Neil Young was working with Apple on super high-def music format - Page 3

I vaguely remember us having this exact argument months, years? ago

Ignorant statements.

Given that he was an old hippie, it isn't all that surprising to me. The reason would have been partly that he no doubt had a huge collection of vinyl. But the fact that he listened to vinyl is no indication that he had any misgivings about the superiority of digital recording. It is exceedingly unlikely that he would have believed that vinyl recordings were inherently superior to digital recordings given a good encoding scheme and adequate bitrate.

It seems $10 vodka leads to deafness.

Yes so do I. It was just over a year ago. Wasn't that where you were totaly pwned by the Mod, Mr H, when he informed you his electrical engineering Phd topic was to do with audio amplifiers?

So one year on, and having pointed lots of people to the compressed/uncompressed file, no one has been able to distinguish between the sections, even when those people tell me the difference even between 320 kbs mp3 and CD is night and day; chalk and cheese. For some strange reason, their powers of discrimination seem to evaporate the instant they listen to that file.

What leads to blindness?

But, your argument is specious. As for the headphones, I mostly listen through a pair of Sennheiser HD580. As for the particular Sennheiser headphones in the post to which you replied, that pair happens to be a DJ model and not one of their best. The closed back has the single advantage of noise isolation, to the detriment of sound quality. The air spring effect can be used to good effect if and only if the driver itself is over damped so as to avoid a high Q resonance. And even in that case, there will still be a cavity resonance, i.e., standing waves set up between the back side of the diaphragm and the wall in back. These are the reasons that Sennheiser's better headphones always have been open back and always will be open back.

But as for your comments, the analogy to display resolution and bit depth and so on is always one that is easy to make. But it and of itself it does not prove anything at all. And the comparison with imaging is a bogus comparison for a fundamental reason. With any image, it is always potentially possible to display it on a display with greater pixel resolution, and for this reason there is always a potential advantage for using greater quantity of pixels in the image file. But bit depth is another matter. It translates into the amount of fine variation in brightness, hue, and saturation. There is inherently a limit to the ability of human vision to detect these differences. To keep it simple, consider the case of grey scale. Initially as you increase the bit depth, the brightness of the reproduced image gets closer and closer to the original, i.e., is neither whiter nor blacker than the original. But at a certain point, the human eye simply can no longer perceive the difference. Double the bit depth and scan and encode again, and you cannot tell any difference at all between that copy and the previous one, or between either and the original. Common sense tells you that eventually this will happen. It is not a question of whether it will happen. It is only a question of what the bit depth has to be, in order for this to happen. And once you have reached that point and are entirely certain that you have reached that point, there is absolutely no reason to increase the bit depth of the scan any further. It is the same with audio encoding, and even with perceptual encoding.

The question of whether perceptual encoding can be indistinguishable from the original is a moot question. The only question that is even worth considering is what amount of compression, given a specific encoding scheme, can be tolerated without introducing some artifact by which any listener would be able to hear any difference between that recording and the original. IF you are entirely certain that the bit rate that you have used is perfectly adequate such that no person could every detect any difference between that recording and a non-lossy recording with arbitrarily high quantization rate and word size, then there is no discernible reason to use a higher bit rate. Because, IF it is true that no one can hear then difference, THEN it is true that no one can hear the difference. The only meaningful, valid questions are what people can and cannot hear. To dismiss all perceptual coding techniques in the manner that manner people do is equivalent to asserting that it is not possible, using a perceptual coding technique, to make a recording that no person would be able to recognize as different from a master using arbitrarily high quantization rate and word size. It is manifestly ludicrous to suggest that this would be the case, yet this is precisely what people are in effect asserting when they criticize perceptual encoding categorically. It is logically preposterous.

He/she may have wasted money spending that much on a stereo system, but if so, it would be entirely the same if they had spent that money and then used some other source.

It is logically, manifestly absurd to suggest that when the bit rate is adequate, that perceptual encoding is inherently inferior, i.e., cannot be indistinguishable from the original or from a master using arbitrarily high quantization rate and word size (sample size). It is utterly ludicrous to suggest otherwise. Common sense should tell you that as long as the bit rate is adequately, that no Human being would ever be able to hear any difference whatsoever. What you assert however in effect, is that no matter how high the bit rate, that there will always be some artifact that someone will be able to hear. This is what you assert in effect, and it is simply ludicrous. In fact, anyone who has any technical comprehension whatsoever of these matters knows perfectly well that these claims with respect to perceptual encoding are logically absurd. Only people who lack the technical comprehension (and the common sense) are prone to disparage perceptual encoding categorically in the manner that you do.

There is also the fact that when the question was seriously undertaken in the form of a scientific study conducted over a period of a couple of years, that the conclusion was that no person could be found who could demonstrate the ability to reliable distinguish between standard CD and the so-called high resolution formats. The companies who backed those formats knew this before they introduced them, and they introduced them only because there appeared to be a market for them. That is not to say that you do not own any high-resolution discs that sound any better than standard CDs typically sound. Those discs that sound better were either recorded originally using better equipment, or else they were remastered in some way that corrected for some limitation that was present in the prior CD release of the same material.

I keep trying to say something that seems to me that should be obvious to everyone, but that for some reason is not. It should be obvious to everyone who has any sense that the question of whether perceptual encoding is inherently flawed is moot. It should be obvious that as long as it is done carefully and with adequately high bit rate, that no person would be able to hear any difference whatsoever between that recording and a master that was recorded with linear PCM using arbitrarily high sample rate and sample size. This ought to be obvious, to everyone. Yet, there are still lots of people who for some inexplicable reason assert, explicitly or implicitly, that there is something inherently wrong with perceptual encoding. This is effectively what they assert when they argue that only non-lossy compression is acceptable. Some people just cannot tolerate the idea that some of the music is being discarded, and for this reason they insist on something that logically is ludicrous: that even though you cannot hear any difference whatsoever between two recordings, that one still sounds better than the other.

I'm sure Apple will become very interested in selling us on the benefits of music formats with larger file sizes when they want us to upgrade from 1TB SSD iPhones and Macbook Airs to 10TB SSD iPhones and MacBook Airs.

I don't have the Pat Methany that you referred to, but I am very interested in the question of whether 128 kbps AAC avoids artifacts that I can hear. My CD collection is also encoded at that rate. I have encoded a few CDs at higher rate, and sometimes thought I could hear a difference, but after a careful comparison, eventually concluded that I could not hear any difference. If there are any other examples that you have, perhaps in classic rock or mainstream classical, I would be interested in knowing.

As for improvements in cost/quality ratio over the past twenty years, I'm not so inclined to agree, but partly this is due to inflation. Twenty years ago we are talking 1990. Just about any audio amplifier of average quality sounded the same as any other except when driven into non-linearity at a level with a more expensive amp with greater power capability would remain linear. And I can't tell that the price has come down, unless perhaps you take inflation into account and compare in terms of 1990 dollars. I'm not so sure about that. And as for speakers, good speakers seem more expensive today than ever. I don't think my hearing has improved, but when I listen to new speakers nowadays, I often hear some obvious coloration and I just don't recall that so much in earlier decades. Memory is a factor there of course, but in the 70s and 80s there were a lot of good sounding speakers that did not cost over a grand for a pair. Nowadays it seems that you cannot find a good sounding pair of speaker for less than about that much per pair. There are lots of really, really good speakers, but most of them seem to be very expensive, upwards of two thousand per pair. (Not to suggest that all speakers that cost that much are good ...)

One area where cost has certainly come down is with CD players. Of course nowadays most disc players are multi-format players. But most cheap DVD players nowadays, including many that sell for perhaps as little as $100, are probably sonically indistinguishable from the most expensive CD players. But even in 1990, the typical CD player costing a couple hundred bucks was probably not sonically distinguishable from the most expensive CD player.

Now, as for "compression". One of the problems with this word is that it means two completely different things, depending on the context. When you talk about compression in the context of FM radio broadcast, you are almost certainly talking about compression of dynamic range, and probably with respect to low-frequency content in particular. But in the context of the discussion here, the word "compression" refers to file compression, i.e., reduction in the number of bits used to record each second of music. It is a completely different thing. Of course there is the possibility that sonically there will be similarity, but this is not something that ought to be taken for granted, and it is desirable to avoid confusing the two meanings of the word "compression".

My favorite post so far in this discussion!

When you did that casual listening test between the two amps, the absolute most you proved is that the two amps sounded different. It seems unlikely that they would, and I am inclined to think that the two of you tricked yourselves into that conclusion, but even if they really did sound different, that in itself is not proof of the superior quality of one over the other. It is a well-known fact that many people prefer the different sound of tube amps. Over the decades various reasons have been proposed for why tube amps would, in theory, sound more accurate, but none of those supposed reasons stand up to careful scrutiny. The difference that is most obvious by far is the difference in the spectral makeup of the harmonic distortion when the amps are driven to clipping, i.e., whether the harmonic components are even-order or odd-order. Here there are differences even within tube amps, so it does not make sense to attribute the effect to the "softer" distortion of one type of clipping over another, and that's before even considering the question of whether the onset and severity of clipping is the same for the two amps. The point is that even if you were truly able to hear a difference, that in and of itself does not constitute proof that the more expensive amp is more accurate than the less expensive amp. Until you can prove otherwise, you have to allow the possibility that you like the sound of a less accurate amp, and the conundrum is that through listening tests alone, it is not possible to prove otherwise. It is a conundrum.

The stuff about the fact that image files can be converted in ways such that the difference is visually apparent does not prove anything at all about the subject at hand and has no relevance whatsoever. Unless of course your point was to argue that it is in fact possible to apply perceptual encoding using a very low bit rate such that the difference would be obvious.

Oh, don't make me laugh so hard that it hurts. And nowadays Monster Cables are considered the practical alternative to overpriced hype. I found some excellent speaker wire at Lowe's a few months ago. I found it in the lawn and garden section, where substantial lengths of it could be purchased at a fraction of the price for "speaker cable". It was packaged and sold as power cable for low-voltage landscape lighting. Great stuff.

And even though you didn't venture into the vinyl fray, you reminded me of Michael Fremer at Stereophile giving a glowing endorsement of a vinyl demagnetizer of some sort. I suppose that might make sense, but only if first you can explain how it would make sense to demagnetize a non-magnetic substance. He of course is the guy who writes all the vinyl and analog stuff for Stereofool.

Check out http://www.hydrogenaudio.org/forums/

The community there have a collection of killer samples that defeat some compression algorithm's. The search function may well lead you to them. They even seem to have a thread discussing the same topic as this thread - Neal young's tosh.

Maybe so. I haven't heard if. But if I did, my first question would be whether the reason is because of remastering that was done, or because of the SACD per se. In other words, if you took that nice new digital master and then re-sampled it carefully to the CD standard, would it sound any different? How many comparisons of that particular sort have you been able to do at all, much less in a controlled environment where you expectations will not be able to influence the result?

But, in the post to which you replied, by Bloodshotrollinred (whatever), that poster was in effect asserting that ALL recordings done using perceptual encoding inherently contain audible artifacts, no matter what bit rate is used, and no matter how carefully the encoding scheme was devised. To me it is obvious that this defies common sense, and I do not understand why other people do not readily deduce that as long as the bit rate is adequately high, that perceptual encoding will not introduce audible artifacts. He was asserting in effect the perceptual encoding always introduces audible artifacts, and you implicitly endorsed his nonsensical position.

Bingo.

If you have an infinitely high sampling frequency and an infinitely high bit depth then the digital waveform is identical to the analog waveform. Therefore, they will sound identical. The higher the sampling rate (e.g., 44.1kHz, 192kHz) and bit depth (e.g., 16 bits, 24 bits), the better the digital waveform approximates the analog (i.e., original) waveform. Humans are not perfect. At some point, even the most discriminating listener will be unable to tell the difference between the approximation and the original waveforms.

All you are asserting in effect is that all music recorded in a perceptual format sounds the same as your MP3 files, no matter what bit rate you used and no matter what bit rate is used for the other perceptual encoding. In fact, you are in effect arguing that MP3 at an extremely low bit rate sounds entirely the same as an MP3 at an extremely high bit rate.

But do not feel alone. Roughly half of the posts here have in effect argued the same thing.

Throw yourself right in with all the other people who came here and wrote a post that said in effect that all MP3 sounds the same no matter the bit rate.

This is true, but the implication is somewhat incorrect by way of being incomplete. You don't need infinity.

A discrete-time (sampled) signal holds all of the information of the analog original so long as the analog signal's frequency content lies entirely below the Nyquist frequency (enforcing that condition is the role of the antialiasing filter). So long as that criteria is obeyed, sampling and synthesis are both lossless.

The quantization of a sampled signal is not lossless, but we can consider it lossless if it covers the entire S/N ratio of the signal in question (or equivalently to our purposes, the ear's entire dynamic range and maximum S/N ratio). 16 bits linear is very close (with dither its even better) and 20 bits covers the entire dynamic range from inaudible to 'severe permanent damage'.

In this sense, even 16/44.1 linear PCM is already a very wasteful encoding; within a critical band, the ear only has 5-6 bits of S/N resolution, though each critical band can cover a dynamic range of almost 20 bits.

So what's the difference between flac and Apple lossless (alac)?

Apart from wasting space on portable devices.

I'm not sure that any of the posts were really trying to say that all MP3 sounds the same. Many were saying that lower-bit-rate MP3 is discernibly different from higher-bit-rate MP3 or lossless, and possibly not as good to listen to. I wouldn't read any more into it than that.

The threshold beyond which the perceived difference vanishes will vary with the quality of the original, the player, the D-A converter, the amplifier, the loudspeakers and the listener's ears.

I bet the only time Steve and Neil ever came close to having a conversation was if Steve was lip syncing to Suite Judy Blue Eyes... the Woodstock version.

Poor 'ol Neil, one burnt out dude.

Well, this business of compressing dynamic range in many modern recordings is certainly not a good trend, and it obviates one of the major advantages that most any digital recording format has over vinyl, that being of course the advantage in dynamic range. Dynamic range compression was commonplace in vinyl recordings, particularly in recording of symphonic music where the dynamic range from the quietest to the loudest passages greatly exceeded the capability of vinyl. This of course was a major reason for the development of the CD format in the first place. Nowadays lots of people seem to have forgotten this, or perhaps the majority of population interested in music nowadays is not even old enough to have owned a collection a vinyl. At one point I had over a hundred easily. Now I don't have a single one, and I do not miss them in the least. There is not one thing about them that I miss. Only a handful of the vinyl recordings in my collection did not sound overtly inferior to a typical, run-of-the-mill CD, and even the ones that sounded half decent still did not sound in any way better to the same recording released on CD. And on top of that they were large and the collection was heavy and bulky. The cardboard folders constantly wore out at the corners and the backs, and the extra vinyl sleeves were cumbersome. There is nothing about that experience that I miss in any way. More recently I ripped my entire CD collection into iTunes using AAC and now this is the very first thing I do whenever I buy a new CD. When I want to listen to my music, I don't even have to get up off the sofa. The computer that I use for a music server stays powered on, and it is located in my office, and streams music digitally to the little Airport that sits close to my stereo. I reach for the remote control and switch the input, and then I reach for the iPad and touch the icon for the iTunes/Airtunes remote control app. In an instant, I am listening to whatever it was I wanted to listen to, without even getting up. And in general the sound quality is vastly superior to what it was with those old vinyl records. Why would anyone prefer to have to get up and walk to wherever that old orange crate full of dusty vinyl records is kept, flip through them to find the one of interest, pull it out of the sleeve and place it on the turntable, run the brush, lift over the tone arm, and then have to go and put it away afterwards. I would just as soon have a car that I have to start by turning a hand crank, and that has no heater to speak of for winter driving and no air conditioning for the summer, and hard bench seats, a three speed transmission that clunks and grinds every time I change gears, and that constantly drips oil down on to my garage floor.

Yeah, right. (wink, wink). Anyone who knows the first thing about it knows that long before the CD was introduced, vinyl was lamented for several reasons, one of which was that you had to filter out practically all the bass before you could squeeze the signal into that little groove with undulating walls (and then vainly hope to restore a reasonable facsimile of the bass upon playback), and another of which was that even after filtering out almost all of the bass, you still could not record anywhere near the dynamic range of most symphonic music. It was generally necessary to compress the dynamic range as part of the signal processing that had to be applied before the signal could be placed in that little groove. That was the sad reality of vinyl, and still is, yet you claim that "CDs started the compression era". You've got it completely backwards. Dynamic range compression was one of the notable problems with vinyl that CDs sought to solve, and did. If you are talking about data compression, there is no data compression of CD unless the sound is recorded originally at a higher sampling rate and sample size. But even if so, that it itself is a bogus criticism for reasons that have been beaten to death and that are well understood by anyone who truly desires to understand such things.

Totally agree, getting up every 22 minutes to change "sides" still looks no better through rose coloured glasses, so the crackle, hiss and pop has gone, so what, it was just crackle, hiss and pop.

Give me a 20,000 playlist and the surprises that "shuffle" brings, any day of the week.

The quality of iTunes match is surprisingly good, even the 128k MP3 tracks I "acquired"* from Napster last century have gained a new lease of life.

*as "fair use" backups of work I already owned.

What you're advocating is the McDonalds mindset. Never mind the quality feel the width - Why pay more, It's good enough, It's a niche market, blah blah.

I don't eat at McDonalds, I will pay more for a really good bottle of wine, I will pay more for good quality fresh produce. I would happily pay more for HD music just as I will pay more for HD video on iTunes over SD video.

There is most definately a market there. It may not appeal to the kids or 20 somethings with a crappy iPod speaker system in their bedroom but it would appeal to 40 somethings like me who have invested in a good quality hifi system and would like to move from CD to something more convenient and better quality like 24bit.

I came from the tape, avoided vinyl when I could and now sacrifice with CD when SACD titles are not available or in my budget. I thought the range compromises with vinyl and the degredation made for bad tech so I avoided when I could. I did oversimplify my argument about CD being first mainly because most people think CDs aren't lossy. I'll still argue that removing information from the source material to save space is a form of compression whether it's sampling rate or bit depth. I'm not a vinylphile and overstated for the sake of the general audience.

Nice insult at the end though.

Edit: Damn - reading more of your posts and you like to not be polite. "Neil Young does not know one damned thing about anything." ??? I'm gonna check the forum rules to find where it states you have the exclusivity on hyperbole.

Exactly.

Nothing to back up my statement?

OK, then - you remember how Apple was handling everything back in the mid-90's before Jobs came onboard? No?

I rest my case.

Having a bad day, today?

At a guess I would say about 1:35 it changes.

Also there is a noticeable difference in the vocals between the start and end, the music in that one is difficult as there is a lot of bass

Nope, but I'm afraid yours is sophistic.


Thanks, I'm sure I'd never heard of issues of acoustics & resonance before. Did you know room acoustics are subject to similar issues? :roll eyes: Or is your argument that the person talking about his Sennehisers can't hear the difference because they're 'not one of their best' and just DJ models? (Yet presumably better than the ear buds that come bundled with the phone/pod.)


Ummm... yes, it was an analogy, not a proof. I find find pictures difficult to listen to. But as for your comments, you're essentially spending a lot of energy explaining that infinite resolution isn't needed because at some point we can't perceive it? No kidding?


"IF it is true that no one can hear the difference, THEN it is true that no one can hear the difference." Jesus, seriously? What a bit of pedantic tossing. To summarize: infinite resolution isn't needed and at some point people can't hear the difference between compressed and uncompressed. Astonishing.

The original argument was whether the lossy low-bit-rate encodings used for PMPs are so good that nobody can tell them from the high res audio that's out there. I've heard the difference, so I'm going to go with no, they're not that good. Unfortunately you haven't actually helped to resolve whether that's the case or where that point may be.

I want access to the higher res audio for home, while the lower bit rate on my phone is good enough for on the go. Apparently I shouldn't want the better sound because infinite resolution is too much? What!?!?

Epic discussion. Where did you guys come from? I'll have to take a few hours to read through this.

i don't have a link to double blind test that you ask for. but 24/192 files smoke the 16/44 and the 256k files from anywhere hands down no question about it. My 256 k rips i make from 24/192 files also smoke the files from iTunes and they sound as good or better than the cd. They sound far superior on any piece of equipment I have listened to them on. iPod, iPhones, and either of them plugged into car stereo or through the apple tv/ home theatre. i am not just saying that to troll you. I am not an audiophile by any means, but I know what I hear when playing the hi-res files. The volumes you can reach with clarity is unprecedented. And every voice and instrument in these recordings are dripping wet with this clarity.
you don't have to have these hi-res files to enjoy music, but there are some albums and some musicians where you appreciate the higher format. Flaming Lips-yoshimi battles the pink robots is an awesome sonic experience in hi-res compared to its cd16/44 counterpart.

I was on the thinking of, SJ is a true Audiophile, as any audiophile would know, no digital media, CD, DVD what so ever with tens of thousand worth of equipment would replace the sound of Vinyl.

It is also the reason we still keep using Tube Amps!.

I was on the thinking of, SJ is a true Audiophile, as any audiophile would know, no digital media, CD, DVD what so ever with tens of thousand worth of equipment would replace the sound of Vinyl.

It is also the reason we still keep using Tube Amps!.

Everyone knows this story is rubbish. Steve Jobs got up on stage and said he got rid of his expensive audio equipment and only used the iPod hi-fi unit, so one of them is lying