Poll: Can you hear sound over 20kHz? - Page 18

Originally Posted by stv014 

Originally Posted by kiteki 

However it seems like digital processing draws straight lines, such as in this example - http://www.head-fi.org/t/571259/hi-rez-another-myth-exploded/15#post_7748362

 

I would have thought reverb is typically a linear time invariant effect (i.e. convolution with an impulse response, which does not produce any aliasing), but even if time varying delays and/or slight distortion are added to increase the complexity, it is not clear why it would need to be more than 20 times slower than real time on currently available hardware to avoid aliasing; this sounds like snake oil.

In the follow up he says he paid $1.5k for the software, I think, not sure.

Originally Posted by bigshot 

I didn't bookmark it. But feel free to try it yourself. Take a CD and a good digital equalizer and set up a low pass filter at 10kHz. You'll find out exactly what that last octave contributes. Then set up a high pass filter at the same spot and hear what's up there. There really isn't much.

The important frequencies for sound quality are the ones in the middle, not the ones at the edges of our ability to hear.

sorry what, so you want me to do a test something like this?

I could test that easily in Foobar along with 44.1 -> 32kHz resampling, which is cutting off frequencies above 16kHz

I'm happy to use my Sony MDR-EX700 IEM which incidentally, apparently measures something like this

Anyone interested in high-rez discussion - http://www.head-fi.org/t/571259/hi-rez-another-myth-exploded/135#post_8375720

Originally Posted by bigshot 

Yep. It's always good when talking about numbers to know what those numbers *sound* like. The top octave in music (10kHz to 20kHz) is the least important octave when it comes to getting good sound. When you actually rack up a CD in an equalizer and listen to what is going on in the various frequency ranges, you immediately can see which ones are important to the reproduction of music.

Yes, that's why 96 kbps MP3 sounds 'acceptable', since they've focused on the 'most important' part of the frequency range and deleted the least necessary information.  However when talking about highest fidelity and realism then the information above 16kHz is still extremely important.

Quote:

Originally Posted by bigshot 

The information above 16kHz is the LEAST important part of the audible spectrum. I know that sounds counter intuitive, but it's true. Take your best sounding recording and listen to just the sound above 16kHz. I bet you a doughnut that you will find very little up there.
16kHz is NOT treble. It's nothing but faint upper harmonics of upper harmonics and the occasional noise from a cymbal crash. That's not what constitutes high fidelity realism. Balance throughout the core frequencies creates that.

I'm happy to read that, as my hearing currently goes no longer than 17 kHz 

I think I'd rather delete some bass like everything under 50Hz rather than delete everything above 16kHz, yes naturally something like 9~10kHz is more important.

I haven't played with an advanced equalizer but I can vividly hear a difference between 32kHz and 44.1kHz sample rates that's all.

On that note ♪ http://www.aes.org/e-lib/browse.cfm?elib=15398

If you want to figure out how to get great sound, it isn't enough to just know the numbers. You have to know exactly what those numbers represent. Theory is all well and good, but we don't hear theory. We hear sound.

16kHz-20kHz is about a third of an octave- three notes- at the very edge of human perception. 20Hz-50Hz is an octave and a half- 12 notes- extending well into the normal range of hearing. I know those numbers may sound like they're equivalent to you, but they aren't even close to being equivalent.

Play around with an equalizer a little bit and you'll see what I'm talking about.

Quote:

Originally Posted by stv014 

 

Quote:

Originally Posted by the wizard of oz 

Is there an app to generate sine tones on the DX100? I don't trust the 16bit 48kHz ALAC's I downloaded. With those on the DX100 I hear all the way up to 23kHz, but from 21.5kHz on it can't be the actual tone I'm hearing but rather hardware generated noise, since it doesn't follow the crescendo in treble, similarly to what the Mac Mini's headphone output was producing. So either these noises were created while recording these tones, either both the Mac Mini's headphone out and the DX100's headphone out are inadequate in handling those tones.

xnor's FLAC files that are available with 48 kHz and 96 kHz sample rate (the latter is recommended if the hardware supports it) are fine. If there is any problem playing those, it is specific to your system.

 

I realize I hadn't reported back on this yet. The FLAC files do play back better than the ALAC ones, funnily enough. Something with the decoding process maybe.

So with the 24bit/96kHz FLAC files on the DX100 paired with my W5000's I can hear up to 21.5kHz @ 230 volume setting, on Lo Gain; up to 22kHz on Mid Gain; and 22.5kHz on Hi Gain; 230 being the highest volume I listen to with my W5000's, staying mostly on Lo Gain, Mid being only for certain albums that can use some extra oomph. I never use the Hi Gain setting since not only does it get way too loud, but the sound signature turns to a very pronounced V shape.

My usual listening volume is 210~220, and at that volume my hearing extends barely over 20kHz, I can just about make out 20.5kHz.

Nothing to hear, but things to feel IMHO, that add to the sonic experience, especially for acoustic & voice. But I agree that if it hasn't been recorded in the first place, then it won't be heard nor felt.

& too loud on the DX100 is 232~255. I don't go there.