PONO

[stringtapper]

Did you read that whole paper or just the abstract?

From the article:

Our previous studies showed that ultrasound at 27, 30 and 33kHz strongly masked air-conducted sounds ranging from 9 to 18kHz.

Masking sounds in the audible range sounds like a bad thing, no?

The stimuli were presented to the mastoid by a devised ceramic vibrator used in our previous study.

Sounds pretty different from a "normal listening experience."

The standard of hearing levels for BCU has not yet been estab- lished, and so we operationally defined 0 dB normal hearing level (nHL) from the average thresholds of the control group for each frequency measured once a week for total 5 times.

So they never really say what dB the BCU portions of the test were done at, only that they were applied in increments of 1dB. That is some key information needed in order to link this study back to high sample rates because if most of the ultrasonic content of a 192kHz recording (about 20kHz–96kHz) is below the dB levels of their test threshold then applying this study to the subject of high sample rates at least goes beyond the bounds of practicability if not the bounds of perceptibility. Having recently looked at the sonograms of recordings I've made at just 96kHz I would be surprised if anything captured above 30kHz by any recording system would ever meet the dB threshold of the BCU portion of those tests.

[fishmonkey]

yes, i've read the article, and some associated ones.

i'm not defending mega-hires playback, i was simply responding to the blanket claim that ultrasonic frequencies are only detectable by our sensory systems as distortion products.

also, you are misconstruing what 'masking' refers to in this experimental context. the presence of masking simply implies that a particular sound is definitely detectable. whether it's a good or bad thing is another matter. if ultrasonic frequencies in natural sounds have a measurable impact on the 'audible' frequencies, then one possible interpretation is that those sounds are therefore useful contributors. that may or may not be the case, but i don't believe we currently have a definitive answer.

masking of course also occurs a lot in the audible spectrum, and is a feature of the way our ears work. masking experiments are the classic way of determining auditory tuning curves, for example.

[eyeknow]

The most hilarious part I've found in researching all this is finding out the highest frequency response limits of the best mics and speakers.

The highest mic I can find goes up to 30kHz. The highest for speakers that I could find was 40kHz.

That means even 96kHz recordings are going to contain around 8kHz that the speakers can't reproduce and 18kHz that the mic never picked up in the first place!

This is not entirely true. check these out:

these are "modeled" after hardware:




Please don't ask me to explain why these "models" have frequencies that people are not supposed to hear, but I hear them. So all I'm trying to say is just because the FR is "allegedly" out of our hearing, it doesn't mean nothing is significant about having it.

That said, I don't think the pono thing is worth a hill of soggy baked beans.

[stringtapper]

^ Not sure what those plugins have to do with microphones and speakers nor am I sure how they make my quote above untrue. Can you explain?

What I am not unsure of is whether you are hearing ultrasonic frequencies. You aren't. But believe what you like.

[Forge.]

Please don't ask me to explain why these "models" have frequencies that people are not supposed to hear, but I hear them. So all I'm trying to say is just because the FR is "allegedly" out of our hearing, it doesn't mean nothing is significant about having it.

That said, I don't think the pono thing is worth a hill of soggy baked beans.

that's probably more likely to mean the plugin is producing some kind of distortion in the ranges you can hear

at the end of the day, to me this is like arguing about whether the "Q" is a few pixels too wide in the Helvetica bold font.

It just doesn't matter.

I've been recording everything the last 6 months in 88/32 and I helped out in a session the other day that was in 48 and the raw audio sounded pretty much exactly as good as the recordings I've been doing.

Dan lavry made the argument for 60kHz. I think I decided to use 88 because Partly because it was the next best in that it provides more "extra stuff that we can't hear anyway", and Because Native Instruments "Monark" displayed an angry message telling me that it hadn't been designed for such lowbrow frequencies.

I think Some plugins inside the sequencer though very definitely do sound different at higher sampling rates. Especially reverbs and delays, But this is not relevant to the Mic/Speakers aspect

[Forge.]

in other words ... you won't get anything better out of the raw audio *that matters*

but in the case of the reverb and delay and other plugins you use on it, it might be worth it.

So this either means, recording at lower sample rates and upsampling the audio, or just working in the higher one in the first place.

Recording at high rates takes twice as much HD space, and thus takes twice as long to backup after sessions

I'm beginning to think the upsampling option might be the better one.

[eyeknow]

^ Not sure what those plugins have to do with microphones and speakers nor am I sure how they make my quote above untrue. Can you explain?

What I am not unsure of is whether you are hearing ultrasonic frequencies. You aren't. But believe what you like.

I'm not trying to start a thing. Was simply pointing out that 25 to 27k is supposedly out of our hearing range but you can clearly hear a significant difference when adjusting them (and the hardware which would work with mics)

[stringtapper]

^ Not sure what those plugins have to do with microphones and speakers nor am I sure how they make my quote above untrue. Can you explain?

What I am not unsure of is whether you are hearing ultrasonic frequencies. You aren't. But believe what you like.

I'm not trying to start a thing. Was simply pointing out that 25 to 27k is supposedly out of our hearing range but you can clearly hear a significant difference when adjusting them (and the hardware which would work with mics)

Nothing's getting "started." First of all he limits of hearing are well understood and have been tested many times over. If you're hearing a difference it's not because you're actually hearing ultrasonic frequencies, regardless of what the numbers on those dials say. Secondly, it's software. There could be anything going on behind that interface.

Again none of this addresses the fact that even most professional microphones and speakers can't capture and reproduce frequencies above the range of around 30kHz–40kHz, and that's being generous. Even those Opals Tarekith rock are only rated up to 22kHz.

There's also a simple test to see if you're hearing something around those 27kHz dial markers: throw a high pass filter on the output and steadily move the cutoff up to 27kHz.

[eyeknow]

I hear you, but that's not what I mean.

All I'm saying is just because the frequency response is beyond our hear capability doesn't mean there is simply no advantage. Sorry if I'm not getting my point across.

Keep in mind, I laughed when I saw "soundcity" and dude was saying the got the algorithms all wrong. I KNEW he was up to something with that statement. Turns out I was right.

I don't commute on a train or train at the gym (which wouldn't matter anyways since they are SO FUCKING LOUD ) but I wouldn't have the need for one of these doohickeys. I'm all for better sound but I'm terribly suspicious of this product.