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
  #2881
Gear Head
 

Quote:
Originally Posted by Karloff70 View Post
Maybe re-read chrisj's twice reposted post and see if that might change your perspective on the possible and the to be assumed.
I don't know. Chrisj knows about those things. I don't. But if those converters in the Pono works as described, does that mean that a 44 kHz recording will sound worse on the Pono than on my iPhone if not upsampled beforehand?
Old 12th April 2014
  #2882
Quote:
Originally Posted by SmoothVibe View Post
the impact Pono will have on music will be a relatively big one.
dream on…..
Old 12th April 2014
  #2883
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by H.E. View Post
I don't know. Chrisj knows about those things. I don't. But if those converters in the Pono works as described, does that mean that a 44 kHz recording will sound worse on the Pono than on my iPhone if not upsampled beforehand?
I don't know either. To me it merely means there is a possibility that this thing could actually sound better than a standard converter run at high rates.
Old 12th April 2014
  #2884
Gear Guru
 
Sounds Great's Avatar
 

Quote:
Originally Posted by mixmixmix View Post
Artists who endorsed Pono don't realize how phony they look praising this gadget. It is painfully obvious they just want to get paid / stay in the public eye etc.

Rock 'n' Roll started out as an act of rebellion. Today those corporate *****s will do anything to get a buck. And Neil Young is nothing but a pimp. Digital age pimp, hawking Pono to naive general public.

He use to write great songs. What happened to him? Did not he make enough money?

Pfff.

Green monster got you in a cage. Open the door my friend, there's no lock on that door.
Old 12th April 2014
  #2885
Gear Guru
 
Sounds Great's Avatar
 

Quote:
Originally Posted by Karloff70 View Post
Maybe re-read chrisj's twice reposted post and see if that might change your perspective on the possible and the to be assumed.
Everyone should read chrisj's post because he said it well.

And then stop arguing about high frequencies and 100's of db of dynamic range. If everyone would agree that we only need about 20-20k, and about 80 db of range (at the extreme), the discussion could stay on track about the part we actually hear. Dog whistles indeed.
Old 12th April 2014
  #2886
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Karloff70 View Post
I don't know either. To me it merely means there is a possibility that this thing could actually sound better than a standard converter run at high rates.
I can predict what that means within the context of 'only around 20K is perceptible, really, in normal use'.

A standard converter run at 44.1K or thereabouts produces ripple and overshoot as part of the reconstruction, and this is the thing Pono gets to compare against (notably, this is still true for mp3s but the extreme highs are trashed because you can't get the bandwidth to accurately track all the data. Extreme highs off mp3 are ALWAYS a rough approximation like how even moderate highs are an approximation at 96K and under. There isn't the data rate to reproduce it even at 320K but people don't resolve such high frequencies as cleanly as they resolve 2K.) So, all normal audio listening, mp3 or wav, is typically using a brickwall filter at 22.050K of known behavior.

A standard converter at high rates will still have the same behavior but 'pitched up', which I think will help make it much harder to notice as a problem. Also the input data simply isn't there to drive the reconstruction that hard: it will overshoot a lot less, because there just isn't a bunch of audio content at 40K under any circumstances. Possibly synthetic sounds?

There are DACs out there that don't use a reconstruction filter. They literally output stairsteps, but no overshoot or ripple. Pono COULD do that, or provide it as an option: it doesn't mean you wipe out 20K content, but it rolls off faster and it's 'coarser', not really technically correct and from what I've read there's a type of graininess you get off it. But it totally lacks the 'glare' of the stressed-out reconstruction filter, which is very likely a factor of overshoots and pre-ring, something we know happens.

There's DSD/SACD. That delivers really high resolution the lower in frequency you go: it's not restricted to 24 bit PCM, it continues to get more accurate as frequency drops. The downside is that, beyond the range of hearing, it disintegrates into really high noise and chaos that has to get filtered out in other ways (and it's uneditable). Monty (xiphmont)'s objections having to do with IM distortion are way more relevant here, DSD's design almost guarantees playback hardware will be stressed out, and this can translate into listening fatigue when the >20K is such a war-zone of noise and filtering.

The Pono moving-average from 96K or 192K is ingenious. It's going to slightly help out resolution in the passband because averaging the sample values of low frequency information gets you closer to the real value by suppressing noise. It's NOT actually getting you 96K frequencies to enjoy, but we're all pretty much in agreement that this is not the point. It's giving you the frequencies around 20K and up essentially unattenuated (especially for 192K: at 96K I think moving average will get you all the smoothness you want but just a hint of roll-off. surface of water, anyone?) and it is transitioning from passband to stopband with great smoothness: significantly better transition using 192K.

This will apply to upsampled 44.1K material. If you upsample with good SRC you get a representation of your audio that does not itself ripple or pre-ring: it's a technique fancy boutique converters have used for many years. I don't think it'll sound anything like '20,000 Leagues Under The Sea'. I think it'll sound like more natural-sounding, unfatiguing CD highs, with exactly as much 'something lacking in the 30K' as would normally be heard, but highlighted by a total lack of unwanted ripple/overshoot that might otherwise suggest super-highs.

I think without the ripple and overshoot, it'll be significantly easier to pick out something lacking in stuff brickwalled at 22K, and this is why the ABX was too 'easy' at high volumes when it was 'supposed' to be impossible. The test bandpass supplied was exactly what we're talking about: upsampled, with no overshoot or ripple, and the upsampled one lacking >22K was noticeably duller. This is probably what Pono will do. Playing back at 44.1K without upsampling would have worked the reconstruction filter harder and made it tougher to tell a difference, but the increased treble energy isn't desirable, it's pre-ring and overshoot and is only compensating for loss of the real energy… so the 'underwater' of upsampling and playing back at double or quad rate through THOSE filters is actually desirable, tonally, even though it seems to give you a 'dull' sound compared to the cruder approach of direct 44.1K brickwalled playback…

Cliff notes: once you get rid of reconstruction filter problems, 44.1K IS dull compared to 96K or 192K. But that's because the ringy filter is part of what fills in the sonic gap for listeners, they can't really pick out that it's unwanted energy unlike the real input sound, yet it's subtly annoying in a way the Pono filters cannot be.
Old 12th April 2014
  #2887
Lives for gear
Quote:
Originally Posted by Sounds Great View Post
Everyone should read chrisj's post because he said it well.

And then stop arguing about high frequencies and 100's of db of dynamic range. If everyone would agree that we only need about 20-20k, and about 80 db of range (at the extreme), the discussion could stay on track about the part we actually hear. Dog whistles indeed.
Quick conclusions there cowboy. When we align to higher and higher resonance levels, our DNA shifts as well.
Old 12th April 2014
  #2888
Lives for gear
 
Ephi82's Avatar
 

Quote:
Originally Posted by chrisso View Post
I strongly disagree.
It means nothing to the main consumer demographic. Tech nerds and musicians might care about the ultimate sound quality, but then we don't need to attract that demographic back to buying music. We need to attract back the average 16 to 40 year olds.
Not only does Pono do nothing in that regard, it makes us look stupid.
We need to get young people buying new music again, in a way that nurtures and supports new music. The way to health in our industry is innovation and new blood. To propose charging premium prices for albums recorded 20 to 40 years ago is going to change anything - is actually a disaster and could make you angry if you thought anyone was listening. But thankfully I think the vast majority of potential music customers neither know or care about Pono.
Chrisso,

I think PONO has the potential to become relevant to people buying new music.

Although there are other "Hi Res" PORTABLE audio players out there, they have not been marketed broadly enough and the value proposition hasn't been defined very well.

Having a high profile "guru" like Neil Young is important to get people to at least listen to the "story". The story is a long one, and complex, so its not going to be easy.

I agree with several posters that PONO is making a marketing mistake by focusing too much on the sample rates and not enough on the quality of conversion and analog stages in the PONO device.

They also have to educate the consumer that brickwalled mastered recordings are not going to benefit from the format or the hardware. (even if they are 192/24) Listening stations in retail stores might have to be the way?

Having written all that, it seems like an very uphill battle.

However, I would think that most artists and audio engineers would love to know that there is a market for carefully mixed/mastered material with wide dynamic range. In addition, satisfaction in knowing that their work is being heard through equipment where the listener HEARs all that was intended.
Old 12th April 2014
  #2889
Lives for gear
 

Quote:
Originally Posted by Ephi82 View Post
Chrisso,

I think PONO has the potential to become relevant to people buying new music.

Although there are other "Hi Res" PORTABLE audio players out there, they have not been marketed broadly enough and the value proposition hasn't been defined very well.

Having a high profile "guru" like Neil Young is important to get people to at least listen to the "story". The story is a long one, and complex, so its not going to be easy.

I agree with several posters that PONO is making a marketing mistake by focusing too much on the sample rates and not enough on the quality of conversion and analog stages in the PONO device.

They also have to educate the consumer that brickwalled mastered recordings are not going to benefit from the format or the hardware. (even if they are 192/24) Listening stations in retail stores might have to be the way?

Having written all that, it seems like an very uphill battle.

However, I would think that most artists and audio engineers would love to know that there is a market for carefully mixed/mastered material with wide dynamic range. In addition, satisfaction in knowing that their work is being heard through equipment where the listener HEARs all that was intended.
Its obvious that this is targeted towards the consumer market. As has been pointed out, there are already hi res players that exist for audiophiles. This will be mass consumers first exposure to Hi res.
Old 12th April 2014
  #2890
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Ephi82 View Post
They also have to educate the consumer that brickwalled mastered recordings are not going to benefit from the format or the hardware. (even if they are 192/24) Listening stations in retail stores might have to be the way?
There is no such thing as a brickwalled master. Brickwall is strictly about the reconstruction filter which is part of the hardware, not even the format. So, any high resolution OR normal resolution digital recording could benefit from the hardware.

Again, this is very old news in audiophilia: upsampling and then using a different reconstruction approach is commonplace in converters that cost about as much as a house. The difference is, this is priced to be consumer-accessible, but it's very much the same thing.

Any and all 96K or 192K sources will work just fine no matter how they were produced (the mastering doesn't apply the brickwall filter at all). Upsampled stuff will probably come off as a bit 'mellow' compared to high sample rates as the brickwall will no longer be hyping the highs and distorting them. I think 'underwater' is putting it a bit strong, but hey, at least it's 'liquid', right? (and there are gains to be had in the averaging, for mids and bass)
Old 12th April 2014
  #2891
Motown legend
 
Bob Olhsson's Avatar
 

The real test will be how many show up immediately on E-Bay.
Old 12th April 2014
  #2892
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Bob Olhsson View Post
The real test will be how many show up immediately on E-Bay.
The real test will be whether they command higher prices on eBay for being collectible!
Old 12th April 2014
  #2893
Lives for gear
 
Ephi82's Avatar
 

Quote:
Originally Posted by chrisj View Post
There is no such thing as a brickwalled master. Brickwall is strictly about the reconstruction filter which is part of the hardware, not even the format. So, any high resolution OR normal resolution digital recording could benefit from the hardware.

Again, this is very old news in audiophilia: upsampling and then using a different reconstruction approach is commonplace in converters that cost about as much as a house. The difference is, this is priced to be consumer-accessible, but it's very much the same thing.

Any and all 96K or 192K sources will work just fine no matter how they were produced (the mastering doesn't apply the brickwall filter at all). Upsampled stuff will probably come off as a bit 'mellow' compared to high sample rates as the brickwall will no longer be hyping the highs and distorting them. I think 'underwater' is putting it a bit strong, but hey, at least it's 'liquid', right? (and there are gains to be had in the averaging, for mids and bass)
Perhaps I misunderstand, but when I said brickwalled I meant the limiting technique used in the mastering stage for maximum loudness. From what I hear in my small collection of SACD and DVD audio discs, some are a pleasure to listen to and fine details (like reverb tails) can easily be heard and appreciated. Others, clearly mastered with little dynamic range, don't benefit at all from being at 48/24 or 88 or 96/24 as compared to mp3 and cd.
Old 12th April 2014
  #2894
S21
Lives for gear
 
S21's Avatar
 

iTunes won't go to a higher resolution unless forced by a product that has made serious inroads. That product would need to be shipping millions of units each quarter to scare Apple.

Look, providing hires content would be validating a competitor's position. Hires content would not be hard to provide from iTunes. This is just a move on the chessboard that Apple don't need to make and won't make.
Old 12th April 2014
  #2895
Lives for gear
 
Ephi82's Avatar
 

Quote:
Originally Posted by Bob Olhsson View Post
The real test will be how many show up immediately on E-Bay.
Hey Bob,

I've meant to ask you this before. Have you ever heard the remaster (for cd, and DVD audio (surround and stereo) called The Marvin Gaye Collection that came out in 2001?

If so, what's your opinion of it?

I think the earlier recordings (the stuff with Tammi Terrell) lose all that great mono mellowness, but the later stuff (What's Going On, Mercy, mercy Me) are revealing. The vocals have lots of room in the soundstage and the bass is just incredible (I can hear lots of nuances)
Old 12th April 2014
  #2896
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Ephi82 View Post
Perhaps I misunderstand, but when I said brickwalled I meant the limiting technique used in the mastering stage for maximum loudness. From what I hear in my small collection of SACD and DVD audio discs, some are a pleasure to listen to and fine details (like reverb tails) can easily be heard and appreciated. Others, clearly mastered with little dynamic range, don't benefit at all from being at 48/24 or 88 or 96/24 as compared to mp3 and cd.
Oh! Sorry, I misunderstood with all our talk of filters.

Yes, a lot will be lost if people just go for massive overlimiting and clipping. One thing that'll happen is that obnoxious stuff will come out SLIGHTLY more listenable on Pono simply because that's the stuff that stresses the reconstruction filters worst, so there's a layer of analog clipping (check on SSL's X-ISM meter to see) which will be totally removed. I think without it that stuff will feel pretty oppressive, though: overloud but without that artificial 'sparkle' to make it seem like there's more energy beyond the overloud.

I'm trying to push an approach to mixing/mastering with much more headroom, specifically for Pono, on the grounds that 'this is different' is a GREAT excuse to overhaul that expectation. Actually have a plugin out which is getting some real momentum, that resists over-loudness by breaking up when the RMS is pushed too hard (peak energy is of course fine: it's a little tricky to work with).

We don't have the power to stop the industry (that is, US) from competing with masters louder than the next guy. What we do have is a new context designed to promote attentive listening to audio on its own terms, and I think it does open up a huge opportunity. Basically anything you do and label 'mastered/mixed FOR PONO' gets to be examined outside the context of the existing industry.

People could run their albums through a ProCo RAT and call that 'mastered for PONO' if they want, to see if that flies. But given the known characteristics of what Pono's gonna be and the limits (basically 'none') of the digital audio it'll carry, it's a big opportunity to frame mixes shockingly less loud than usual, not just 'slightly' less loud. Do it so exaggeratedly that people cannot possibly consider them the same thing, and they'll turn up the 'new for PONO' mix and hear it as intended. Half-ass it, and you lose the shock of the new. I can't overemphasize the importance of presenting naive listeners with a BIG difference on every level, and leaving (and using) tons of headroom will help sell the idea that it's a new and different thing.
Old 12th April 2014
  #2897
Quote:
Originally Posted by michael cleary View Post
This will be mass consumers first exposure to Hi res.
I can't see that happening when they are charging full blown CD retail when mass consumers are used to Spotify, illegally free and (at best) $2 songs from iTunes.
Old 12th April 2014
  #2898
Lives for gear
 
Ephi82's Avatar
 

Chrisj, "mastered" or "made for" Pono is exactly what is needed, and what must be marketed.

It's why PONO believes it must be in the music distribution business. (Aside from the money! No one will appreciate their player if they don't have great stuff to play on it.

You know, that's why they should load every new. PONO with a famous Neil Young song, maximum volume mastered mp3 and cd versions, and then a 70 style era mastering at 96/24 of the same song, just so that the consumer, and most important, all his or her friends, can hear the difference.

Neil, I'll email you my address so you can send the check!
Old 12th April 2014
  #2899
Lives for gear
 
Ephi82's Avatar
 

Quote:
Originally Posted by chrisso View Post
I can't see that happening when they are charging full blown CD retail when mass consumers are used to Spotify, illegally free and (at best) $2 songs from iTunes.
I disagree. If people believe it sounds better, they will buy it at a modest premium.

Obviously, free is hard to beat, but a lot of that stuff is lossy and compressed to heck
Old 12th April 2014
  #2900
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by S21 View Post
iTunes won't go to a higher resolution unless forced by a product that as made serious inroads. That product would need to be shipping millions of units each quarter to scare Apple.

Look, providing hires content would be validating a competitor's position. Hires content would not be hard to provide from iTunes. This is just a move on the chessboard that Apple don't need to make and won't make.
Why 'scare' Apple?? How about 'inspire' Apple to use their already great network/iTunes to distribute a new product more efficiently than anyone else can? Money to be made, forget fear. Besides, if anything Apple being Apple they can't afford to look behind the curve, so not much choice but forward. They will need to make it, or risk looking behind, and THAT would scare Apple.
Old 12th April 2014
  #2901
tkr
Gear Addict
Quote:
Originally Posted by chrisj View Post
There is no such thing as a brickwalled master. Brickwall is strictly about the reconstruction filter which is part of the hardware, not even the format. So, any high resolution OR normal resolution digital recording could benefit from the hardware.

Again, this is very old news in audiophilia: upsampling and then using a different reconstruction approach is commonplace in converters that cost about as much as a house. The difference is, this is priced to be consumer-accessible, but it's very much the same thing.

Any and all 96K or 192K sources will work just fine no matter how they were produced (the mastering doesn't apply the brickwall filter at all). Upsampled stuff will probably come off as a bit 'mellow' compared to high sample rates as the brickwall will no longer be hyping the highs and distorting them. I think 'underwater' is putting it a bit strong, but hey, at least it's 'liquid', right? (and there are gains to be had in the averaging, for mids and bass)
I have been reading up on reconstruction filters, and must admit that I find it a pretty tough read. After a couple of hours I seem to have gotten the gist of it, but there is one big question in my head:
What is the difference between what Pono (supposedly) does and simple upsampling of for example a 16/44.1 signal to 24/96 before d/a conversion? Is it the reconstruction filter design that is so special?
Old 12th April 2014
  #2902
Gear Guru
 
Karloff70's Avatar
 

Quote:
Originally Posted by chrisso View Post
I can't see that happening when they are charging full blown CD retail when mass consumers are used to Spotify, illegally free and (at best) $2 songs from iTunes.
Why would that stop people? They will torrent the crap out of the hires stuff if they realise they like it better instead of paying CD retail prices. Maybe that will drive the price down over time.

But it will get really interesting once this stuff can be streamed. Then we would 'only' have to solve paying content makers properly.....lol
Old 12th April 2014
  #2903
Quote:
Originally Posted by mixmixmix View Post
... What happened to him? ...
Sadly, it's the nature of the beast for people to go through youthful idealistic phases and then if they're unlucky enough to be successful, they retreat into a self-deluded state of what we'd define as "comfort" or more pointedly "corruption." It would take a really dedicated, almost heroic effort to resist this.
Old 12th April 2014
  #2904
Gear Guru
 

Quote:
Originally Posted by S21 View Post
iTunes won't go to a higher resolution unless forced by a product that as made serious inroads. That product would need to be shipping millions of units each quarter to scare Apple.
Pono does not even exist yet, and already Apple is going to higher resolution:
http://www.macrumors.com/2014/04/10/...sic-downloads/

Quote:

Look, providing hires content would be validating a competitor's position. Hires content would not be hard to provide from iTunes. This is just a move on the chessboard that Apple don't need to make and won't make.
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
Old 12th April 2014
  #2905
Lives for gear
 
bogosort's Avatar
Quote:
Originally Posted by chrisj View Post
Not only that, they're also executing the double and quad sample rates with a very different reconstruction filter… a moving average.
Do you have a cite for this? How much detail did they provide? Moving average seems like a terrible choice for an anti-imaging filter.

Quote:
The truly funny thing about this is, it COMPLETELY throws away the super-high frequencies we're all talking about
Incorrect. A moving average filter can be thought of like a backwards LPF: if we define a LPF as the convolution of the input with a sinc function, we see that its frequency response is a rectangle. A moving average has this backwards: it is the convolution of the input with a rectangle, so its frequency response is a sinc function. But a sinc function rings forever!

Quote:
The people shown in their videos clearly have never heard high sample rate digital audio used to extend the rejection band short of Nyquist. They have always listened to stuff with Nyquist at 22.050K and a really steep brickwall filter that causes ripple, pre-echo and overshoots.
You think all those people in the video have never heard 96k (Nyquist at 48k) sampling rates in the studio? In any case, for the last decade at least, pro converters use oversampling, which leads to gentle analog filters. The heavy lifting is done by digital filters (with non-equal coefficients), where you can specify the passband ripple, overshoot, etc.

Quote:
Hell, if they're doing different degrees of moving average for 96K and 192K, they may be throwing away everything above 20K identically no matter which rate you pick. Or, 30K, something else much more reasonable. I'd love to know how many FIR taps they're using for their moving average, because we could calculate exactly where they're rolling off.
Look up the frequency response of a moving avg filter (equal coefficients). They don't simply roll off.
Old 12th April 2014
  #2906
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by michael cleary View Post
Its obvious that this is targeted towards the consumer market. As has been pointed out, there are already hi res players that exist for audiophiles. This will be mass consumers first exposure to Hi res.
I agree, there will be. Then... what happens when people realize they can't tell the difference between 24/96 and 320kbps mp3? Confirmation bias only gets you so far.

This is the scenario many of us are worried about and say over and over. Some say "well at least they are trying to improve audio quality for the consumer"... but if there is no real meaningful audio quality increase, people will only be fooled for a short time. Then they will feel taken advantage of.

We all know how many times many of us have bought the same content. Record companies don't have the greatest reputation as it is... trying to market people into buying yet ANOTHER format, when, guess what, there is no major audible benefit?

Every previous generation of music had a real benefit. LPs sounded better than 78's and had more music. 8-tracks were the first alternative for cars. Cassettes were another alternative to LPs for people to listen in their car, but offering FF and REW capability, plus doing away with the problem of 8-track stretching. CDs had improved audio quality over all previous formats and added a portability, and worked in cars (though with some skipping in the early players). MP3 came with immense increase in convenience, initially with lower quality but MUCH higher convenience. As MP3 improved it is all but indistinguishable at 320kbps from a CD for most people.

What is the advantage of Pono? Getting rid of "underwater" listening? It's based on confirmation bias.

One thing all sane people will agree with... there is no "night and day" sonic benefit. If there is a benefit, it's extremely subtle. Most people don't have the playback systems to be able to pick up subtle differences.

So this all adds up to the reasons why many of us who love music think Pono is a bad idea. It adds more fuel to the fire of evil record companies and their greedy plans to sell us the same old music again because they can't find any good NEW music to sell to us. That IS what people think in many circles. It's not smart IMHO. You can fool some people some of the time...
Old 12th April 2014
  #2907
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by chrisj View Post
There's DSD/SACD. That delivers really high resolution the lower in frequency you go: it's not restricted to 24 bit PCM, it continues to get more accurate as frequency drops.
SACD IS 24 bit PCM. DSD is not. You are making a big mistake combining them into one bullet point.
Old 12th April 2014
  #2908
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by joelpatterson View Post
Sadly, it's the nature of the beast for people to go through youthful idealistic phases and then if they're unlucky enough to be successful, they retreat into a self-deluded state of what we'd define as "comfort" or more pointedly "corruption." It would take a really dedicated, almost heroic effort to resist this.
I think that is a very wise view. It's very difficult to resist the temptation to read and believe your own press.
Old 12th April 2014
  #2909
Lives for gear
 
nuthinupmysleeve's Avatar
 

Quote:
Originally Posted by Ephi82 View Post
I disagree. If people believe it sounds better, they will buy it at a modest premium.

Obviously, free is hard to beat, but a lot of that stuff is lossy and compressed to heck
Hipsters and the equivalent audiophiles might buy it. Remember that is a very small percentage of music consumers. The majority of music is still bought/streamed by younger people and they don't own turntables and don't really care much about audio quality.

My daughter loves music and she can hear the difference... but most often she listens from her computer speakers even though I bought her a decent external dock that she can plug into. I sometimes have to walk into her room and plug the speakers in because I can't stand it a couple rooms away. THAT is what you are up against if you think people will buy it. Sure, some will (obviously from kickstarter). But the masses?
Old 12th April 2014
  #2910
Lives for gear
 
bogosort's Avatar
I'm not an expert on converter technology, and I don't believe chrisj is either. But for balance I want to point out that almost everything chrisj says in this post seems controversial to me. Due to the technical nature of the subject, I don't think most readers will understand what he's talking about and so may take his conclusions at face value.

It'd be great if a true expert could weigh in here.

Quote:
Originally Posted by chrisj View Post
I can predict what that means within the context of 'only around 20K is perceptible, really, in normal use'.

A standard converter run at 44.1K or thereabouts produces ripple and overshoot as part of the reconstruction, and this is the thing Pono gets to compare against (notably, this is still true for mp3s but the extreme highs are trashed because you can't get the bandwidth to accurately track all the data. Extreme highs off mp3 are ALWAYS a rough approximation like how even moderate highs are an approximation at 96K and under. There isn't the data rate to reproduce it even at 320K but people don't resolve such high frequencies as cleanly as they resolve 2K.) So, all normal audio listening, mp3 or wav, is typically using a brickwall filter at 22.050K of known behavior.

A standard converter at high rates will still have the same behavior but 'pitched up', which I think will help make it much harder to notice as a problem. Also the input data simply isn't there to drive the reconstruction that hard: it will overshoot a lot less, because there just isn't a bunch of audio content at 40K under any circumstances. Possibly synthetic sounds?

There are DACs out there that don't use a reconstruction filter. They literally output stairsteps, but no overshoot or ripple. Pono COULD do that, or provide it as an option: it doesn't mean you wipe out 20K content, but it rolls off faster and it's 'coarser', not really technically correct and from what I've read there's a type of graininess you get off it. But it totally lacks the 'glare' of the stressed-out reconstruction filter, which is very likely a factor of overshoots and pre-ring, something we know happens.

There's DSD/SACD. That delivers really high resolution the lower in frequency you go: it's not restricted to 24 bit PCM, it continues to get more accurate as frequency drops. The downside is that, beyond the range of hearing, it disintegrates into really high noise and chaos that has to get filtered out in other ways (and it's uneditable). Monty (xiphmont)'s objections having to do with IM distortion are way more relevant here, DSD's design almost guarantees playback hardware will be stressed out, and this can translate into listening fatigue when the >20K is such a war-zone of noise and filtering.

The Pono moving-average from 96K or 192K is ingenious. It's going to slightly help out resolution in the passband because averaging the sample values of low frequency information gets you closer to the real value by suppressing noise. It's NOT actually getting you 96K frequencies to enjoy, but we're all pretty much in agreement that this is not the point. It's giving you the frequencies around 20K and up essentially unattenuated (especially for 192K: at 96K I think moving average will get you all the smoothness you want but just a hint of roll-off. surface of water, anyone?) and it is transitioning from passband to stopband with great smoothness: significantly better transition using 192K.

This will apply to upsampled 44.1K material. If you upsample with good SRC you get a representation of your audio that does not itself ripple or pre-ring: it's a technique fancy boutique converters have used for many years. I don't think it'll sound anything like '20,000 Leagues Under The Sea'. I think it'll sound like more natural-sounding, unfatiguing CD highs, with exactly as much 'something lacking in the 30K' as would normally be heard, but highlighted by a total lack of unwanted ripple/overshoot that might otherwise suggest super-highs.

I think without the ripple and overshoot, it'll be significantly easier to pick out something lacking in stuff brickwalled at 22K, and this is why the ABX was too 'easy' at high volumes when it was 'supposed' to be impossible. The test bandpass supplied was exactly what we're talking about: upsampled, with no overshoot or ripple, and the upsampled one lacking >22K was noticeably duller. This is probably what Pono will do. Playing back at 44.1K without upsampling would have worked the reconstruction filter harder and made it tougher to tell a difference, but the increased treble energy isn't desirable, it's pre-ring and overshoot and is only compensating for loss of the real energy… so the 'underwater' of upsampling and playing back at double or quad rate through THOSE filters is actually desirable, tonally, even though it seems to give you a 'dull' sound compared to the cruder approach of direct 44.1K brickwalled playback…

Cliff notes: once you get rid of reconstruction filter problems, 44.1K IS dull compared to 96K or 192K. But that's because the ringy filter is part of what fills in the sonic gap for listeners, they can't really pick out that it's unwanted energy unlike the real input sound, yet it's subtly annoying in a way the Pono filters cannot be.
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