Is there any importance above 15.000 Hz?

[simon.a.billington]

Yeah, I wouldn't be charging everything with a top end boost, only things that need it. Cymbals, vocals, even a bit or reverb proving its not too dark.

There's a lot I would sculpt away from the top too, that's why we have low pass filters, just be gentile with them.

I'm not sure, however, whether a mic that tops at 15k would be good enough for vocals. Maybe, it does depend on the type of vocals and your mix decisions. It may also not be as effective in the aforementioned drum cymbals. Brass instruments and screechy violins generally have too much top end so a mic like this may work well on them as well.

[Andy Gone]

I believe it depends a lot on the program material. Most often there are no 'useful' fundamental tones there, but there might be important harmonics.

One interesting exception that comes in mind is very high-pitched percussion on Exciter album by Depeche Mode. It is really loud, sometimes louder than vocals, but does not come in the vocals way.

[llze]

by using such frequency ranges like 24Khz on a speaker or even 48Khz on some headphones, they make sure that the energy in the higher frequency range is reflected in a "working range" and not fisseling around the limits. Basically to have an equal responce. It's just the electronics that work better in green than in red maybe

[simon.a.billington]

Yeah the big misnomer about us not hearing above 20k or so is that it just straight forward assumes we can't hear anything past that point.

I mean, that's of course a logical way to think, and perhaps maybe not worded as accurately as I would like but...

While it's true we don't hear 20+KHz frequencies directly, we hear the byproducts of those frequencies interacting with other frequencies and those byproducts interacting with other frequencies.

We are hearing the sum and the differences of those main frequencies, those are the byproducts. Though, while we still can't hear the byproduct of say 20kHz + 25kHz, we can still 25kHz -20KHz, which is 5kHz.

So yeah there is point well above 20kHz that still contains some useful information, just how high that useful info goes - I'm not so sure.

I do know that I can hear a difference, with higher sampling rates producing more lively, open sounding results, with the lower sampling standards of 44.1 or 48kHz it seems to sound more stiffled, a bit less energetic.

I do think 96kHz sampling rate may be a bit excessive, but it's not like anyone is offering us 60-70kHz sampling solution.

Plus I'm also with the impression that the ear'rs frequency response needs to be retested. We've all known about 20Hz-20KHz being the ear's range for decades now, but how were the ears tested. Was it by directly using sine waves or did they have people compare high frequency content bursts against each other...?!?!

We've also learned so much more about sound since then. I'm think it's about time for a more accurate retest... Unless there's something I'm not aware of there.

[lfm]

I do know that I can hear a difference, with higher sampling rates producing more lively, open sounding results, with the lower sampling standards of 44.1 or 48kHz it seems to sound more stiffled, a bit less energetic.

Yes, I've run Metaplugin with overampling active now for some time, and about to do that with all synths - without having full projects in 96k. It depends on material a bit, but strings as one is quite noticable difference how those complex waveforms are translated to each pitch with oversampling or not. Guessing it's all about aliasing that you drop as you oversample. More clean and smooth sounding.

But having real expensive and large sample libraries where each pitch is sampled and not resampled - difference is less by oversampling. Cheaper libraries and synths benefit more is my assumption.

About ranges for ear, I got the impression looking at Fletcher-Munson curves that it's all about energy you output. At a certain spl you need to put so and so on 10,15k etc to follow ears sensitivity, on another spl you just need some less energy.

It's not quite as we build amps and say that range is 20k because level is down 3dB compared to 10k or 1k.

So not sure how they draw the line, if ear needs 100 dB more than half frequency or something - to say where ear has a limit. It would be a wall more or less in FM-curves. No straight lines anywhere really.