content over 20khz?

[ZAKO]

Does any body remember the old Farting Contest Olympic record made in the early 1930s ???? Quite funny..

[jacksonbart]

I heard it, but thought the recording was a little muddy.

[ZAKO]

Muddy through the K402 horn of course.

[Mr. RF62]

Guess one could ass_ume it is muddy if there is substance coming out as well.

[ZAKO]

Is that called Laquidity ???

[ClaudeJ1]

Anything above what you can hear is simply window dressing IMHO.

Otherwise known as pure marketing BS (oxymoron or redundancy?) to sell $50K plus speakers.

[ClaudeJ1]

For what it's worth, I came across an article on the B&W 803Ds which said:



"Granted, human hearing tops out at around 20 kHz, but there is substantial evidence to show that the range we do hear is powerfully affected by what's going on in the range we don't hear."

What evidence? Let's see and hear the "evidence." I can't hear much past 12 Kz. so does 20 Khz. help me hear better at 12K. I think NOT.

[colterphoto1]

Could that be him appearing at 1:02 in this video? It looks to be a mid-range horn.

The Purple Helmets:

Thanks for that Pat, I needed a good laugh this morning. Yes it does look like Monty Python gone terribly wrong. Hilarious

[colterphoto1]

For what it's worth, I came across an article on the B&W 803Ds which said:



"Granted, human hearing tops out at around 20 kHz, but there is substantial evidence to show that the range we do hear is powerfully affected by what's going on in the range we don't hear."

What evidence? Let's see and hear the "evidence." I can't hear much past 12 Kz. so does 20 Khz. help me hear better at 12K. I think NOT.

In a few short years you'll be able to delete the tweeters from your system saving precious bucks and time aligned height! []

[Islander]

For what it's worth, I came across an article on the B&W 803Ds which said:

"Granted, human hearing tops out at around 20 kHz, but there is substantial evidence to show that the range we do hear is powerfully affected by what's going on in the range we don't hear."

I've read often that adding a sub to a system has an effect on the midrange as well as the bass, sometimes adding barely audible cues (or clues?) as to the size and shape of the performance space that add to the realism of the listening experience, even though they're barely perceived by the listener.

Could it be that something similar happens from the top frequencies down into the audible range?

[Islander]

Thanks for that Pat, I needed a good laugh this morning. Yes it does look like Monty Python gone terribly wrong. Hilarious

Glad to hear you enjoyed that, Michael! The Purple Helmets perform at bike and car racing events across England and are very popular.

[DrWho]

However for electronics (amps) there may be a valid reason for going 100K or better because of the feedback from output back to input. It seems reasonable to me.

The amplifying stages need a high gain bandwidth product, but the final system doesn't need to go out to light...

Amps that go out to 100kHz are usually going to have high gain bandwidth stages in them though, so it's more of a correlation than a causation.

[Islander]

Does any body remember the old Farting Contest Olympic record made in the early 1930s ???? Quite funny..

I remember my dad referring to a record called The Crepitation Contest. Luckily for us, practically everything can be found on Google. It's not the Olympics, it's the British Empire Championship being contested.

Here's a clip: http://oldtimeradio-in-tx.homedns.org/otr/crepitation%20contest/

More info here: http://plancksconstant.org/blog1/2007/03/the_international_cr.html

You can buy Parts One and Two from Amazon for only 89 cents each, but it's unclear in what format.

[DrWho]

SACDs and LPs have lots of high frequency information

SACD is good to 50kHz I believe, but LPs? I wouldn't expect anything but noise. Even tape (which usually comes before the LP) has a natural HF rolloff to it.

The problem with all supertweeter demos I've seen (which certainly isn't all of them) is that xover points and whatnot are never perfectly steep. Even crossing as high as 40kHz to a super tweeter can have an audible effect at frequencies as low as 2.5kHz with a single order filter (which seems to be the xover of choice). I can't help but wonder if the perceived difference isn't caused by the resonant frequeny of the super tweeter, or whatever minor levels of equalization that result from unbalanced drive levels and non-linear impedances, etc...

[Islander]

SACD is good to 50kHz I believe, but LPs? I wouldn't expect anything but noise.

Didn't John Lennon once say that the Beatles put some 15KHz content on one LP just to see if some dog owners noticed any reaction from their pets? That would imply that most LP content is well under 15KHz, or at least it was in the 1960s.

[ZAKO]

Like I said before if the mic does,nt pick it up,, The super tweeter will not reproduce it... I dont think any body on this forum could pick out the resonent frequency of a super tweeter

[Don Richard]

For what it's worth, I came across an article on the B&W 803Ds which said:

"Granted, human hearing tops out at around 20 kHz, but there is substantial evidence to show that the range we do hear is powerfully affected by what's going on in the range we don't hear."

So if I leave the B&Ws at the dealer's, 60 miles away where I certainly will not hear anything from them, will these mysterious inaudible effects still be manifest? Seems like more inaudible would cause a more "powerful effect", eh?

[DrWho]

I dont think any body on this forum could pick out the resonent frequency of a super tweeter

Except for the ones where it falls into the audible band.

I think most impressions of super tweeters have more to do with the fact that xovers are never perfectly steep, so there is always some output below the xover frequency. It becomes an even stronger argument in light of the fact that most people (that I've seen) are usually going with first order highpass filters.....so you're going to get all sorts of content into the audio band unless you xover really high.

Let's see, a 40kHz xover only puts you 6dB down at 20kHz (and that's assuming a flat impedance response, which is also never the case). If you're crossing lower to pickup where most tweeters roll off (usually between 16kHz and 23kHz or so), then you're guaranteed to have output in the audible band......and with the simple xovers there is no accurate tweaking of the overall amplitude. Even as little as 0.1dB over a wide bandwidth can have very noticeable impacts on the sound.....much more than reproducing higher frequencies (that probably don't exist on the recording anyway).

I think it's another case of correlation vs causation.

[garyrc]

colterphoto1: " What evidence? Let's see and hear the "evidence." "

Yes, indeed, and how was the evidence collected?

As Dr. Who may have suggested, we may see a lot of "correlation but not causation" error in our conclusions. Even when restricting ourselves to the same speakers (all with the same supertweeters capable of going far above 20K), and using filters to cut off response just above 20K, in one condition, and with the kind of procedures discussed below, we still wouldn't know if something about the damn filters was causing the difference. Of course, we could build in "type of filter" as yet another analyzable independent variable in our multivariate analysis, but what if they all have artifacts?

To try to answer the 20K question, and many other questions in audio, I would like to see double blind experiments set up, even though some of the "Golden Ears" are allergic to research of this type. In the case of the "above 20K" question, we might ask three or more questions:

• Does content (if any) above 20K Hz make a difference to groups of participants who are randomly assigned to "cut off above 20K" v.s. "no cut off" listening groups? This could be a "between by within" experiment with each individual hearing a collection of music twice, but half (randomly determined) hearing the cut-off version first, with the others hearing the full range version first. They would need to use rating scales already shown to be sensitive to audio quality differences. They could rate on several response measures, such as preference, realism, musicality, what-have-you.
• Does content (if any) above 20K Hz make a difference to groups of "special people" -- musicians, audiophiles, people who earn their living playing or recording music -- who are randomly assigned to "cut off above 20K" v.s. "no cut off" listening groups? The experimental design would be the same as above.
• Does content (if any) above 20K Hz make a difference to a given individual -- YOU, for instance, when exposed to a series of "cut off above 20K" v.s. "no cut off" blind trials? Similar design.

The often heard objections to A-B testing (e,g,, the brain is functioning differently when making immediate comparisons -- judging is different than relaxed but attentive listening) would still be a problem, but might be partly overcome by building in the variable of "type of auditioning," with levels of "immediate judgments " vs "end of prolonged listening session judgments." This might be done in the participants' homes over several days, but would be hard to set up, and full of problems, but in the "prolonged listening" condition the participant, blinded to whether the filter is on, could listen as usual for several sessions, then fill out a dated card on leaving the music room. The equipment would have to automatically record whether the filter was in "pass through" or "cut off" on that particular day.

[ZAKO]

Benz,,, You still hav,nt addressed how a 40k signal harmonic effects a studio mic,,whos bandwith craps out a 20k... It still doesnt get on the master cut.