The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Launch of Pono Studio Headphones
Old 15th April 2014
  #3061
Lives for gear
Umm, did anyone else notice that the "party piece" filter in the Pono player is a digital filter? The next device mentioned is the analog amplifier. So the analog filter in use is the one provided by the chip manufacturer? I'm starting to lose confidence that this is even a high quality reproduction device quite apart from the 192 vs 44 kHz brouhaha.
Old 15th April 2014
  #3062
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by UnderTow View Post
It makes perfect sense if you just joined the space cadets academy. I think it is covered in the "How to push your best selling plugin to unsuspecting neophytes" course.

But then again, I look at posts here and after reading C's response to those posts I wonder if he ran the text of those posts through his Righteous plugin before reading. It is rumored that it removes all the glaringly obvious, and the extra bloom provided by Righteous seems to give him a very spacial interpretation of the content. A pitty the average level of his responses remains at a lowly -18 dB FS. If only he would limit his posts a bit more...

Oh gawd. Why am I posting at close to 5am? Obviously not to be taken seriously.

Alistair
Casually throwing in that Chris is only here to sell his plugs is way under the belt even for you, Alistair. Definitely not to be taken seriously indeed. You have managed to sink to a new low. Congrats.
Old 15th April 2014
  #3063
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Karloff70 View Post
Casually throwing in that Chris is only here to sell his plugs is way under the belt even for you, Alistair. Definitely not to be taken seriously indeed. You have managed to sink to a new low. Congrats.
I didn't say he was only here to sell his plugins but my misguided nocturnal attempt at humour does point to something more pertinent: Prejudice on the whole topic on his part. All the rest was just poking fun at his ridiculous marketing. And then came his last post... (<--- that one is fully deserved).

Alistair
Old 15th April 2014
  #3064
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by UnderTow View Post
I didn't say he was only here to sell his plugins but my misguided nocturnal attempt at humour does point to something more pertinent: Prejudice on the whole topic on his part. All the rest was just poking fun at his ridiculous marketing.

Alistair
If that was your attempt to climb back out of an embarrassing hole, you didn't quite make it to the surface. Should have gone for at least 192.
Old 15th April 2014
  #3065
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by jasman View Post
From the Pono site:
------------------------------------------------------------------------
I'm an audiophile. Explain what is so special about the PonoPlayer technology

The PonoPlayer was designed with a “no compromises” approach to sound quality. We partnered with the engineering team at Ayre (Ayre Acoustics) to include some of their world-class audio technology in our PonoPlayer. The Ayre team describes their contribution to the PonoPlayer design as follows:

• The digital filter used in the PonoPlayer has minimal phase, and no unnatural (digital sounding) pre-ringing. All sounds made (including music) always have reflections and/or echoes after the initial sound. There is no sound in nature that has any echo or reflection before the sound, which is what conventional linear-phase digital filters do. This is one reason that digital sound has a reputation for sounding "unnatural" and harsh.

• All circuitry is zero-feedback. Feedback can only correct an error after it has occurred, which means that it can never correct for all errors. By using proprietary ultra-linear circuitry with wide bandwidth and low output impedance, there is no need for unnatural sounding feedback.

• The DAC (Digital-to-Analog Converter) chip being used is widely recognized in the audio and engineering community as one of the best sounding DAC chips available today.

• The output buffer used to drive the headphones is fully discrete so that all individual parameters and circuit values and parts quality can be fully optimized for the absolute finest sound quality. The output impedance is very low so that the PonoPlayer delivers perfectly flat frequency response and wide volume range using virtually any set of headphone
---------------------------------------------------------------------------

If that 2nd bullet doesn't peg your red-flag-o-meter, than your meter must be broken. Although I suppose if you're looking for the answer to the title question, you sort of asked for it.
Now maybe I am tripping, but that description isn't a million miles away from what Chris has been describing.

Wouldn't it be hilarious if this thing actually does take those hi res files and through its filter difference to your normal converter is able to make the AUDIBLE range sound rather better through making it respond better in the time domain, needing the hi res to do it without eating up the highs in the process, as the filter decends way down compared to a brick wall. And then all the neysaying would go really quiet rather swiftly.
Old 15th April 2014
  #3066
Gear Head
 
bandpass's Avatar
 

Quote:
Originally Posted by Karloff70 View Post
Wouldn't it be hilarious if this thing actually does take those hi res files and through its filter difference to your normal converter is able to make the AUDIBLE range sound rather better
Better vs. transparent:

Quote:
Originally Posted by NwAvGuy
I believe most of the filter options some DACs offer are mostly a marketing gimmick. Some create audible levels of distortion and, as such, may indeed sound different in a blind test. Others don't audibly change the sound of playing music however much they might change the performance reproducing unrealistic waveforms.

You can insert an A/D and D/A into a signal path and not even tell both devices are there. That's exactly what Meyer & Moran did in their AES high resolution audio study. Over 500+ trials recording engineers, audiophiles, and others, couldn't tell when an extra 16/44 A/D and 16/44 D/A were added to the output of a SACD player. They proved transparent.

The M&M study speaks volumes as to the transparency of DACs. Not only could none of the listeners hear the rather boring conventional digital filter in the DAC, they also couldn't hear the DAC itself, the op amps, power supply, cables, and even the entire A/D process. Meyer & Moran not only called SACD into question, they provided strong evidence DACs can be audibly transparent despite all the erroneous audiophile claims they all have a sound of their own.
Old 15th April 2014
  #3067
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Karloff70 View Post
If that was your attempt to climb back out of an embarrassing hole, you didn't quite make it to the surface. Should have gone for at least 192.
We've gone on for what, 103 pages? I've been called worse.

I will do whatever I have to, to (a) further digital output formats until _I_ think they go far enough, which doesn't necessarily mean they have to go full DSD mind you, and (b) get in the way of the loudness war that's the other half or more of the problem.

Heck, I'd even go so far as to say that's the real problem, because I know you can get away with a LOT in 16/44.1 if you really try. I just think it's uninspiring how close that is to high-rate mp3, and it's unrepresentative of what digital audio really can do.

The claim about zero feedback is much like the 384K thing: it's not unusual to see high-ender analog designs marketing themselves in this way, but it's only one option of many worthwhile ways to get to a good analog stage. It's being touted because in real life that is what you do: you think of good-sounding lines of patter to try to communicate the legit technical work you've done to a curious but untechnical audience. It's like the 'stairsteps' analogy, which is representative of quantization but has no actual parallel in DAC output unless the DAC is designed broken. 'Zero feedback' essentially means 'look at this analog stage designed to sound good'. And it may (probably does) sound really good, but it doesn't hold that ONLY zero feedback designs sound good.

…and it'd be pretty weird if I didn't put my feelings about digital audio into practice with my own software. I literally make stuff to dither better, for instance, so I've actively worked with digital word length and the nuts and bolts of PCM audio, which also necessitates understanding what the DAC and reconstruction filter are going to do with it, and having strong opinions on what's best. That's what got me into the business in the first place, was total frustration with the state of digital audio and being stuck with it. You got it backwards: the whole plugin thing is an excuse to engage with this very subject and try to do what Pono is doing.

It took years to solve some of the problems and there's a lot I'm not talking about because it's too out there for the thread. It's not just 'persistent ABX confidence over 50% indicates fugitive observations that must be taken seriously on a statistical basis: only 50% confidence really suggests perceptions dropping below the threshold for good'. I've got a concept that due to reconstruction, isolated sample values are meaningless and they're better understood in pairs. Anyone throwing pathological data into my plugins (such as Dirac impulses) will sometimes get unusual results because I actively use these concepts. But I'm sticking to more mainstream ideas in this thread… and I am here to praise Pono, not bury it
Old 15th April 2014
  #3068
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Karloff70 View Post
Wouldn't it be hilarious if this thing actually does take those hi res files and through its filter difference to your normal converter is able to make the AUDIBLE range sound rather better through making it respond better in the time domain, needing the hi res to do it without eating up the highs in the process, as the filter decends way down compared to a brick wall.
The response is "better" in the time domain (compared to a brickwall filter) because the filter is so shallow, in other words, it is filtering less, but that just means there are more high-frequencies left (which is kind of the definition of a shallower filter). Hence the need for higher sample rates. But in this scenario the need for higher sample rates is to fix the fact that it is using a sub-optimal filter. The design fixes a problem in... the design.

I certainly hope the filter won't be filtering anything out of the audible band. Removing noise? Like reverb tails? Is that what the people are hearing? Music with the reverb (and highs) removed? That explains the "up in your face" comments. (I don't really think this is what is happening. If the DAC is significantly removing stuff in the audible band, it is a broken DAC IMO).

Alistair
Old 15th April 2014
  #3069
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by chrisj View Post
I've got a concept that due to reconstruction, isolated sample values are meaningless and they're better understood in pairs.
Now we are slowly getting somewhere. IMO many of the issues in digital audio have to do with processing looking at individual samples rather than the actual signal (not just two samples) being encoded in the data stream. You have alluded to Inter Sample Peaks, that is a flagrant example of what happens when dealing with sample values rather than the actual signal represented by the data stream. (Note for readers: It doesn't just affect absolute peaks that go over 0 dB FS when reconstructed: Many many reconstructed peaks are higher than the samples they are between but they mainly become a problem if they clip the digital anti-imaging filters during reconstruction or even analogue stages if not enough headroom is included in the design). The famous 3 dB difference in RMS meters (addressed by AES17) is another example...

This stuff, and all the loudness crushing that is a big cause of ISPs amongst other issues, is the biggest problem in audio quality at the moment IMO. I think this whole "hi-res" stuff detracts from these very real issues.

Also, the Pono crew are talking to consumers so their marketing BS can be somewhat explained if not excused but in discussion on what should be an audio engineering forum, we don't need marketing buzz words and ridiculous under water analogies (or format driven musical genres or other silliness).

Alistair
Old 15th April 2014
  #3070
Lives for gear
 
O.F.F.'s Avatar
 

Quote:
Originally Posted by Karloff70 View Post
Scientists are the constructors of temporary truth. For us all to hold onto for the time being. But I'll stop annoying you with my 'prattling on' now. As you were.
Scientists are the constructors of a detailed model of reality.
The only temporary part are fine details some of which still need to be discovered (although those are quite few) while others still need to be correctly connected to each other.
Old 15th April 2014
  #3071
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by O.F.F. View Post
Scientists are the constructors of a detailed model of a reality.
The only temporary part are fine details some of which still need to be discovered (although those are quite few) while others still need to be correctly connected to each other.
A useful vowel was missing I reckon.

On that second sentence we'll just have to disagree.
Old 15th April 2014
  #3072
Lives for gear
 
O.F.F.'s Avatar
 

Quote:
Originally Posted by Karloff70 View Post
A useful vowel was missing I reckon.

On that second sentence we'll just have to disagree.
I disagree with the first of your sentences above. ;-)

As for the second one: We may well have very different ideas of what constitutes details but I do not foresee a paradigm shift in knowledge/science/the model soon or at any time.
After all the model has to explain every observation made so far so there isn't a lot of wriggling room.
Old 15th April 2014
  #3073
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by O.F.F. View Post
I disagree with the first of your sentences above. ;-)

As for the second one: We may well have very different ideas of what constitutes details but I do not foresee a paradigm shift in knowledge/science/the model soon or at any time.
After all the model has to explain every observation made so far so there isn't a lot of wriggling room.
I am sure you do. That's cool. Your prerogative, isn't it.

Indeed, I think it is the quantifying aspect of 'detail' where we part. You are saying most of 'it' is grasped and only mere detail is to be finalised. I think what little detail we have nailed down is a spit in an infinite sea and find it strange for humans to really believe we know that much about 'everything'. As it all only exists in a place with no bookends holding it together anyway. In effect we don't even know where we are.

In the end it is all about belief, although you and likely others will jump me for saying so. You either chose to believe that what our science has nailed down by leaning one logic on the other is 'what we know' and that there isn't much more to it, or you believe that it is rather bigger and our science is a mere model to hold onto, as being a human requires this to a large degree as if it gets too abstract we can't function. And sure, I am not excluding myself. I obviously live by that model, too. It's ant stack functionality, and we're all ants. That doesn't necessarily mean though that one hence has to believe that just because the model is a great tool for functioning as a human/ant, that therefore it is all there is to it.

I am not one to tell you what you should believe, I just always find it a wow experience when people say with conviction things to the effect of 'it's all sussed out'.
Old 15th April 2014
  #3074
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by Karloff70 View Post
You either chose to believe that what our science has nailed down by leaning one logic on the other is 'what we know' and that there isn't much more to it, or you believe that it is rather bigger and our science is a mere model to hold onto, as being a human requires this to a large degree as if it gets too abstract we can't function.
But you're making a false dichotomy. No scientist believes that there isn't much more to it -- such a belief necessarily stops science. I assure you, no one is more aware of the fact that the models are just models than the model makers. It is the very premise of science; without such awareness, one risks becoming impassioned with his world-view. But that is anti-science; a scientist must dispassionately follow the data, wherever it may lead him.

Yes, it is a flawed premise from the outset: our window into the universe can never be objective, can never reveal what's out there as it is. We can only perceive the universe through our finite, perception-warping senses; and we can only interpret what we sense with these patchy, quirky brains that seem to have hundreds of millions of years worth of mixed programming. So we're doomed to an incomplete, barely coherent world-view no matter what we do.

But the astoundingly spectacular success of science has clearly shown that if we follow the scientific method, we end up with really good models. These models are almost certainly incomplete (and likely terribly skewed by our frame of reference), but they approach the truth. This is self-evident: if they didn't harbor some semblance of truth, then they wouldn't work!

So on the one hand we have flawed but partly-true models derived from a system that's specifically designed with the knowledge that we are flawed model makers. On the other hand we have the believers. They try to castrate science by pointing out its mistakes, claiming there is more than meets the eye. But that is all believers can offer: the promise of "more" based solely on the supposed failings of science. But that is an empty promise based on wishful thinking, at best.

If we are to do things in this world, then give me the discipline of science, the art of engineering, and the tools of mathematics to get it done.

And it'll get done.

Quote:
I am not one to tell you what you should believe, I just always find it a wow experience when people say with conviction things to the effect of 'it's all sussed out'.
Maybe it's ironic given all my gas-bagging above, but the essential processes in PCM encoding -- quantization and sampling -- have indeed been sussed out.
Old 15th April 2014
  #3075
Gear Guru
 
Sounds Great's Avatar
 

Quote:
Originally Posted by bogosort View Post

Maybe it's ironic given all my gas-bagging above, but the essential processes in PCM encoding -- quantization and sampling -- have indeed been sussed out.
But what hasn't been sussed out is whether there is anything being missed in this encoding/decoding.

The claim is that there have been enough tests done to prove that there isn't. Many people, myself included, find that claim unproven. Linking to a website of "some guy" who claims he will test you and write an article if you can refute that claim does not help the argument.
Old 15th April 2014
  #3076
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by bogosort View Post
If we are to do things in this world, then give me the discipline of science, the art of engineering, and the tools of mathematics to get it done.

And it'll get done.



Maybe it's ironic given all my gas-bagging above, but the essential processes in PCM encoding -- quantization and sampling -- have indeed been sussed out.
Which is why I said the model is a necessity for us to indeed, like you say, get stuff done. Ideally, while at the same time not allowing it to stop us from seeing possibilities the model can not account for yet.

It may have been sussed out. But if this is supposed to mean that just because 16/44.1 pcm can hold all the frequencies that a human ear can hear up and down to it is now perfect and the end station of audio, that makes me smile. That's akin to saying all speakers with the same frequency plot sound the same.

Within the audible range I for one am convinced whatever has been sussed out already, we can improve a LOT with the quality of audio. I mean, even tape manages to record certain aspects arguably better than digital (albeit with loads of other issues baked into it). Aspects we might not have a measurement for yet, but that we all can sense already.

I still want to hear this Pono device, as I don't know enough to make me certain the filter they say it has won't sound better than any normal ones I have heard already, the possibility of which makes me excited. I also think there are likely few here in this thread who are truly on a level of knowledge that can with certainty say this is not a possibility and completely rule it out, even if they think they are. Maybe someone like JJ. But I didn't see him say it. Hence meanwhile I'll take everyone else's neysaying as likely hot air and stay open to the possibility this Pono lark might shift the post a little somehow. Until proven different.
Old 15th April 2014
  #3077
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by Sounds Great View Post
But what hasn't been sussed out is whether there is anything being missed in this encoding/decoding.
If you believe that a time-varying voltage can represent the audio signal, then that time-varying voltage can be digitized to at least the level of precision of the time-varying voltage itself. This has been sussed out on both the theoretical and engineering level. What exactly do you believe is missing?
Old 15th April 2014
  #3078
Let's put it on a balancing scale.

One one side we have the vague and evidentiarily disconnected notions of Sounds Great et al that there must be 'something else going on' -- but who can't put any sort of finger on it or point to any credible evidence for it -- and, on the other side of the balance scale we have well over a century of increasingly intense study of sound and human perception of it that overwhelmingly points to the current model of how human perception of sound works.

Is it the final model? Will there be no refinements?

Of course not. That's not how science works, as folks keep pointing out.

Science builds on the evidence we have gathered -- and, with its self-correcting ethos is continually re-exploring and retesting prior data and understanding, building more solid and coherent understanding and squeegeeing out inevitable human error along the way.

Unfortunately many lay people only seem to see science through the lens of superficial and sensationalistic 'science reporting' -- typically the work of a young intern or cub reporter desperate to get OFF the science desk. (I started out uni with a small journalism scholarship.) And, sadly, it is those periodic 'everything-you-know-is-wrong' articles -- often based on a single preliminary study and sometimes not even that and almost always superficially and sloppily written -- that 'sell' the occasional science headline but which continually distort the nature and efficacy of scientific endeavor.


If the process of science was SO imperfect and SO fraught with defects of process and understanding -- how does all this crazy technology manage to work together and how do our scientifically trained engineers and system designers manage to use this supposedly deeply flawed science in order to create highly complex technical systems of all sorts that whose simplest to most complex operations and interactions are dependent on a minutely detailed and profoundly deep understanding of the practical and theoretical science that went into their design.

But that's technology. While we poets and artists have probed the 'eternal questions' of man's existence in the world -- how far along are we on that? -- the scientists bent their minds to the far more mundial and pragmatic task of figuring out how it all fits together -- examining and attempting to explain everything from why apples fall from trees to how the planets and galaxies move through outer space and atoms and electrons move through inner.

I dunno, Science seems to have made some small progress along the way, despite the profound flaws attributed to its process and driving ethos by some folks. Honestly, I think this dialog has gone into a place where we see two profoundly different and largely incompatible approaches to trying to understand our physical world. One attempts to apply observation, measurement, experimentation, and logic to that task. The other substitutes personal hunches, suppositions and wildly disconnected conjecture, even 'conspiracy theory,' for logic and evidence.
Old 15th April 2014
  #3079
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by theblue1 View Post

Is it the final model? Will there be no refinements?

Of course not.
You say that now, but there are several people here who reckon 16/44.1 pcm is the final stop for audio. Will never be improved as far as any human being able to tell the difference. I mean, really?
Old 15th April 2014
  #3080
Motown legend
 
Bob Olhsson's Avatar
 

The word "science" is getting thrown around a lot here.

What we know from real peer-reviewed research that has been agreed upon by real scientists is:

1. a 21-22 bit window is required to capture what a conductor hears while conducting a symphony orchestra.

2. a 50-60 kHz. sample rate is required for absolute transparency below 20kHz. using real world filters.

Anything less than this becomes a question of what can be gotten away with subjectively.

This also assumes that one is simply recording and playing back audio with no digital signal processing involved. The minute you apply any DSP, all bets are off as to what is required to achieve transparency. Again it becomes about what can be gotten away with subjectively as opposed to any kind of proven scientific fact.

People also want to believe that our hearing is linear and what one person hears is what everybody will hear. This couldn't be less true. Our brain can't comprehend all of the information presented by our ears and can only be focused on a tiny portion of the information at one time. Likewise any hearing damage destroys the ability to mask distortions in the frequency range involved. Unfortunately which band is more sensitive can be completely different from individual to individual. This means, among other things, that rock stars are far more likely to hear distortion than average listeners and recording engineers. The better our hearing is, the more we are actually flying blind!
Old 15th April 2014
  #3081
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by Karloff70 View Post
Which is why I said the model is a necessity for us to indeed, like you say, get stuff done. Ideally, while at the same time not allowing it to stop us from seeing possibilities the model can not account for yet.
Name the last scientist you heard say, "This is it people, we're done."

Quote:
It may have been sussed out. But if this is supposed to mean that just because 16/44.1 pcm can hold all the frequencies that a human ear can hear up and down to it is now perfect and the end station of audio, that makes me smile. That's akin to saying all speakers with the same frequency plot sound the same.
Weird analogy you chose, but I'll run with it. If two different speakers have the same frequency response (measured with the same parameters), then they will indeed sound the same (with those parameters). I take it you realize that a speaker's frequency response will change with amplitude, source impedance, and input type, and that no two speakers will have the same frequency response for all amplitudes, impedances, and input types. So not really sure what you're trying to say here.

Quote:
Within the audible range I for one am convinced whatever has been sussed out already, we can improve a LOT with the quality of audio. I mean, even tape manages to record certain aspects arguably better than digital (albeit with loads of other issues baked into it). Aspects we might not have a measurement for yet, but that we all can sense already.
What aspects does tape manage to record better than digital? If they are unmeasurable, then how do you know they exist? The actual differences between a specific tape and digital recording are easily measurable. On top of that, they're readily explainable: the physics of magnetic storage are well-understood.

What makes you think that some unmeasurable, inexplicable thing is missing?

Quote:
I still want to hear this Pono device, as I don't know enough to make me certain the filter they say it has won't sound better than any normal ones I have heard already, the possibility of which makes me excited. I also think there are likely few here in this thread who are truly on a level of knowledge that can with certainty say this is not a possibility and completely rule it out, even if they think they are. Maybe someone like JJ. But I didn't see him say it. Hence meanwhile I'll take everyone else's neysaying as likely hot air and stay open to the possibility this Pono lark might shift the post a little somehow.
Which is just wishful thinking. Nothing wrong with that, but it's not a credible argument.
Old 15th April 2014
  #3082
Gear Guru
 
Karloff70's Avatar
 

Thank you for the fresh wind, Bob!
Old 15th April 2014
  #3083
Airwindows
 
chrisj's Avatar
And there's inevitably subjectivity about the tradeoffs made in quantization and sampling. The fact that these things can be defined doesn't always give the final significance right away.

The idea of 22.050K and 16 bits was worked out in just this way, on the assumption that you could have a brickwall filter of whatever sort and it wouldn't matter so long as it was nearly perfect at deliniating the line between passband and stop band, with total rejection and a really sharp cutoff.

Now, we've been linked to an AES presentation talking about nearly 20 bits as the necessity, and there's widespread interest in things like apodizing filters on the grounds that pre-ring of the filter is particularly annoying. Me and UnderTow agree strongly on the problem of intersample overs (I think this, more than anything else, is what gives people the impression of 'digital glare in the highs' because analog output doesn't give you output swings significantly beyond 0db, and digital driven into stress and intersample overs does that continually while the rest of the audio is limited and clipped to provide a contrast to this; in practice, hyperloud and bright CD audio constantly puts out extreme high frequency peaks in excess of everything else, when working as designed. This is an unintended consequence of a design decision that became ubiquitous, in a format never designed to be abused in that way.

Choosing, as Pono has, such a different approach also has consequences. A naive moving-average calculation will result in a pretty lame EQ by accuracy standards, or consistency to a pre-intended curve. It may still be rolling off a bit at 20K, depending on how it goes. That will adversely affect sounds that are both bright and close-miked: definitely not 'reverb tails' because all supersonic frequencies die away long before the reverb even is a 'tail', you don't get supersonic highs out of even a second of transmission through open air. But another characteristic of that decision is optimal averaging of lower frequency data: in fact at 192K using four samples (that's not confirmed, though, it could be three) you're doing some nice averaging of frequencies 16K and up.

Given that people tend to hear 10K and up as 'super-highs' (and rightly so), getting this level of added accuracy on them could be an advantage, in the event that ADCs need help at capturing the full quality of the sound. You're basically helping the ADC do its job: if in some frequency range it's more fuzzy and not really delivering the accuracy, it becomes four times as accurate even if the converter itself is at its limit. In images we do this through taking the same digital picture repeatedly and then averaging to deepen color richness, or scanning repeatedly and doing the same thing. In audio, the examples are staggered very slightly but it's the same principle as taking four simultaneous samples from four converters and then averaging them.

If we were doing this at 16 bit capture, and the AES presentation claims 19 bits or so are needed to be properly accurate at audio frequencies, the technique works and gives us the desired result even while doing a 16 bit capture (4X). We're doing 24 bit. I'm guessing that's enough for anybody, as I struggle to get even a fugitive sense of things like dithering to 24 bit: I have to go by what's technically correct, as I wouldn't be able to hear 24 bit truncation… but the technique Pono's using gives us four times that amount of data, with a total and utter lack of time domain weirdness like ripple. If I'm not mistaken, it'll be rounding off the corners of squarewaves and delivering perfectly flat squarewave tops with no ripple at all, even the natural ripple that's a part of lowpassing. It'll effectively synthesize a 'smooth line' output through whatever the audio waveform must be, even if it means rounding off some of the corners at frequencies approaching 20K.

Traditional digital audio ignores ripple so completely that by design it produces extensive pre-ring as well as very severe supersonic ripple. The apodizing filters and alternate approaches put the ripple behind the attack transient, but it's still there, because it's the brick wall filtering and that is the shape of that audio condition when done properly. It's not the sound of digital, it's the sound of really high filter slope and it means 'full frequencies up to 22K, and then none'.

The 'broken' Pono moving average design does something completely different. Its frequency behavior is odd, even pathological, but the thing totally lacks ripple, and if it turns out ripple is the key factor to aversion to digital highs… then we've learned something interesting that the initial creators of sampling didn't know at the time.
Old 15th April 2014
  #3084
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by bogosort View Post
Name the last scientist you heard say, "This is it people, we're done."



Weird analogy you chose, but I'll run with it. If two different speakers have the same frequency response (measured with the same parameters), then they will indeed sound the same (with those parameters). I take it you realize that a speaker's frequency response will change with amplitude, source impedance, and input type, and that no two speakers will have the same frequency response for all amplitudes, impedances, and input types. So not really sure what you're trying to say here.



What aspects does tape manage to record better than digital? If they are unmeasurable, then how do you know they exist? The actual differences between a specific tape and digital recording are easily measurable. On top of that, they're readily explainable: the physics of magnetic storage are well-understood.

What makes you think that some unmeasurable, inexplicable thing is missing?



Which is just wishful thinking. Nothing wrong with that, but it's not a credible argument.
No, they wouldn't say that, they are never done. But that doesn't mean they wouldn't possibly miss the wood for the trees, because of orienteering based on the already 'known'. Usually when the border moves forward it will take someone leaping out of the present status quo headstate to even see the next possibility.

I am trying to say that the things we can measure about our audio today are missing measurements for parameters we have not named yet. There are no measurements of a loudspeakers today that can tell you what it actually sounds like by reading them. Only vague clues. The only thing measuring it really is your ear/brain when you hear it.

Tape captures something about the content of the audio that isn't to do with any measurements we have. Call me fool for saying so, disagree, fine. And the fact you say that the physics of magnetical tape storage are well understood is about as valid in ascertaining these things as saying "I know the chemical composition of the paint used on the Mona Lisa" would be in order to determine why it feels a certain way to look at and has a certain value of expression coming off it. Tape transports a feeling in a different way. Differently flawed, but in some ways more efficiently than today's digital audio IMO. How do I know this? Well, I sense it, sorry this is where you can call me a fool again and say I must be imagining things. That's ok.


That last one wasn't meant to be 'an argument'.
Old 15th April 2014
  #3085
Quote:
Originally Posted by Karloff70 View Post
You say that now, but there are several people here who reckon 16/44.1 pcm is the final stop for audio. Will never be improved as far as any human being able to tell the difference. I mean, really?
I believe if you take a close look at those who state their claims carefully, I think you will find that the claim is that 16/44.1 is for most purposes entirely adequate as a delivery format, in that, with a good implementation, it can deliver 20 kHz of bandwidth with a ~90 dB signal-to-noise ratio and what perceptual scientists would suggest should be inaudible distortion. That gives that format (if not necessarily individual devices -- a key point that some seem intent on blurring or avoiding) native capability that surpasses the playback capabilities of the overwhelming majority of repro systems out there. It provides a dynamic range that far surpasses the program content requirements of almost all the music being made today -- and it's worth noting that any recorded material that pushes the lower limits of that SNR with ultra low content will likely need have the volume adjusted during everyday playback.*

So, for the overwhelming plurality of musics that people listen to, on the systems they listen to it, the CD format is far more than adequate.



*I'm a classical music fan (seen over 160 symphonic concerts, etc) and I like my music dynamic, without question. But even in the concert hall, one can go from straining to hear nuance in a single instrumental solo (and in that world, solo means just that) to having one's ears boxed by 100 musicians playing as loud as they can (and believe me, that man-size bass drum can put out some SPL, believe you me), it can be extremely demanding, problematic even. While it's great to have the technology that can contain those dynamic extremes -- it can also be vexing listening at home. (Hey, just NOT having the freaking audience along for the ride goes a long way to 'improving' the dynamic experience. When someone coughing 20 rows away is louder than the soloist, it's an issue. And you know, the classical audience, they aren't getting any younger and healthier. God love 'em.)
Old 15th April 2014
  #3086
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by theblue1 View Post
Unfortunately many lay people only seem to see science through the lens of superficial and sensationalistic 'science reporting' -- typically the work of a young intern or cub reporter desperate to get OFF the science desk. (I started out uni with a small journalism scholarship.) And, sadly, it is those periodic 'everything-you-know-is-wrong' articles -- often based on a single preliminary study and sometimes not even that and almost always superficially and sloppily written -- that 'sell' the occasional science headline but which continually distort the nature and efficacy of scientific endeavor.
Unfortunately this is quite true. There is a giant disconnect between scientists and the general public -- they don't even speak the same language -- and the only bridge between the two comes from science-beat reporters who almost invariably are ill-equipped to understand the science. Compounding the problem is that working scientists almost invariably are ill-equipped to explain the science to a lay-person in an email or short phone conversation.

Ed Yong, who does good science-reporting, has given advice to both sides about the problem.

Quote:
While we poets and artists have probed the 'eternal questions' of man's existence in the world -- how far along are we on that?
Zing! But I always tear up in the scene from Contact where Jodie Foster is traveling through space and she can only stammer, They should have sent a poet.

I like to think that at the core of every great scientist, mathematician, and engineer is a poet searching for beauty.
Old 15th April 2014
  #3087
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by theblue1 View Post
I believe if you take a close look at those who state their claims carefully, I think you will find that the claim is that 16/44.1 is for most purposes entirely adequate as a delivery format, in that, with a good implementation, it can deliver 20 kHz of bandwidth with a ~90 dB signal-to-noise ratio and what perceptual scientists would suggest should be inaudible distortion. That gives that format (if not necessarily individual devices -- a key point that some seem intent on blurring or avoiding) native capability that surpasses the playback capabilities of the overwhelming majority of repro systems out there. It provides a dynamic range that far surpasses the program content requirements of almost all the music being made today -- and it's worth noting that any recorded material that pushes the lower limits of that SNR with ultra low content will likely need have the volume adjusted during everyday playback.*

So, for the overwhelming plurality of musics that people listen to, on the systems they listen to it, the CD format is far more than adequate.


Actually, no, in the other thread the guys are saying with totality that 16/44.1 pcm is the end. Your human perception is not equipped to appreciate anything any better. As a delivery format, yes, but no other quantifying that statement whatsoever.
Old 15th April 2014
  #3088
Quote:
Originally Posted by Bob Olhsson View Post
The word "science" is getting thrown around a lot here.

What we know from real peer-reviewed research that has been agreed upon by real scientists is:

1. a 21-22 bit window is required to capture what a conductor hears while conducting a symphony orchestra.

[...]
Who here has a repro system with an end to end signal-to-noise ratio of 126 dB?

Step right up.
Old 15th April 2014
  #3089
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Karloff70 View Post
I still want to hear this Pono device, as I don't know enough to make me certain the filter they say it has won't sound better than any normal ones I have heard already, the possibility of which makes me excited. I also think there are likely few here in this thread who are truly on a level of knowledge that can with certainty say this is not a possibility and completely rule it out, even if they think they are. Maybe someone like JJ. But I didn't see him say it. Hence meanwhile I'll take everyone else's neysaying as likely hot air and stay open to the possibility this Pono lark might shift the post a little somehow. Until proven different.
Pay better attention!

Quote:
Originally Posted by j_j View Post
BUT you don't use a "moving average" filter. No, not ever. You use a filter with a carefully designed frequency response that does not cause any in-band dips. A simple sum of 4 samples is far, far from the optimum filter, either perceptually or analytically.

There is a slide deck at PowerPoint Presentations from recent (or not so recent) meetings. that shows some of the process of oversampling and noise shaping. It's old, but the mathematics doesn't change, unsurprisingly. Look for the "convertor tutorial".
Alistair
Old 15th April 2014
  #3090
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by chrisj View Post
Traditional digital audio ignores ripple so completely that by design it produces extensive pre-ring as well as very severe supersonic ripple. The apodizing filters and alternate approaches put the ripple behind the attack transient, but it's still there, because it's the brick wall filtering and that is the shape of that audio condition when done properly. It's not the sound of digital, it's the sound of really high filter slope and it means 'full frequencies up to 22K, and then none'.

The 'broken' Pono moving average design does something completely different. Its frequency behavior is odd, even pathological, but the thing totally lacks ripple, and if it turns out ripple is the key factor to aversion to digital highs… then we've learned something interesting that the initial creators of sampling didn't know at the time.
No competent filter designer ignores ripple, though our ears are demonstrably much less sensitive to time-domain distortions than frequency-domain distortions. The great benefit of oversampling the local ADC/DAC rates is that the heavy-lifting can be done digitally, where the filter characteristics can more easily be specified.

I'm not sure why you complain about ringing in digital filters but ignore the ringing in analog filters. Every analog filter greater than first-order will ring, yet we seem to happily live with our analog EQs. If an apodizing digital filter rings no more than an analog EQ filter (or even the natural RLC filter inherent in all analog stages), then why is it suddenly nefarious in the digital domain but harmless in the analog?

It does not seem likely that the key to audio nirvana lies in reduced filter ringing.
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump