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 27th March 2014
  #1951
Lives for gear
 
bogosort's Avatar
Well I can definitely hear a difference between your 10k square waves when sampled at 44.1 and 192:

PHP Code:
foo_abx 1.3.4 report
foobar2000 v1.2.9
2014
/03/27 16:34:21

File A
C:\downloads\square-10khz_at_44ksr.wav
File B
C:\downloads\square-10khz_at_192ksr(1).wav

16
:34:21 Test started.
16:35:10 01/01  50.0%
16:35:20 02/02  25.0%
16:35:28 03/03  12.5%
16:35:59 03/04  31.3%
16:36:31 04/05  18.8%
16:36:40 05/06  10.9%
16:36:47 06/07  6.3%
16:37:04 07/08  3.5%
16:37:11 08/09  2.0%
16:37:22 09/10  1.1%
16:37:53 10/11  0.6%
16:38:09 11/12  0.3%
16:38:16 12/13  0.2%
16:38:22 13/14  0.1%
16:38:30 14/15  0.0%
16:39:06 Test finished.

 ---------- 
Total14/15 (0.0%) 
While it's possible I got very lucky, it's unlikely (I wasn't guessing, and I hid the results until after I completed the test). I used ReplayGain to level-match the files, but the ABX "tell" for me was a slight difference in level.

Looking at this a bit closer, I think I see what's happening. The ReplayGain algorithm sees all the extra harmonic energy in the 192k capture and so lowers the total level correspondingly. But none of that energy is audible! So the 44.1 capture turns out to be louder than the 192 version, which is an obvious tell. (Note that even without ReplayGain, there is a slight volume difference in the files: I could easily ABX the two. Level-matching will be the biggest challenge in this experiment.)

This little muck up exemplifies the many difficulties in setting up a proper ABX test. Caveat auditor!
Old 27th March 2014
  #1952
Lives for gear
I should of had a mic handy, I could of got a sample of 1 Million dollars being wasted.

damn.

D
Old 27th March 2014
  #1953
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by jkorten View Post
Please contribute to my new signal generator fund - and I will provide distortion free recordings of sinewaves and squarewaves!
Don't take it as an attack. I it is peculiar that the sine wave isn't clean. That shouldn't happen even on a cheap generator. I suspect something else might be going on. It is only about 43 dB down from the main sine which is significant!

Quote:
None of this will affect your ability to tell the difference between the 192 version and the 44.1 version as they are the same signal being sampled. Can you tell the difference?

Can you tell the difference between the sinewave and the squarewave at 10kHz at 192K?

J
I haven't had time to listen yet but I will.

Alistair
Old 27th March 2014
  #1954
Gear Nut
 

Quote:
Originally Posted by bogosort View Post
Well I can definitely hear a difference between your 10k square waves when sampled at 44.1 and 192:
While it's possible I got very lucky, it's unlikely (I wasn't guessing, and I hid the results until after I completed the test). I used ReplayGain to level-match the files, but the ABX "tell" for me was a slight difference in level.

Looking at this a bit closer, I think I see what's happening. The ReplayGain algorithm sees all the extra harmonic energy in the 192k capture and so lowers the total level correspondingly. But none of that energy is audible! So the 44.1 capture turns out to be louder than the 192 version, which is an obvious tell. (Note that even without ReplayGain, there is a slight volume difference in the files: I could easily ABX the two. Level-matching will be the biggest challenge in this experiment.)

This little muck up exemplifies the many difficulties in setting up a proper ABX test. Caveat auditor!
Hmmm, I wonder if using Audacity's "normalize" feature on the files would help as I think that is normalizing to RMS levels if I understand the function correctly. You'd want to normalize to something like -20db or so in order not to cook your eardrums.
Old 27th March 2014
  #1955
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by jkorten View Post
Hmmm, I wonder if using Audacity's "normalize" feature on the files would help as I think that is normalizing to RMS levels if I understand the function correctly. You'd want to normalize to something like -20db or so in order not to cook your eardrums.
Yes, good call. After normalizing both files to -20 dB, I tried another ABX and gave up without completing even one trial because I was guessing.

I've uploaded the two files here; I don't believe anyone can ABX these.
Attached Files

square-10khz_at_44ksr_normalized.wav (938.3 KB, 49 views)

square-10khz_at_192ksr_normalized.wav (4.40 MB, 61 views)

Old 27th March 2014
  #1956
Lives for gear
 
nuthinupmysleeve's Avatar
 

Tried it using ABXTester on the Mac.

Ran the mac's built in converters on my macbook pro retina at 96khz (so that's a compromise already).

I couldn't tell at all. Not even close, just guessing.

Thanks for posting the files!
Old 28th March 2014
  #1957
Lives for gear
 

Quote:
Originally Posted by jkorten View Post
The relative strengths of the harmonics are totally different though, and I'm not sure of the utility of speaking of our hearing in the frequency domain. We don't do an FFT, we have hair cells in a continuous fluid column that are picking up these signals. I am also going to guess that there is interaction going on between harmonics etc. at that level as well.
On a vastly oversimplified level, the human auditory system has a bit in common with an AD converter, except the biological method requires many different structures to derive the combined, coherent signal which we perceive. During this journey the auditory structures supply information chunks which are sometimes temporally discrete (being derived from change in action potentials) and are even synthesized at some points (during binaural fusion for example).
In other words, don't be so sure that a "continuous fluid column" means the incoming information is continuous like a sine wave. We perceive it as continuous, but that's different to your implied assertion regarding "continuous fluid". The inner ear is very similar to a spectrum analyzer via arrays of mechanical receptors (hairs) which detect frequencies. In this regard, we do indeed "do an FFT" in the sense of doing a pass of more or less discrete spectrum analysis channels which are combined in various ways downstream. Each hair is a different band.
From this Yamaha document: "the cochlea is one of the most amazing organs in the human body, as it basically constitutes a very sensitive 3,500-band frequency analyser with digital outputs - more or less equivalent to a high resolution FFT analyzer"
Quote:
Originally Posted by jkorten View Post
But again - I go back to my perception. If I can hear the difference, the difference is important to me. If I cannot, then it wouldn't. But it is! So I care.
It's all about you, eh?

Even after disregarding the fact that perfect sine tones do not occur in nature, hearing a difference using a sine wave as the source is not an indicator of much at all, given the many variables applicable to your "test".

It is a meaningless distraction to use a half-baked home-brewed test to dispute countless prior tests which were made adhering to the scientific method.
Old 28th March 2014
  #1958
Gear Nut
 

Quote:
Originally Posted by diggo View Post
On a vastly oversimplified level, the human auditory system has a bit in common with an AD converter, except the biological method requires many different structures to derive the combined, coherent signal which we perceive. During this journey the auditory structures supply information chunks which are sometimes temporally discrete (being derived from change in action potentials) and are even synthesized at some points (during binaural fusion for example).
In other words, don't be so sure that a "continuous fluid column" means the incoming information is continuous like a sine wave. We perceive it as continuous, but that's different to your implied assertion regarding "continuous fluid". The inner ear is very similar to a spectrum analyzer via arrays of mechanical receptors (hairs) which detect frequencies. In this regard, we do indeed "do an FFT" in the sense of doing a pass of more or less discrete spectrum analysis channels which are combined in various ways downstream. Each hair is a different band.
From this Yamaha document: "the cochlea is one of the most amazing organs in the human body, as it basically constitutes a very sensitive 3,500-band frequency analyser with digital outputs - more or less equivalent to a high resolution FFT analyzer"

It's all about you, eh?

Even after disregarding the fact that perfect sine tones do not occur in nature, hearing a difference using a sine wave as the source is not an indicator of much at all, given the many variables applicable to your "test".

It is a meaningless distraction to use a half-baked home-brewed test to dispute countless prior tests which were made adhering to the scientific method.
Interesting comparison to the FFT - it is approximate of course as the hair cells occupy physical space that can't make the transformation "point-wise". The point I was responding to had to do with why we'd hear a difference between a high frequency sine wave vs square wave, that the missing (to our hearing range) harmonic might have another effect rather than having to be perceived. (Thinking over the night, perhaps the rise time of the waveform is what is being perceived rather than the harmonic.)

Yes - my world is all about me, and the choices I make about audio have to do with what I percieve and want, not you Diggo, or anybody else. Sorry to break it to you. This is why some people spend $500 on a stereo and some spend $50,000.00.

There are also no perfect square waves in nature. These two signals are mathematically tractable which is why they were chosen (we know their frequency transforms etc.). As bogosort showed - even my signal generator can't generate a perfect sinewave though. I'm not sure why this is relevant. What is relevant is if you can discriminate a sine and a square where the first harmonic of the square is "beyond the range of human hearing" then why do we hear a difference?

We unfortunately got stuck on 10kHz, but thinking over it, we chould choose a sine/square comparison where the first harmonic is way lower, but still beyond perception - what would that be 7kHz? (21kHz for the first harmonic). Why would we hear a difference? If you think it is a meaningless distraction it doesn't hurt my feelings. Away with those negative feelings I say.
Old 28th March 2014
  #1959
Quote:
Originally Posted by bogosort View Post
Yes, good call. After normalizing both files to -20 dB, I tried another ABX and gave up without completing even one trial because I was guessing.

I've uploaded the two files here; I don't believe anyone can ABX these.
At first, I was sure I could hear a difference but then I downloaded the free ABX tester from the mac app store and realized I couldn't.

So stoked to have that app now! Going to be a lot of fun with clients and other AEs.
Old 28th March 2014
  #1960
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by jasonwagner View Post
At first, I was sure I could hear a difference but then I downloaded the free ABX tester from the mac app store and realized I couldn't.

So stoked to have that app now! Going to be a lot of fun with clients and other AEs.


This type of experience is often ignored by the critics of ABX testing. They (justifiably) question the broad implications of a particular study while ignoring or dismissing the undeniable insight that can be provided by an individual simply testing himself.

We've all experienced what we think are night and day differences when switching between gear. But when those supposedly profound differences literally disappear when we put the blindfold on, something amazing happens. Rather than cast a shadow of doubt over our hearing ability or mental fortitude, the newly formed insight actually sharpens our perception (take off the blindfold and the differences are still gone!), and the increased self-awareness leads to greater confidence in making audio-related decisions.

Even if one believes that ABX is fundamentally incapable of sussing out subtle or even subconscious differences, one cannot deny its power at exposing night and day differences as being subtle, at best. As audio engineers, an occasional reality check of our perception is a Good Thing.
Old 28th March 2014
  #1961
Motown legend
 
Bob Olhsson's Avatar
 

Real working engineers battle expectation bias every single time we do anything. Subtle differences to one person can become real distractions to others especially upon repeated listening.

I have guitar shredder mastering clients who can spot a change in dither settings in a heartbeat even though I find it very subtle. This whole mindset of "good enough" doesn't exist among the people I've known who made successful recordings.
Old 28th March 2014
  #1962
Quote:
Originally Posted by bogosort View Post
[...]

Looking at this a bit closer, I think I see what's happening. The ReplayGain algorithm sees all the extra harmonic energy in the 192k capture and so lowers the total level correspondingly. But none of that energy is audible! So the 44.1 capture turns out to be louder than the 192 version, which is an obvious tell. (Note that even without ReplayGain, there is a slight volume difference in the files: I could easily ABX the two. Level-matching will be the biggest challenge in this experiment.)

This little muck up exemplifies the many difficulties in setting up a proper ABX test. Caveat auditor!
THIS.

ABX testing is one of our best tools when dealing with issues of audibility and preference -- but for it to be a 'fair' test, everything must be set up with great care -- and things that may not, at first, seem obvious can have effects that can seriously distort one's conclusions if they are not factored out.
Old 28th March 2014
  #1963
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by Bob Olhsson View Post
Real working engineers battle expectation bias every single time we do anything. Subtle differences to one person can become real distractions to others especially upon repeated listening.

I have guitar shredder mastering clients who can spot a change in dither settings in a heartbeat even though I find it very subtle. This whole mindset of "good enough" doesn't exist among the people I've known who made successful recordings.
Well, it's never really that simple, is it? On the one hand, "good enough" suggests that X is good but Y is better, with the implication that if we choose X we're just being lazy. But that's hardly ever the case, right? Audio engineers are audio junkies; they're the least likely of all the people in a recording arrangement to cut corners that would affect the sound, especially if their name is going to be attached to it.

On the other hand, everything in the real world is a matter of "good enough". No one has unlimited resources, so we make Pareto charts (whether we realize it or not) and prioritize our efforts. Surely you agree that part of being a working engineer is knowing when to say when. The "good enough" mindset, whether we want to recognize it as such or not, is a very real part of all our lives.

To the main gist of this thread, many of us are claiming that 44.1 as a delivery format is not just "good enough", but that it provides exactly the same information as a bandlimited 192k capture. And because that bandwidth corresponds to the audible bandwidth, there is no reason to deliver with higher sample rates.

Edit: Removed cheeky attempt at humor.
Old 28th March 2014
  #1964
Lives for gear
 

I can see a use of high sample rates for recording bats, then pitching it down so that humans can hear it.

Chris
Old 29th March 2014
  #1965
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Bob Olhsson View Post
Real working engineers battle expectation bias every single time we do anything.
I bet, but how do you battle it? Do you "test" in some way to eliminate expectation bias as why you might be hearing a difference? Seriously curious!
Old 29th March 2014
  #1966
14,732 backers; $ 4,998,891 pledged; 17 days to go.
Old 29th March 2014
  #1967
Gear Head
 
bandpass's Avatar
 

Don't forget that these are signature models—a good percentage of them will probably be put on eBay as soon as they are delivered, by folk trying to make a fast buck before the hype wears off.
Old 29th March 2014
  #1968
Gear Guru
 

Quote:
Originally Posted by Chris93 View Post
I can see a use of high sample rates for recording bats, then pitching it down so that humans can hear it.

Chris
Ultrasonic sniffers are used in industry to detect tiny leaks in pressurized systems. They pitch the stuff up around 40k down to audible frequencies to help locate the leak.

For the casual bat-watcher, they can be pretty expensive, but this company expresses the cost of the unit in terms of how many hours an 1/8" pinhole in a compressed air line will cost you the same money! heh
Old 29th March 2014
  #1969
Gear Guru
 
Kenny Gioia's Avatar
 

Quote:
Originally Posted by diggo View Post

Extraordinary claims require extraordinary proof, especially regarding bias.
Probably the most important sentence in this thread.

For those saying they hear a difference, while some argue that you can't, I'll argue that maybe you're right. Maybe you do hear a subtle difference.

But if you hear an extraordinary difference, then an ABX test shouldn't be a problem.
Old 29th March 2014
  #1970
Pono Music Player Climbing Kickstarter’s Rankings, Now 4th Behind Veronica Mars
Quote:
Thanks everyone for all of your support. It has been and continues to be gratifying to me to see how many of you are supporting Pono and our revolution in the way people can listen. Now we will have a choice that is as convenient as it is great sounding. We are celebrating our Freedom of Choice.
Neil Young
Source: Pono Music Player Climbing Kickstarter's Rankings, Now 4th Behind Veronica Mars | Crowdfund Insider
Old 30th March 2014
  #1971
Old 30th March 2014
  #1972
Here for the gear
 

Quote:
Originally Posted by KFW View Post
Ridiculous. Hardware and speaker quality is what will give you a noticeable difference (and I'm sure that's what made the music in that car sound great). I can't imagine who has fooled themselves to think this nitpicking of resolution makes a difference. You can even do it yourself in your recording program--just render a super high quality mp3 and a lower quality wav file, and be honest with yourself for a moment.

An article worth reading, that actually explains in great detail why 24/192 actually sounds WORSE in playback.

24/192 Music Downloads are Very Silly Indeed

And I have no idea where the argument that "people don't care about music fidelity nowadays" comes from. I'd say they care more than ever--hence why Beats have gotten so popular, and plenty of other headphones that actually reproduce the bass that modern productions have. Technology is improving, and it's not rare to see a casual music consumer drop over $100 on headphones so they can enjoy music more. I'd say that's a pretty far cry from 80's Walkman headphones on a portable tape player.
I very much appreciate that link. Sounds pretty definitive to me.
Old 30th March 2014
  #1973
Quote:
Originally Posted by bogosort View Post

[*]Those who think they hear a difference.[*]Those who don't hear a difference but would rather err on the side of "future proof".

In my experience, those in the first group are the least sophisticated: they don't understand digital audio and they haven't tried testing their belief; ignorance is bliss. The second group means well but they don't understand Shannon, so they think maybe one day in the future we'll be able to use the "extra information".
If nothing else, your side of the debate has routinely portrayed anyone who 'thinks they can hear a difference', or are even slightly open to the bandwidth not being settled as morons.
Funny that.
I don't really have a horse in the race. I don't support Pono or 192khz.
I agree it's a sideshow that is highly unlikely to help the cause of average artists.
I just had to comment that every time I dip into this thread I see those who are convinced on 44.1khz making negative character assessments of those who aren't.
Surely you can make your argument without calling people 'ignorant' and making patronising comments about 'well meaning' people who 'don't understand'?
Old 30th March 2014
  #1974
Just a heads up that we aren't really up for adhomenim attacks any longer.

We find them hard to police as they don't use profanities but anyone indulging in them might find infractions issued as we play "catch up".

Sneaky, just under the radar comments might get through and perhaps only the more obvious one might get infracted. But that's your look out.

To give you a glimpse behind the curtain here at Gearslutz.com adhominem attacks in threads with a science / physics theme - really pisses us off. So we are keen to edge people doing it towards the exit door.

Any how - for those indulging in attacking the person and not the argument - You might find yourself looked out of the forum as a result.

Advice going forwards is - apologise / edit your posts if you have been doing it. Don't continue doing it.
Old 30th March 2014
  #1975
Thanx jules for that...

On the subject at hand ...
Old 30th March 2014
  #1976
Gear Guru
 
Kenny Gioia's Avatar
 

Quote:
Originally Posted by bogosort View Post
  1. Those who think they hear a difference.
  2. Those who don't hear a difference but would rather err on the side of "future proof".
  3. Those who believe ultrasonic frequencies are a factor in music appreciation
In my experience, those in the first group are the least sophisticated: they don't understand digital audio and they haven't tried testing their belief; ignorance is bliss. The second group means well but they don't understand Shannon, so they think maybe one day in the future we'll be able to use the "extra information". The third group is just batty. (See what I did there? )
I don't know if this is part of the second group or a new fourth group but I would imagine that there are some people that run their mixes at higher frequencies in case the label wants to one day release them as high res files.

Whether it sounds different is not the point. It's a business. If labels want a song mixed at 384kHz, mixers will deliver them.

And as an aside, ignorant does not mean you're stupid or an idiot. It means you don't have the proper information. For example, everyone in this thread is ignorant as to what the people in that video experienced.

Thanks
Old 30th March 2014
  #1977
Gear Guru
 
Kenny Gioia's Avatar
 

Quote:
Originally Posted by luizdepalma View Post

On the subject at hand ...

Article
My apologies but that article is pretty darn useless.

Unless I missed something, did they compare the Pono to anything?

I have to hand it to them. They took away confirmation bias by not even listening to another source.

What the hell?

And is the "mainstream" going to listen to the Pono on that system?

An 8k amplifier with a pair of 37k speakers?

You would have thought they could have found a $30 CD player to compare.
Old 30th March 2014
  #1978
... that's why I posted it K. It is just a review from an audiophile who experienced Pono first hand.
Old 30th March 2014
  #1979
Here for the gear
 
Sean M Robinson's Avatar
 

This just in on the ultra-high frequency news network. Pono's going to have to step up it's game (i.e. sampling rate) if it's going to compete in the emerging audio market of portable dolphin communication--

Scientists Translate Dolphin Whistles | I ****ing Love Science

"Thad Starner at the Georgia Institute of Technology built CHAT using pattern-discovery algorithms that are designed to analyze dolphin whistles, extracting features that we wouldn’t know to look for. After all, dolphins produce sounds with frequencies up to 200 kilohertz, or 10 times higher than the highest pitch we can hear.
Read more at http://www.iflscience.com/plants-and-animals/scientists-translate-dolphin-whistles#lc5mS87bKF1oOiAJ.99"
Old 30th March 2014
  #1980
Lives for gear
 
doncaparker's Avatar
 

"So long, and thanks for all the fish."
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