93710.0.html

And don't forget to give them enough big beans to eat !

I’m not sure exactly what I want to ask. It is something along the lines of “what is the meaning of this?” or “what is the significance of this?” This needs a little explanation. It started with a another forum’s thread on this thread’s topic. Someone posted saying he could consistently differentiate the 96kHz version vs the resampled 44.1kHz version of the triangle sample  (triangle-2_2496) from the PCXAB site
http://64.41.69.21/technical/sample_rates/index.htm

He resampled with Audition 3 using a quality setting of 999. Remembering the admonition about overly high quality settings, I ran a few tests. I only have CE2K, so I don’t know how results might be different with Audition 3.

Comparing resampling with quality 999 vs quality 250 by means of Mix Paste Inverted, I found a very large residual signal. A little more fiddling led me to make the same comparison between quality 250 and quality 251. Again there is a large data difference, much more than between a CD track and an mp3 encoded version of same, for instance.

I know that misalignment by even one sample makes for a major difference, but that does not seem to be the problem here. The samples line up, one to one, and the wave form appears to make all the same changes on corresponding samples. The sample values themselves are the difference.

Comparing the 250 and 251 versions (via Synchronize Cursor across Windows), the individual peaks as shown by the waveform drawing appear to be just about the same, even though the samples nearest those peaks have different values. Between the 250 and 999 versions, however, the individual peaks could differ by at least 2dB. Analyze/Statistics on the entire file (all 4.4 seconds of it, most being the fade out) gave almost identical peak and RMS values, in spite of the difference in individual peaks.

We know that making this kind of data comparison with mp3s and other perceptual encoded audio is meaningless. It isn’t the numerical dissimilarity of the data that is significant. That apparently is also true here, but perhaps based on something other than psycho acoustics? I don’t think the program is concerned with that kind of algorithm for resampling, is it?

Hmm... you have to bear in mind that the quality settings on SRCs don't relate to the conversion itself, but to the residual calculations in the pre/post filtering. And I have to say that if you are going to present a sample that has significant content above your resampled Nyquist frequency, then altering the filter characteristics could potentially make more detectable difference to those frequencies - simply because there is rather more content than there otherwise would be.

As for somebody saying that they can consistently tell the difference between two of those samples - well, that says rather more about their reproduction chain than anything else. Since it's well known that sound cards don't behave identically by any means at different sample rates, I'd say that this 'result' doesn't mean anything at all. You cannot possibly achieve a valid result with these tests unless you use a known-adequate replay mechanism. I'm sure that there are several options here, but I'm pretty sure that none of them involve 'domestic' soundcards. Even with some pretty good cards, you can see significant differences in the performance at different sample rates, simply by using RightMark.

That poster finally revealed he is using an Audigy 4 soundcard. That alone makes his claims suspicious, but I haven't been following the ins and outs of Creative soundcards. Do they still all resample everything to 48kHz? Creative seems to have kept high quality in the Emu line. Perhaps some of that quality has migrated to their consumer cards?

He claims that an Audition 3 resampling quality settings of 700 does not sound "as good as" one of 800 which does not sound as good as one of 999, and so on as the setting is lowered to 150. The absolute differences in sound are supposedly supported by ABX tests.

Anyway, this is not about the hardware, only about the program. Independent of the playback chain, the data is different. To change the presentation slightly, I used a lowpass FFT filter (Blackman Window Function, FFT Size 24,000) cutting off at 21kHz to hopefully eliminate any considerations about higher frequencies effecting lower frequencies. Saving that as a new file, I first resampled to 44.1kHz with a quality setting of 250, then with setting 251. There is still the same major data difference. This makes sense to me as I would not expect the program to know if there is anything at any particular frequency. It should do the same thing every time when resampling, regardless.

Comparing the setting 250 results from the two sources (original and low passed), I find no difference below somewhere in the vicinity of 21kHz, except for a faint haze of color (in Spectral View) at all frequencies during the first 50 samples. This seems to mean that (probably) every change of quality setting, no matter how small, makes for numerically large data differences. Perhaps someone with Audition 3 can report as to whether of not it also functions this way.

Since the resampled sample values are different with different quality settings, is there any way to say that one result is superior to another? If people can truly hear a difference, then each can judge which is "best," but is there a mathematical evaluation (or any other data oriented consideration) that makes any sense?

No such luck - they all resample. But some have a 96k pass-through. And this passes all of the processing through an entirely different route, so his claims would stand up to nothing as far as SRCs are concerned - but hearing a difference in results wouldn't be at all surprising. Different reason though...

Well, the soundcard information is interesting, but does anyone have any thoughts about the topic of my post: CoolEdit’s (and possibly Audition’s) largish resampling data differences from even small changes in the quality setting?