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 12th April 2014
  #2941
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Karloff70 View Post
I am not convinced. I have hope. I had hope this Pono initiative would turn out a result of positive nature for audio from the beginning. Chris just pulled out some stuff to make that seem more likely, making me smile. I still have hope, regardless of whether what Chris is saying is applicable here or not. All you lot still have your cynicism. Be like that. Time will tell how this turns out. Meanwhile I have hope and you have doubt. Enjoy your doubt.
You are really a funny guy. It's not at all cynical to conclude that people can't hear the difference... that's exactly what the evidence shows in multiple a/b/x tests in scientific settings. That's not "cynical." That rational.

What is irrational is to believe, in spite of all scientific evidence to the contrary, there is "hope." That is magical thinking.. at least at this point.

If evidence comes out to the contrary, I'm there with you. Believe me.

I have no horse in this race. I'm neither for nor against 96/24 as a format. I only know I cannot hear the difference, and I only know that nobody has been able to hear the difference in a real world scientific a/b/x test. I appreciate Chris' testing, BTW... but what he did doesn't appear to be a real world test to me.

I will suggest that you don't seem to have a handle on the damage this can do to an already damaged record company reputation when people realize there might not be a difference even though they paid a lot more money. How many times will you buy the same music? Me... I'm done re-buying stuff.
Old 12th April 2014
  #2942
Lives for gear
 
Ephi82's Avatar
 

Quote:
Originally Posted by nuthinupmysleeve View Post
What remains to be seen is whether anyone will care.

Please try yourself to do an a/b/x test. Grab a 96/24 file you know well. Use high quality sample rate conversion to 44/16, the convert back to 96/24. This will effectively remove all higher frequencies and lower the bit depth. Be sure your system is running at 96/24. Then run an a/b/x test yourself. Please report back! The more data the better.
I agree that people will have to care (quite a bit) about better sounding music for success of PONO.

I have closely followed discussions here and other forums about ABX testing and the results. I am fully comfortable with the idea that there's a human bias built on expectation, but the entire recording industry is tracking, mixing and archiving in a 96+/24 bit environment.

Why not produce at 44/16 and archive with lossy MP3 320's?

No one can convince me that i dont hear a much weightier and detailed sound when tracking and mixing the recordings I make in my studio now because of the quality of my RADAR's converters, analog stages (btw, I track at 44/24 bit because I only get 8 tracks at 96khz)

As I have been saying, its not a format alone issue, its also raising the bar on the quality of the hardware.
Old 12th April 2014
  #2943
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by paul brown View Post
to qualify, are you saying that it necessitates a new filter design to be able to hear the differences?
Yes. I'm prepared to say that it necessitates a moving-average filter design to get THAT much of a difference simply from moving to 192K. Using normal digital audio, all you get is extended frequency response that's useless, up at 96K. Using the moving average means you take the existing data you have, and use it in a way people can actually hear. Both in the complete and total absence of ripple (irrelevant if the ripple would be at 96K anyway) and in the resolution gain through averaging AUDIO frequencies. For all practical purposes, even the high frequencies would end up averaged, it's just that at 192K they will include absolutely everything you could want in audible sound. i.e. 20K and under.

Quote:
Originally Posted by nuthinupmysleeve View Post
But it's interesting that everyone else disagrees with him so far... people like Dan Lavry and other audio experts. But... all it takes is one guy on the internets and you're convinced?!?
On the contrary, Dan Lavry would be the FIRST person to tell you that 96K audio content is useless and this averaging technique is the only relevant way to handle that information. There may be converter designers who would want you to keep the 96K frequencies in the audio output, but Lavry is not one of them.
Old 12th April 2014
  #2944
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by nuthinupmysleeve View Post
You are really a funny guy. It's not at all cynical to conclude that people can't hear the difference... that's exactly what the evidence shows in multiple a/b/x tests in scientific settings. That's not "cynical." That rational.

What is irrational is to believe, in spite of all scientific evidence to the contrary, there is "hope." That is magical thinking.. at least at this point.

If evidence comes out to the contrary, I'm there with you. Believe me.

I have no horse in this race. I'm neither for nor against 96/24 as a format. I only know I cannot hear the difference, and I only know that nobody has been able to hear the difference in a real world scientific a/b/x test. I appreciate Chris' testing, BTW... but what he did doesn't appear to be a real world test to me.

I will suggest that you don't seem to have a handle on the damage this can do to an already damaged record company reputation when people realize there might not be a difference even though they paid a lot more money. How many times will you buy the same music? Me... I'm done re-buying stuff.
I am long done rebuying anything, more concerned with new music anyway. But unlike you I don't feel that there is any sort of reputation to protect in record companies. Are there really still people who do not realise that record companies do just about anything to make another buck??

In any case, feel free to call me a 'funny guy'. Doesn't cost me none. We'll see if this thing comes out in a positive. I'll keep hoping so, totally without evidence and you can stay sceptical.
Old 12th April 2014
  #2945
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Ephi82 View Post
No one can convince me that i dont hear a much weightier and detailed sound when tracking and mixing the recordings I make in my studio now because of the quality of my RADAR's converters, analog stages (btw, I track at 44/24 bit because I only get 8 tracks at 96khz)

As I have been saying, its not a format alone issue, its also raising the bar on the quality of the hardware.
Buying hand made IEC power cables for your gear is also "raising the bar on the quality of hardware." But will it make an audible difference? There's lots of ways to raise the hardware quality bar.. what's the point if it doesn't improve the sound?

I appreciate your perspective but an improvement just for improvement sake doesn't make sense to me.

Again, I'll just ask... please try and a/b/x test of material you know. It was eye opening to me when I tried it. Nobody is trying to convince you what you hear is true or false. The only way to know for sure is to do an a/b/x test to eliminate the human real world condition of expectation bias.

Once you see for yourself that there is or is not a difference, eliminating your bias, you can make a decision based on real world data.
Old 12th April 2014
  #2946
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Karloff70 View Post
feel free to call me a 'funny guy'.
You're a funny guy because you accuse people who are basing their decisions on known science as being "cynical." If cynical is accepting known science... than I'll wear that banner with pride. But that is not really the definition of cynical.

I have no problem with your belief and hope, that's awesome. But trying to mislabel a rational conclusion based on all known data as "cynical" is... well, rather funny.
Old 12th April 2014
  #2947
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by nuthinupmysleeve View Post
You're a funny guy because you accuse people who are basing their decisions on known science as being "cynical." If cynical is accepting known science... than I'll wear that banner with pride. But that is not really the definition of cynical.

I have no problem with your belief and hope, that's awesome. But trying to mislabel a rational conclusion based on all known data as "cynical" is... well, rather funny.
We'll see what's funny when it all comes out in the wash.
Old 12th April 2014
  #2948
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Karloff70 View Post
We'll see what's funny when it all comes out in the wash.
Wow, you really have a hard time following... you're missing the point about "funny" but OK.

If it "comes out in the wash" that somehow higher frequencies make a difference, you would be right... but right via magical thinking and wishing, not rational understanding.

I and the rest of the rational world will change our conclusions based on the new science and go happily on our way. When you don't have a belief system you are trying to justify, it's all good.
Old 12th April 2014
  #2949
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by nuthinupmysleeve View Post
Once you see for yourself that there is or is not a difference, eliminating your bias, you can make a decision based on real world data.
j_j would tell you (DID tell you) that you are doing no such thing.

I'm honestly not sure why some audio fans become so impressively singleminded in their dedication to proving the single concept 'you never heard that'. I could tell you a little story, about Eric Sarafin.

He called me up on the phone one day. Was very frustrated and asked me to help him solve a problem. He had audio files recorded on some new converters he was trying to build a project studio around, and he was getting a high frequency squeal that drove him crazy. The manufacturers denied such a noise existed. He insisted it did, but they didn't listen. He sent me the audio files (24 bit) to listen to.

I listened to them, on my Lavry, on speakers, on headphones. Quiet, normally, loud as hell. I never, NEVER heard this noise, any more than the manufacturer did, this noise that Eric ranted about and said was infuriating and obvious to him.

I ran 'Amadeus' and displayed the audio as sonograms, fooling around with the levels until the high frequencies he was complaining about (seemingly FINE) were on the threshold of showing up, the background noise faintly registering and all the actual audio content blowing the graph away.

The noise was there.

It was a persistent tone at exactly 11.025K. I don't know what was causing it. I never had a chance to hear it, and you know me, if I figure I can hear something I'll say it, but THIS was way beyond anything I could register.

Eric found this noise completely maddening. Ever since then, I've known without a shadow of doubt that there are people who hear things WAY better than me.
Old 12th April 2014
  #2950
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by chrisj View Post
Eric found this noise completely maddening. Ever since then, I've known without a shadow of doubt that there are people who hear things WAY better than me.
Thanks for the MM story, that is good! I agree BTW... I have NO DOUBT that others can hear WAY WAY better than I can!

To add.. actually do not doubt that others might be able to hear a difference where I do not, even 24/96 vs 16/44.

However, what is it about a/b/x testing that you disagree with? I suggested it is the only way we know to eliminate expectation bias from the equation. Do you disagree with that?
Old 12th April 2014
  #2951
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by nuthinupmysleeve View Post
Wow, you really have a hard time following... you're missing the point about "funny" but OK.

If it "comes out in the wash" that somehow higher frequencies make a difference, you would be right... but right via magical thinking and wishing, not rational understanding.

I and the rest of the rational world will change our conclusions based on the new science and go happily on our way. When you don't have a belief system you are trying to justify, it's all good.
I have no hard time following my friend. And there seems nothing magical to me in how the information in the higher regions of the signal can be used for the lower, audible spectrum as Chris explained. Or have you a hard time following that?
Old 12th April 2014
  #2952
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Karloff70 View Post
I have no hard time following my friend. And there seems nothing magical to me in how the information in the higher regions of the signal can be used for the lower, audible spectrum as Chris explained. Or have you a hard time following that?
Passive agressive insults. I guess we're done.
Old 12th April 2014
  #2953
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by nuthinupmysleeve View Post
Passive agressive insults. I guess we're done.
Reactions to the same. But yes, we are done.
Old 12th April 2014
  #2954
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by nuthinupmysleeve View Post
Thanks for the MM story, that is good! I agree BTW... I have NO DOUBT that others can hear WAY WAY better than I can!

To add.. actually do not doubt that others might be able to hear a difference where I do not, even 24/96 vs 16/44.

However, what is it about a/b/x testing that you disagree with? I suggested it is the only way we know to eliminate expectation bias from the equation. Do you disagree with that?
You can't get proof out of it, just preponderance of evidence. I think that's what was bugging j_j, everyone involved behaved as if it was proof of this or that or the other, and it's never proof, just probability.

The part of ABX testing that most interests me, you don't have to go along with if you don't want to. I'm certain that audio phenomena are not repeatable in the sense of measurement with instruments (even that's debatable). In my opinion, there are some audio conditions where if you did an infinite series of ABX tests you'd get a median value like 75% confidence. A binary 'there/not there' statement implies a condition that will inevitably trend towards either 50% (guessing) or 100% (perception).

Some of this audio stuff, like 30K frequencies or 24 bit word length, is sometimes 'there' and sometimes not. It's fugitive, and that causes all the trouble. Humans can't nail it down every single time, but respond to the information positively on a less reliable basis. The insistence on making observations repeatable 100% is like pushing for a 'proof' that statistics can't ever give you.

I think a lot of us would be happy with the concession, "I guess some of you guys can hear some of this some of the time. Very well, let's include it in the audio format for those times you do".

What we get is, "Since you can't prove you hear this stuff every single time, it means you hear it none of the time, so it must be taken away from your audio format because it's snake oil and wasteful'.

That's my issue with the ABX tests. People so rarely respond with 'Okay, have your 192K 24 bit moving-average audio, where's the harm in that?'. We get 'ban it, it's fraud!'. I think some of the wackier excesses of high end audiophilia are tantamount to fraud but Pono is certainly not one of those cases.

edit: and try to be kind, folks, please? I know it's hard. Shows we care.
Old 12th April 2014
  #2955
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by chrisj View Post
You can't get proof out of it, just preponderance of evidence. I think that's what was bugging j_j, everyone involved behaved as if it was proof of this or that or the other, and it's never proof, just probability.
Got it... I wasn't suggesting anything was proof though. I was suggesting it was the only way I know to eliminate the impact of confirmation bias.

As far as proof goes, it isn't proof, it's a data point... a point in time. The data point is that this particular person using these particular sample files through this particular audio chain can't tell the difference above random chance on this particular day.

But given additional data points, one can make additional computations of probability. That is what interests me.

Quote:
In my opinion, there are some audio conditions where if you did an infinite series of ABX tests you'd get a median value like 75% confidence. A binary 'there/not there' statement implies a condition that will inevitably trend towards either 50% (guessing) or 100% (perception).
Well, to me, anything significantly above 51% is interesting. I'm not a binary person.

As we established, I think most of us agree the possibility exists for outliers... but those outliers will have data that supports the fact that they are outliers.

Quote:
I think a lot of us would be happy with the concession, "I guess some of you guys can hear some of this some of the time. Very well, let's include it in the audio format for those times you do".
I would absolutely agree with this if the data points showed that to be true. So far though, I don't see it, but I'm open to the possibility.

re: "be kind" I have blocked Karloff so from my part it's over.
Old 12th April 2014
  #2956
Old 12th April 2014
  #2957
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by chrisj View Post
I _ship_ moving average filters in my current line of plugins.

I'm sorry, but none of what I said was controversial or speculative at all.
Well, your description of the moving avg goes against every source I can find, so I would consider that to be, at the very least, controversial.

Here's the venerable Steven Smith on the subject:
http://www.dspguide.com/ch15/3.htm

In short, the moving average is an exceptionally good smoothing filter (the action in the time domain), but an exceptionally bad low-pass filter (the action in the frequency domain).

[Emphasis mine.]

Quote:
It might be good to take the trouble to become MORE of a converter expert . . . I have to know about this stuff because I develop software to deal with the limits of digital audio. Bruno Putzeys sent me email once expressing approval of my 'Contingent Dither' design that ended up becoming 'Spatialize' in the Ditherbox plugin. Do you know who Bruno Putzeys is? I just feel you are not qualified enough to tell people not to listen to me.
Do you consider yourself a converter expert? And what in the world does your email correspondence with Bruno Putzeys about dither have to do with any of this?? Would it matter if I once had lunch with Claude Shannon's niece?

You're right, though: I'm not qualified enough to tell people not to listen to you . . . which is fine, because I never did such a thing. Hopefully an expert will join in and set us both straight.
Old 12th April 2014
  #2958
Gear Head
 

Maybe someone in the future will feel so damn cheated that he or she will drag those lyin sumbitches to court. Then we will have the once and for all ABX test. Right there in the courtroom. Let the jury decide!
Old 12th April 2014
  #2959
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by Silver Sonya View Post
Sly little visit from the threadstarter.....nice.
Old 12th April 2014
  #2960
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by bogosort View Post
Well, your description of the moving avg goes against every source I can find, so I would consider that to be, at the very least, controversial.

Here's the venerable Steven Smith on the subject:
http://www.dspguide.com/ch15/3.htm

In short, the moving average is an exceptionally good smoothing filter (the action in the time domain), but an exceptionally bad low-pass filter (the action in the frequency domain).

[Emphasis mine.]

So....wouldn't that make complete sense then? I mean, like a 'bad' lowpass that is put high up enough to not matter by the high rate and resulting great time domain performance in the audible band.......equaling something rather more vivid??
Old 12th April 2014
  #2961
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by bogosort View Post
Well, your description of the moving avg goes against every source I can find, so I would consider that to be, at the very least, controversial.

Here's the venerable Steven Smith on the subject:
http://www.dspguide.com/ch15/3.htm

In short, the moving average is an exceptionally good smoothing filter (the action in the time domain), but an exceptionally bad low-pass filter (the action in the frequency domain).

[Emphasis mine.]
'Bad' is a strange description. It rolls off slowly, which we want, and it doesn't completely reject those 96K tones… which we can't hear.

Sounds like an optimal choice to me. We're not looking for it to be a brickwall infinite-slope in the first place. It doesn't have to be. Neither you nor Steven Smith have any grounds to claim it's 'exceptionally bad' in terms of how it sounds in practice. It's perfect for the task, which is much more a smoothing filter than a high performance low-pass filter. We don't NEED it to have exceptional rejection at 96K. We already can't hear that, and modest rejection will be more than sufficient to keep playback gear from having intermodulation distortion problems.
Old 12th April 2014
  #2962
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by chrisj View Post
You're free to be of the opinion that the information at 24 bit was already so far below any practical noise floor that it wouldn't matter, but their averaging technique still does have that effect of improving the accuracy of the signal… at audio frequencies, like 2K or 60 hz. It's effectively averaging together two or four samples in the space of one, so the data's being used, it's just not being used to deliver 96K signals.
Sorry, but this seems to be really missing the point. A typical FIR LPF also averages samples, but -- unlike the moving avg filter -- the coefficients are not equal. In effect, the typical FIR LPF is a weighted moving avg. You get optimal time-domain response with the straight moving avg but at the expense of frequency response. It is not more accurate at 2k or 60 Hz or whatever.
Old 12th April 2014
  #2963
Lives for gear
 
3rd Degree's Avatar
 

Quote:
Originally Posted by theblue1 View Post
[bold added]

What Gearslutz were YOU on the last few years? heh

I know I saw a lot of threads on them. But, of course, there weren't as many folks here sticking up for the Beats phones as there are for Pono -- but, then, while I may cringe at the outrageous Pono marketing BS, I still hold hope that the device itself can be a good one, even if I don't buy into quad rates in the slightest. But it's pretty well impossible to convince oneself that the Beats headphone line, with its lightshow earpieces and blingtastic whatnot, has much concern with quality, high fidelity sound.
I was speaking on the response I was talking about, meaning the reflection things like these have, or the responsibilities these companies should or should not have in the music industry. My point was that Pono or Beats are consumer products and are in most ways, unrelated to the music industry. Plenty of people talked about Beats or Dr. Dre in many ways but I don't recall anyone putting responsibility or stating opinions on that person or product in any way that directly reflected how albums should be sold or how markets should be approached (which I think is ludicrous and much of that is reflected in that post).
Old 12th April 2014
  #2964
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by chrisj View Post
You can't get proof out of it, just preponderance of evidence. I think that's what was bugging j_j, everyone involved behaved as if it was proof of this or that or the other, and it's never proof, just probability.
From what I read, you were the only person claiming "proof" with an ABX.

Quote:
In my opinion, there are some audio conditions where if you did an infinite series of ABX tests you'd get a median value like 75% confidence. A binary 'there/not there' statement implies a condition that will inevitably trend towards either 50% (guessing) or 100% (perception).
No one is disputing the mathematics of probability. You keep bringing up "infinite trials" or "million trials" as if anyone is capable of doing that. What's most assured -- and what you seem to ignore -- is that until you run enough trials such that the law of large numbers is in play, you're guaranteed pretty much any result. Your argument that 75% confidence in an ABX is indicative of "fugitive hearing" is only true in the limit. And since we don't have the luxury of such large numbers of trials, we must discount the 75% scores. They literally don't mean anything.
Old 12th April 2014
  #2965
Gear Guru
 

Quote:
Originally Posted by nuthinupmysleeve View Post
It's not at all cynical to conclude that people can't hear the difference... that's exactly what the evidence shows in multiple a/b/x tests in scientific settings. That's not "cynical." That rational.

What is irrational is to believe, in spite of all scientific evidence to the contrary, there is "hope."
what I would love to know is WHY people hope for one thing over another. What's it to them in the long run?

What science has done is take a measurement. The distance from the Earth to the Sun is 92,955,807 miles. Are there people out there "hoping" for a larger number? Or a smaller number? Or do they just accept the measurement?
Old 12th April 2014
  #2966
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by chrisj View Post
'Bad' is a strange description. It rolls off slowly, which we want, and it doesn't completely reject those 96K tones… which we can't hear.

Sounds like an optimal choice to me. We're not looking for it to be a brickwall infinite-slope in the first place. It doesn't have to be. Neither you nor Steven Smith have any grounds to claim it's 'exceptionally bad' in terms of how it sounds in practice. It's perfect for the task, which is much more a smoothing filter than a high performance low-pass filter. We don't NEED it to have exceptional rejection at 96K. We already can't hear that, and modest rejection will be more than sufficient to keep playback gear from having intermodulation distortion problems.
Again, it seems you're missing the point. The reason we use anti-imaging filters is to attenuate the spectral images that form around the sampling frequency (fs), 2fs, 3fs, etc. If we push the sampling frequency so far "to the right" that we don't need brickwall filters in the DAC, then we don't need brickwall filters in the DAC. Doesn't matter if we're using moving avg or windowed sinc functions, if the spectral images are "way up there", then whatever filter we use can be gentle.

This is why all pro DACs are oversampling.
Old 12th April 2014
  #2967
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by joeq View Post
what I would love to know is WHY people hope for one thing over another. What's it to them in the long run?

What science has done is take a measurement. The distance from the Earth to the Sun is 92,955,807 miles. Are there people out there "hoping" for a larger number? Or a smaller number? Or do they just accept the measurement?
Oh, don't be so bloody awkward Joe. Is it really so tricky to relate to hope for a standard of a better sounding delivery format? You don't like the idea of that?
Old 12th April 2014
  #2968
Lives for gear
 
paul brown's Avatar
Quote:
Originally Posted by chrisj View Post
In my opinion, there are some audio conditions where if you did an infinite series of ABX tests you'd get a median value like 75% confidence. A binary 'there/not there' statement implies a condition that will inevitably trend towards either 50% (guessing) or 100% (perception).
being a pedant, how is it possible to start percentile divisions when discussing infinity?

Quote:
Originally Posted by chrisj View Post
Some of this audio stuff, like 30K frequencies or 24 bit word length, is sometimes 'there' and sometimes not. It's fugitive, and that causes all the trouble. Humans can't nail it down every single time, but respond to the information positively on a less reliable basis. The insistence on making observations repeatable 100% is like pushing for a 'proof' that statistics can't ever give you.
but you are insisting that the measured limits of human hearing based on the biological constraints are incorrect. whilst i agree that something might be there, sometimes not, it cannot be a change in the biological make-up of the ear. filter artefacts. yes, i think the differences are audible sometimes under abx conditions of different filters. is it more likely we are hearing the effects of what is a very complex and fundamentally weak link in our audio chain, the transition/change from digital to analogue.

Quote:
Originally Posted by chrisj View Post
I think a lot of us would be happy with the concession, "I guess some of you guys can hear some of this some of the time. Very well, let's include it in the audio format for those times you do".
agreed. but remember it is the sensitivity to filter artefacts that is being heard. if a filter can work closer to perfect at a higher sample rate, then a higher sample rate should be used. i'm interested in this revolutionary filter in the pono. i'm off to investigate more! do you have any links? is it proprietary?

EDIT: found a white paper from AYRE
Old 12th April 2014
  #2969
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by bogosort View Post
Sorry, but this seems to be really missing the point. A typical FIR LPF also averages samples, but -- unlike the moving avg filter -- the coefficients are not equal. In effect, the typical FIR LPF is a weighted moving avg. You get optimal time-domain response with the straight moving avg but at the expense of frequency response. It is not more accurate at 2k or 60 Hz or whatever.
Yikes.

If you're taking samples of a 60 hz signal, and you're taking them at 192K and averaging them together so you output the average of four adjacent samples, it's going to be more accurate than if you had just used the one. This works in scanning, in photography, and certainly in audio. Weighting them might reduce the effect because some samples will dominate, keeping coefficients purely equal will optimize the effect, but as long as there's sampling error in the data point taken at a given sample, averaging will improve accuracy of low-frequency information sampled, and suppress noise.

That's distinct from handling it in normal sampling-theory manner, where you'd get 96K output but would probably end up with more noise than with a 44.1K sampling (a point I think Monty made).
Old 12th April 2014
  #2970
S21
Lives for gear
 
S21's Avatar
 

Quote:
Originally Posted by joeq View Post
Pono does not even exist yet, and already Apple is going to higher resolution:
http://www.macrumors.com/2014/04/10/...sic-downloads/


looks like they have already made that move. According to the article, Apple has been "stockpiling" high resolution masters for a few years now.

Check.
and Mate
Preparing for a possible course of action is not the same as taking that action.

99.9% of users of apple devices are not aware that hires audio exists. 99.9% of users of apple devices are not aware that their DACs and headphone amps are not good enough.

99.9% of users of apple devices care about the music, not the audio. Caring more about the audio than the music is a disease.
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