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 4th May 2014
  #4621
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by chrisj View Post
You know, I keep beating on the difference between 'telling the difference' and 'telling the difference every time' and it'd be nice to see the abstraction more closely approach people's normal experience with an audio device.
I understand what you are saying I think, but I prefer to deal with the known rather than the unknown or unprovable.

As you know poeple's listening experience is filled with emotion and expectaiton bias and peer pressure and a whole load of other stuff. How do you suppose we "more closely" approach all that stuff that is not measurable and not knowable?

I'm at a loss to understand exactly what one should do, in your opinion.

I think you know what you know. If you believe there is something there that cannot yet be measured, that seems reasonable, but don't expect others to go with you on faith, right? I'm pretty pragmatic, I've done a fair amount of testing and I can't hear it... so to me it just seems like expectation bias or other audiophile mumbo jumbo. Once you or others have found this X factor and indentified it/measured it, then I will go along with it.

When I was doing radio production there was one particular studio in the facility where I had problems getting things to sound right. I couldn't put my finger on it. Finally I had the engineer go through the studio and he found swapped polarity on one of the tape deck channels. So.. I believe there are sometimes things that we can hear but can't identify like this, but eventually we need to be able to get there and identify it. If there is something, there is an answer...
Old 4th May 2014
  #4622
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by chrisj View Post
You know, I keep beating on the difference between 'telling the difference' and 'telling the difference every time' and it'd be nice to see the abstraction more closely approach people's normal experience with an audio device.

You are absolutely right, because humans are inconsistent and try to pull generalizations out of noisy data. We get a sense of 'hey, that's kinda flat' one time in six and it's the end of the world, suddenly the CD audio or whatever is HORRIBLY flawed and we're ready to swear we could pick it out every time, blindfolded, upside down with our heads in a bucket of Jello.

And we can't, that's established. It requires alarming efforts to bring ANY phenomena of that nature up to the 'pick it out every single time' point (I speak from experience as one who's made just such alarming efforts)
Except experiments (and experience) have shown it works the opposite way. Once we hear something, we never stop hearing it. The amazing thing is that we continue hearing it even under sub-optimal conditions. The first time I experienced this, decades ago as a young guitarist who didn't know anything about audio, I was shocked to hear a quiet guitar part in a Zep song I thought I knew intimately. My friend's stereo system and speakers were much better than mine, so I had never heard the subtle mix elements so clearly before. It was a revelation. But what blew me away was that when I went home to my crappy Panasonic system, I could still hear the guitar part, even though I never once heard it on that system before in 100s of hours of deep listening.

If a bass part sounds fine the first dozen times listening, but on the 13th time we hear that it's flat, it will always sound flat from that point on. It's how our brains work.

Quote:
Some of you guys are insisting until you're blue in the face that people must test until they find the line of what's consistently reproducible, and then THROW OUT any perceptions that hint at crossing the line. That's putting the theory ahead of the data. Data is noisy when it's human. Seeking an invariant threshold and insisting that and that alone is reality, is a completely unsupported conclusion, as unreasonable as what Neil's doing (insisting that the most far-flung data points are not only important, but important to everybody).
Actually, you have it in reverse: theory tells us that crazy anomalies are possible, whereas the data -- the things that actually happen in the universe -- tell us otherwise. For example, theory predicts that the sum of random noise sources will follow a normal distribution. Indeed, an unconnected resistor sitting on the table will exhibit these random voltages. According to theory, every so often we should experience huge voltage spikes arcing across the resistor. But in the real world, where resistors have parasitic reactances preventing such anomalies, we never experience it.

The limits of human hearing have physiological (i.e. real, not theoretical) causes. Assuming that our cumulative thresholds are long-tailed is a theoretical stance that is not backed up by the data.
Old 4th May 2014
  #4623
Lives for gear
 

Quote:
When I started doing that with these machines, it started to hurt, and I couldn’t do it for very long, so the part of the record-making experience that I used to enjoy became painful. That was a sign to me that something was wrong.
Dynamic range /thread.

Chris
Old 4th May 2014
  #4624
Lives for gear
 

Old 4th May 2014
  #4625
Lives for gear
 
Ephi82's Avatar
 

Holy Crap! (as the late great Peter Boyle used to say on TV)

This is no longer a debate, but a handful of posters "babbling" the same general statements again, and again.

It's great that the nastiness seems to have gone by the wayside!

( i will admit to an addiction to this thread, and have posted a lot myself. Its like being in a train wreck you can watch on Youtube)

PONO, please rush the release of the product!!!!! We need more "meat" to keep the thread "vital"
Old 4th May 2014
  #4626
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Ephi82 View Post
This is no longer a debate, but a handful of posters "babbling" the same general statements again, and again.

It's great that the nastiness seems to have gone by the wayside!
... Except for the "babble" thing, I guess...
Old 4th May 2014
  #4627
Gear Maniac
 
toine6's Avatar
Screw Pono. Long live the cassette!! Sony Crams 3,700 Blu-Rays' Worth of Storage in a Single Cassette Tape

"Sony just unveiled tape that holds a whopping 148 GB per square inch, meaning a cassette could hold 185 TB of data. Prepare for the mixtape to end all mixtapes."
Old 4th May 2014
  #4628
Lives for gear
 
Ephi82's Avatar
 

Quote:
Originally Posted by nuthinupmysleeve View Post
... Except for the "babble" thing, I guess...
Its ok to agree to disagree.........

My assessment has been that a smallish number of posters have been "yelling" the same things at each other (ie babble) in this thread. (incl me)

If you want to call it expert opinion, or scientific position, or audio oracle, I can handle that.

Doesn't change the content IMO.

Peace.
Old 4th May 2014
  #4629
Quote:
Originally Posted by Ephi82 View Post

My assessment has been that a smallish number of posters have been "yelling" the same things at each other (ie babble) in this thread. (incl me)
.
I think that's basically right.
There's nothing much left to say until people start hearing Pono in person.
Old 4th May 2014
  #4630
Gear Addict
 

Quote:
Originally Posted by chrisso View Post
There's nothing much left to say until people start hearing Pono in person.
Well because I've missed alot of this, I've got a couple paragraphs, and a question:

Quote:
Originally Posted by bogosort View Post
If a bass part sounds fine the first dozen times listening, but on the 13th time we hear that it's flat, it will always sound flat from that point on. It's how our brains work.
.
If this is true, could such a thing be used as justification for the exclusive use of 192? A user compares 44.1 and 192 in a standard A-B test. The 192 "obviously" sounds better: wider, deeper, more open, better detailed, etc. etc. So the user decides to use 192 files exclusively. The user will always experience the "wider, deeper, etc. etc." because he/she knows what's playing is 192 and 192 has this "wider, deeper, better" sound. They don't just "think" it sounds better, it DOES sound better to them, even though it's their illusion. (I guess illusion is a bad word to use, but that's what I call it when I experience it, which I do every day. Haven't gotten used to the phrase "expectation bias" yet).

Having almost no technical knowledge and only recent knowledge of blind test and expectation bias, it just seems to me that's what's going on with the Pono people. I don't think they're intentionally lying. I think 192 really does sound better than 44.1 to them, and to every one of the artists in the video. But unless the player is just broken at 44.1, none of them could pick the files in a blind test. (my guess of course, just based on my experience noticing how much better 192 sounds in standard A-B, then failing ABX.)

I think expectation bias and how "real" the difference is, is one of the most interesting things about the sample rate discussion. And I've wondered how the quest for "quality" got so far into higher sampling rates for audiophile playback. Was there just nowhere else to go for higher quality? Didn't anybody say, "what are you guys doing? higher sampling rates aren't going to give you better quality or resolution or continuity or anything"? I'm sure somebody did, but like I said I've only recently started following this.

Unless I've misunderstood everything, that is the technical reality, isn't it? (I'm asking): The reconstructed analog waveform is just as precise in the audible range whether the digital capture was 44.1 or 192 or 384, given the same filtering? So what's coming out the speakers should be exactly the same in the audible range? And that's been borne out by the (too few) blind tests that have been done?

(sorry for such simple questions so late in the thread but I missed alot.)

Last edited by walter88; 4th May 2014 at 09:44 PM.. Reason: grammar
Old 4th May 2014
  #4631
Lives for gear
Quote:
Originally Posted by chrisj View Post
If you intend to engage with the Pono phenomenon in any rational or useful way, you're going to have to deal with the idea that people formulate their sense of a thing through scattershot experiences and generalizing to make a mental model of what they've got. Some of you guys are insisting until you're blue in the face that people must test until they find the line of what's consistently reproducible, and then THROW OUT any perceptions that hint at crossing the line. That's putting the theory ahead of the data. Data is noisy when it's human. Seeking an invariant threshold and insisting that and that alone is reality, is a completely unsupported conclusion, as unreasonable as what Neil's doing (insisting that the most far-flung data points are not only important, but important to everybody).
Awesome, awesome, awesome post. This is why Chris is such a valuable member of the community. This is exactly what the rigidly "scientific" people on gearslutz keep forgetting: human beings' perception of sound and sound technologies is extremely complex.
Old 4th May 2014
  #4632
Quote:
Originally Posted by bogosort View Post
Strange interpretation. I'd think most people would agree that the graphic is saying...
Just think about that. It's counterintuitive; the graphic clearly shows that to move from mp3 to CD it would take 800ft distance; from CD to 96K 180ft; and from 96K to 192K 20ft. You'd need to travel forty times the distance to get from mp3 to CD (800) compared with 96 to192K (20). The greater improvement in audio quality (as represented in the graphic) is from mp3 to CD...800ft; the least improvement is from 96K to 192K. The opposite of what your figures show. So why is that?

The distance between 96-192 is 20 when that is measured alone but what does that refer to on it's own? 20 of 1000ft gives reference: when measured with the other depth markings then that ratio is 20:1 of 20 terms whilst mp3 to CD is 5:1 of 800 terms.

Quote:
Originally Posted by bogosort View Post
  • 1000:200 -> CD is 5x better than MP3
  • 200:20 -> 96k is 10x better than CD
  • 20:1 -> 192k is 20x better than 96k
1000ft distance...terms of 800; 180; and 20

800:180:20 or

80:18:2

Therefore (in relationship to each other) if:
CD is 80x better than mp3
96 is 18x better than CD, then
192 is twice as good as 96.
Old 4th May 2014
  #4633
j_j
Lives for gear
Quote:
Originally Posted by walter88 View Post
Unless I've misunderstood everything, that is the technical reality, isn't it? (I'm asking): The reconstructed analog waveform is just as precise in the audible range whether the digital capture was 44.1 or 192 or 384, given the same filtering? So what's coming out the speakers should be exactly the same in the audible range? And that's been borne out by the (too few) blind tests that have been done?
If the filters are exactly the same, then you should see a 3dB increase in SNR at 96dB, and another 3dB increase at 192Hz (total of 6dB), assuming there was no noise shaping added (which is generally but not always the case).

As you can see from slide 72 in www.aes.org/sections/pnw/ppt/adc.ppt if you used various orders of noise shaping, you could get some modest SNR gain (bear in mind high-order delta-sigma is kind of touchy, too), but that's not what people have reported doing, and using more bits would actually be an easier solution (which is in the modern day the same as using a better oversampled convertor).

If, by the way, you want a good overview of conversion, that slide deck is pretty handy.
Old 4th May 2014
  #4634
j_j
Lives for gear
Quote:
Originally Posted by MarsBot View Post
Awesome, awesome, awesome post. This is why Chris is such a valuable member of the community. This is exactly what the rigidly "scientific" people on gearslutz keep forgetting: human beings' perception of sound and sound technologies is extremely complex.
And what a lot of other people forget is that understanding of human perception is not a dark art any longer, either.

Expectation bias is extremely powerful, and has been proven to exist over and over again, and in fact nobody tested has proven to be immune to it to the present.

If you're suggesting we should allow expectation bias to allow us to perceive better quality, well, placebos do have their place, but its a rather narrow application, because eventually people discover the placebo as such.
Old 4th May 2014
  #4635
Motown legend
 
Bob Olhsson's Avatar
 

Thumbs up

Old 4th May 2014
  #4636
Lives for gear
 
Ephi82's Avatar
 

Quote:
Originally Posted by chrisso View Post
I think that's basically right.
There's nothing much left to say until people start hearing Pono in person.
Please, bring it on yesterday!
Old 4th May 2014
  #4638
Lives for gear
 

Old 4th May 2014
  #4639
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by j_j View Post
If the filters are exactly the same, then you should see a 3dB increase in SNR at 96dB, and another 3dB increase at 192Hz (total of 6dB), assuming there was no noise shaping added (which is generally but not always the case).

As you can see from slide 72 in www.aes.org/sections/pnw/ppt/adc.ppt if you used various orders of noise shaping, you could get some modest SNR gain (bear in mind high-order delta-sigma is kind of touchy, too), but that's not what people have reported doing, and using more bits would actually be an easier solution (which is in the modern day the same as using a better oversampled convertor).
True, but 24 bit is pretty established. I think it's beyond what we can really sense, under any conditions, but the easiest way to get 6dB more noise floor is doing what you said, manipulating 192Khz to give it.

We can manipulate the sample data to do the noise shaping, too: creating the 24 bit file is where we'd go about it. I understand that TPDF is the technically correct option, but I cut my teeth on dither and noise shaping and still love that stuff heck, I put out free dithers/noiseshapers to widen people's options. Certainly dithering to 24 bit is correct, and if you can do that you can do stuff like POW-R 3 or MegaBitMax/Ozone or my dithers. High energy noise shapers will be less annoying at 24 bit than they are at 16, and this processing is on the file-creating end. If Pono is filtering out the silly highs, all the more reason to go with a noise shaper as it will 'fix' it while giving even more SNR.
Old 4th May 2014
  #4640
Lives for gear
The specific bias demonstrated in the McGurk Effect video is not relevant to the audio discussion because it's talking about how visual input simultaneous with sound can affect the perception of sound. All I have to do to remove that bias is close my eyes, which I often do when mixing or listening. It's easy to compensate for and remove that cognitive bias.

Also, the McGurk effect shows how visual and hearing systems have evolved a kind of synergism to help us process speech in real time. The expectations biases associated with audio are probably socially constructed not evolved biases and not operating in real time. I do understand that other types of biases can be operating when people evaluate sound gear. But I don't think they are the McGurk effect or nearly as powerful/persistent as the McGurk effect.

After years of working in audio and other creative arts, my experience is that the best artists and technicians do a lot of checking to try to deal with expectation bias. Yes some bias will always exist, but humans who are aware of it can often compensate or account for it. Are all expectation biases created equal? Do all of them persist the way the McGurk effect does? Are such persistent biases really what's happening when someone evaluates sound equipment?

So far people seem to be claiming a very specific type of expectation bias: that advertisers' statements or user reviews (social factors) will influence how a person hears audio gear and that influence will persist and be immutable to correction. It seems to me this is a much weaker basis for an expectation bias than something like the McGurk effect, and it might be more subject to correction and/or compensation, especially over time. Chris Johnson alluded to this in a previous post with his idea that we employ a complex mental model when evaluating audio gear.
Old 5th May 2014
  #4641
j_j
Lives for gear
Quote:
Originally Posted by chrisj View Post
If Pono is filtering out the silly highs, all the more reason to go with a noise shaper as it will 'fix' it while giving even more SNR.
Yes, but nobody is very sure what Pono is actually doing.

I'm almost banking on its doing 4x oversampling on 48K material, and then running an LP simulator.

Yes. Really.
Old 5th May 2014
  #4642
j_j
Lives for gear
Quote:
Originally Posted by MarsBot View Post
The specific bias demonstrated in the McGurk Effect video is not relevant to the audio discussion because it's talking about how visual input simultaneous with sound can affect the perception of sound.
Of course, that's completely wrong, expectation exists as long as you always know what you are listening to. Nothing more is required.

The McGurk effect is exactly germane, it is a very simple demonstration of expectation bias. You are simply using a specific example of an absolutely firmly established, proven as well as anything can be proven, behavior of the human organism to argue against the general case, which is just as well established.

So, what was your point again?
Old 5th May 2014
  #4643
j_j
Lives for gear
Quote:
Originally Posted by MarsBot View Post
So far people seem to be claiming a very specific type of expectation bias: that advertisers' statements or user reviews (social factors) will influence how a person hears audio gear and that influence will persist and be immutable to correction.
Now you're conflating two completely different things.

Expectation bias only affects the first half of your sentence. The part after "and that influence" is a straw-man extrapolation that is simply not the argument here.

Sorry, no, don't use the words "immutable" and anything relating to preference in the same sentence where you imply that one applies to the other. That's not what anybody is arguing that I can see, and certainly not what the informed are arguing.

So why are you attacking this position, anyhow? Why did you even invent it?
Old 5th May 2014
  #4644
Motown legend
 
Bob Olhsson's Avatar
 

JJs talk that I posted a link to really puts this all in context.
Old 5th May 2014
  #4645
Lives for gear
 

This will be one of the most successful kickstarter campaigns and at the same time one of the biggest consumer flops.

The only thing saving it from flopping is about the only thing that made it so successful, a famous name attached to, and fronting it along with a bunch of other famous people hyping it! Just look at Will Wheaton's tabletop IndieGoGo campaign, over a Milly for some board game youtube show, to highlight how any kind of fame will result in significant backing of mediocre ideas and things.

It also reeks highly of just another marketing campaign as funnily enough his record and the pono got released at nearly the same time. Great way to get paid to advertise and build hype! Genius really...

Last i heard they were still dodging what their cut was of not only the $6 Milly but every ongoing download made from PONO.

The triangle design is highly flawed and seems like a big step backward in technology when it is flouting to be the latest and greatest thing!

It does have some famous endorsements, but those either came as a posse thumbs up, they were paid, or bigger artists were promised a better cut then they get from Apple. Either way, none of it can be trusted as is.

All that being said, a genius move from Neil albeit I think he may have sort of hurt himself along the way with the design and some of the marketing hype he spewed about it. To get the masses to buy it up like Beats, they either have to be the coolest thing around, which it isn't, or else speak to their pocket book, which it doesn't! They will need some serious advertising that is just so cool to have a chance, however with geekdom clearly still very popular maybe their audiophile jargon will work out for them. Good luck either way, if it takes off Neil will make an absolute killing off the 30% or whatever they will take beyond what he already generated.

There is always a choke point and unless they can convince people to wear studio quality headphones around, beats are not, then the quality is going to be degraded anyways at some point. I mean, what's with the headphone jack? the damn thing is clearly big enough to support the proper connection. This is the biggest point to me that they are seriously targeting the low level sheep consumer and or really believe HQ audio at a higher price all around will take over the market, either way, they messed that up big time!

If they are still unable to get and make viable pristine high quality sound in the lowest of consumer television and viewing experiences, sustainable portable high quality music players at an affordable price will be difficult to achieve. Overall this will be a tough go, beyond their incredible kickstarter grab, especially with Apple etc. already having an iron hand vice grip on the market.

All that being said, it's nice to see quality being cared for in some fashion.
Old 5th May 2014
  #4646
The blind acceptance that musicians are basically corrupt is amazing. Well it would be if I hadn't seen it so many times on Gearslutz before.
Anyway, isn't the cut 70/30? Presumably 70 to Pono.
Old 5th May 2014
  #4647
Lives for gear
Quote:
Originally Posted by j_j View Post
Of course, that's completely wrong, expectation exists as long as you always know what you are listening to. Nothing more is required.

The McGurk effect is exactly germane, it is a very simple demonstration of expectation bias. You are simply using a specific example of an absolutely firmly established, proven as well as anything can be proven, behavior of the human organism to argue against the general case, which is just as well established.

So, what was your point again?
I am not the one who is "using a specific example of a firmly established . . . behavior." That was Chris93 who brought up the specific example of McGurk. I'm just pointing out that specific example doesn't apply in the case of audio expectations biases which are probably social. It's not a good example to use in this discussion.

And I am not arguing against the general case that expectations bias can exist. My post specifically admitted that they do. I'm suggesting that the issue is more complex than simply having unconscious biases that inevitably determine behavior.

McGurk actually demonstrates that an individual could compensate for a known expectation bias. That is not an argument against expectations bias generally. It is an observation that expectations bias may not be such an either/or thing. Also, it's obvious that not all expectations biases will be equally strong, so there is probably a continuum. Would you care to address those specific issues, or would you rather engage in word play as you do in your second post critical of my McGurk comment?

I'm sick of the argument that McGurk applies to audio gear evaluation when it obviously doesn't. Please pick an example that applies to extend the discussion of expectation biases as they relate to audio gear evaluation. I'm perfectly willing to consider examples that are actually relevant to this discussion.
Old 5th May 2014
  #4648
Gear Guru
 
Sounds Great's Avatar
 

Quote:
Originally Posted by Bob Olhsson View Post
Quote:
If you like it, you like it, and that's all there is to it.



Quote:
If we can work together we can solve some of this. 20 years ago I don't think we could have because frankly we couldn't analyze the systems well enough but now we can.
Hold on JJ, what's changed in the last 20 years about analyzing the systems? And if there really has been a change, how can you say in 20 years from now we won't be able to do it even better and find some things we are missing now?
Old 5th May 2014
  #4649
j_j
Lives for gear
Quote:
Originally Posted by MarsBot View Post
I am not the one who is "using a specific example of a firmly established . . . behavior." That was Chris93 who brought up the specific example of McGurk. I'm just pointing out that specific example doesn't apply in the case of audio expectations biases which are probably social.
Um, you're wrong. I don't know quite how to put it any more clearly. You will get expectation bias if you tell somebody that A is "red" and B is "green" with no red and green present, and no other people in the room.

You'll get expectation bias if you label them A and B.

Or 1 and 2

Or 99 and 54

It won't matter.

Your claim that expectations are social is not incorrect, in that they are also social, but they exist without any social interaction or social expectation.

Please, go watch the clip Bob Ohlson posted, ok?
Old 5th May 2014
  #4650
j_j
Lives for gear
Quote:
Originally Posted by Sounds Great View Post
Hold on JJ, what's changed in the last 20 years about analyzing the systems? And if there really has been a change, how can you say in 20 years from now we won't be able to do it even better and find some things we are missing now?
20 years ago I could not find an ADC or a microphone that had a wider dynamic range than the auditory system.

There were some few FM recorders that might come kind of close.

I'm sure we will understand even more in 20 years, but we now have the tools to at least proceed with an analysis that isn't limited by the tools.

On the other hand, this mythical "missing now" is just another appeal to the fallacy of ignorance. The fact that we do not know everything does not mean that we can not know something, and in fact, we do know the absolute sensitivity of the ear, for instance, and as such, finding something "missing" would seem to violate (literally, not figuratively) the understanding of physics and mathematics that would seem very well supported indeed.

Note: The ear can hear very close to the actual atmospheric noise due to brownian motion at the eardrum. If you think you're missing something below that, you're making a very interesting claim in physics and thermodynamics. Note also that microphones can detect this particular noise, so that does set a limit on the necessary sensitivity.

So it's not so hard as you imagine to say that "we have the tools" now. Really.
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