[RRod]

   
As for the first part - I was just using that as an example of how every "new feature" (whether real or imagined) costs extra when it first appears, but quickly becomes "standard". Today you pay extra for CD quality music as compared to AAC or MP3 compressed content - and there is a "market price" for both "ordinary music" and "audiophile quality music"... and I see no reason to believe that this basic fact will ever change. Therefore, eventually, high-res files will fill the niche of "audiophile music" and be priced accordingly... but the price for that niche will level off at "what the market will bear" as it always does.
 
I'm also simplifying the second point to actual reality. I agree entirely; people are very much influenced by their expectations. However, if you want to go that far, then let's take it all the way. If someone buys a high-res file because he or she thinks it will sound better; and then, when they play it, they actually believe that it sounds better, and so enjoy it more, then haven't they in fact gotten their money's worth? Is an expensive restaurant a "cheat" because, thanks to the wonderful ambiance, you imagine the food tastes better? In fact, if someone paid $5 more for that high-res download, and actually enjoyed it $5 more because they deluded themselves into thinking they heard a difference, then aren't YOU depriving them of that $5 of extra value by pointing out to them that they only imagined the difference? 
 
I say that, if people CHOOSE to base their worldview on what other people tell them, to the total exclusions of including their own personal experience, then they deserve what they get - to live in the world as other people imagine it to be.

 
On the first point: we're talking now about a paradigm where "audiophile quality music" is based on differences that are audibly small if at all existent. That's quite different from earlier paradigms were "audiophile quality" meant "has been worked over with special care to make a new mix", as in productions by the likes of MoFi (at least on the content end; hardware is its own can of worms). So yes, while the existence of an audiophile market persists, they way in which is differentiates itself from the "ordinary" market has changed. For instance, you can get AAC of the new Wilson mix of Aqualung for $10 on iTunes, but in hi-res for $18 on HDTracks. So what seems to be separating "ordinary" from "audiophile" in this case is $8 and content that most people probably can't hear, including audiophiles.
 
On the second point: A restaurant with a wonderful ambiance at least has a wonderful ambiance. Our situation would be more like eating at the same restaurant, but having the chef tell you that he's using a new brand of iodized salt. And yes, if someone wants to pay $5 more for that, so be it. My pointing out to them that the salt probably blind-tastes the same as other salt isn't depriving them, it's giving them information upon which to base a decision. People can call that "crapping" all they want. 

 

 

[KeithEmo]

If the download service tells you that the file is 24/192 before you download and after the download you know the content is no better than 16/44, what's that? Fraud would seem to be a word that springs to mind, but the download service is protected, because the file is 24/192 even though the content isn't. You don't think there should be protections for this rip-off?

 
It seems to me like what you're really claiming is that you personally see no benefit to the file being 24/192k - which is a whole different matter.
 
If they said it was a 24/192k file and it wasn't, then that would be fraud. But the fact is that that's NOT what we're talking about here. What they're selling you is a 24/192k file, with a frequency response flat from 10 Hz to 90 kHz, and a dynamic range of about 140 dB. As long as they didn't lie about any of the particulars, then it's up to YOU to decide whether it suits your needs or not. Just because you can't hear a difference does not mean that no difference exists. And, even if I can't hear a difference either, that still doesn't mean that there isn't any. Perhaps I really do want to test whether my pet bat (who can hear to at least 50 kHz) responds better to Mozart than to Beethoven; or perhaps I'm using my AP test set to analyze the upper harmonics of a violin. The point is that, as long as they didn't misrepresent what the product IS, then they didn't defraud you.
 
Common table salt is probably about 99% pure. But, somewhere out there, there's a chemical supply company that will happily sell you chemical sodium chloride that's pure to 99.999999%, and I'm sure they charge a lot more for it than your local supermarket charges for table salt. Would they be defrauding you if they sold it to you for $1000 a pound? No. In fact, if your local gourmet shop decided to sell it as "gourmet super purity salt", then they wouldn't be defrauding you either - because that's precisely what it is. And, even if they said "some of our gourmet customers insist that it tastes better", they STILL wouldn't be defrauding you. And that's true because they have told you exactly what you're buying, which makes it your responsibility to decide whether you need it or not. (And, even if they said "it tastes better", that still wouldn't be fraud - because "tastes better" is a matter of OPINION and not fact. (Perhaps they really do think it tastes better - and, as long as they never claimed that xxx participants could pick it out in a blind taste test, which then turned out to not be true, then they haven't lied. I've seen lots of TV commercials claiming that this or that snack food is "delicious", and then decided, when I tasted it, that I didn't think it tasted good at all. However, I don't think there was any fraud involved.)
 
There is a bit of as grey area - which has happened occasionally with high-res file vendors - where a real plain old 16/44k file was upsampled to 24/192k and sold as such. Now, in that case, you might have a case for claiming that they misrepresented the product, because it technically does not in fact legitimately "have the characteristics of a 24/192k file", and they have also clearly represented that it does. If so, then they've probably made some false advertising claims, and certainly a few misleading ones, which may be legally actionable.
 
There is in fact a most powerful protection you can invoke to keep this from happening to you.....
 
You should use your brain, do your homework, and decide for yourself whether a 24/192k file has any benefit TO YOU and, if not, then you shouldn't buy them. You should also read all the fine print and make sure that they actually do claim, in writing, that the file was converted from the original master to 24/192k, and hasn't been processed in such a way as to negate the benefits of doing so. (Much like, if you do decide to buy some of that 99.999999% purity sodium chloride, you should make sure that the company you buy it from is prepared to certify that it has been handled properly to maintain its purity and avoid contamination.... and that they do so in such a way that their claim is legally binding - so you can sue them if you find out they're lying.)

 

[KeithEmo]

   
On the first point: we're talking now about a paradigm where "audiophile quality music" is based on differences that are audibly small if at all existent. That's quite different from earlier paradigms were "audiophile quality" meant "has been worked over with special care to make a new mix", as in productions by the likes of MoFi (at least on the content end; hardware is its own can of worms). So yes, while the existence of an audiophile market persists, they way in which is differentiates itself from the "ordinary" market has changed. For instance, you can get AAC of the new Wilson mix of Aqualung for $10 on iTunes, but in hi-res for $18 on HDTracks. So what seems to be separating "ordinary" from "audiophile" in this case is $8 and content that most people probably can't hear, including audiophiles.
 
On the second point: A restaurant with a wonderful ambiance at least has a wonderful ambiance. Our situation would be more like eating at the same restaurant, but having the chef tell you that he's using a new brand of iodized salt. And yes, if someone wants to pay $5 more for that, so be it. My pointing out to them that the salt probably blind-tastes the same as other salt isn't depriving them, it's giving them information upon which to base a decision. People can call that "crapping" all they want. 

 

 
I agree entirely.... but it's a matter for the individual considering the purchase to find out what "audiophile quality" means, and then decide whether that matters to them or not. And, again, I have no sympathy if they simply decide to take the lazy way out and assume that it's better because a bunch of other people say it is.... or because it has a new and much cooler name.... Personally, I don't especially like driving vans, and calling them "sport utility vehicles" hasn't changed my mind; but I suspect that a lot of families who wouldn't have even considered buying a van now own a sport utility vehicle.... which is what marketing is all about. (And a lot of people with brand new "audiophile music players" will buy "audiophile music files" to go with them.)
 
I'm not in the least suggesting that you shouldn't do your best to inform people of the facts, and even of your opinions.... Nor am I suggesting that it's at all unreasonable to say that, if they want to convince an informed audience that their product is in fact better, then the vendors selling these products would be well served to produce some test results to justify their claims. I'm just saying that the guys on the other side of the table aren't lying, or cheating, or committing fraud. (The reality is that more people are likely to be influenced by seeing a bunch of "cool people" using a product on TV, or finding out that it's endorsed by a popular celebrity, than are likely to be influenced by actual facts; that's just a fact of modern life.) 
 
 
 
 
 

 

[Roly1650]

It seems to me like what you're really claiming is that you personally see no benefit to the file being 24/192k - which is a whole different matter.

If they said it was a 24/192k file and it wasn't, then that would be fraud. But the fact is that that's NOT what we're talking about here. What they're selling you is a 24/192k file, with a frequency response flat from 10 Hz to 90 kHz, and a dynamic range of about 140 dB. As long as they didn't lie about any of the particulars, then it's up to YOU to decide whether it suits your needs or not. Just because you can't hear a difference does not mean that no difference exists. And, even if I can't hear a difference either, that still doesn't mean that there isn't any. Perhaps I really do want to test whether my pet bat (who can hear to at least 50 kHz) responds better to Mozart than to Beethoven; or perhaps I'm using my AP test set to analyze the upper harmonics of a violin. The point is that, as long as they didn't misrepresent what the product IS, then they didn't defraud you.

Common table salt is probably about 99% pure. But, somewhere out there, there's a chemical supply company that will happily sell you chemical sodium chloride that's pure to 99.999999%, and I'm sure they charge a lot more for it than your local supermarket charges for table salt. Would they be defrauding you if they sold it to you for $1000 a pound? No. In fact, if your local gourmet shop decided to sell it as "gourmet super purity salt", then they wouldn't be defrauding you either - because that's precisely what it is. And, even if they said "some of our gourmet customers insist that it tastes better", they STILL wouldn't be defrauding you. And that's true because they have told you exactly what you're buying, which makes it your responsibility to decide whether you need it or not. (And, even if they said "it tastes better", that still wouldn't be fraud - because "tastes better" is a matter of OPINION and not fact. (Perhaps they really do think it tastes better - and, as long as they never claimed that xxx participants could pick it out in a blind taste test, which then turned out to not be true, then they haven't lied. I've seen lots of TV commercials claiming that this or that snack food is "delicious", and then decided, when I tasted it, that I didn't think it tasted good at all. However, I don't think there was any fraud involved.)

There is a bit of as grey area - which has happened occasionally with high-res file vendors - where a real plain old 16/44k file was upsampled to 24/192k and sold as such. Now, in that case, you might have a case for claiming that they misrepresented the product, because it technically does not in fact legitimately "have the characteristics of a 24/192k file", and they have also clearly represented that it does. If so, then they've probably made some false advertising claims, and certainly a few misleading ones, which may be legally actionable.

There is in fact a most powerful protection you can invoke to keep this from happening to you.....

You should use your brain, do your homework, and decide for yourself whether a 24/192k file has any benefit TO YOU and, if not, then you shouldn't buy them. You should also read all the fine print and make sure that they actually do claim, in writing, that the file was converted from the original master to 24/192k, and hasn't been processed in such a way as to negate the benefits of doing so. (Much like, if you do decide to buy some of that 99.999999% purity sodium chloride, you should make sure that the company you buy it from is prepared to certify that it has been handled properly to maintain its purity and avoid contamination.... and that they do so in such a way that their claim is legally binding - so you can sue them if you find out they're lying.)

That's not what I'm claiming at all. What I can state with 100% certainty, having used my brain and done my homework, that none of the 5 or 6 albums I've downloaded from HDTracks are what they claim they are. They are no better than 16/44, while claiming to be 24/96 or 24/192. The excuse is that they are at the mercy of the record companies and only pass on what they are given, but the 2 Chesky Records I downloaded were the same. Chesky owns HDTracks, so who's kidding who? Upsampled is what it is.
Can they be had for fraud, which is what it is? I think a good lawyer would argue that they are 24/96/192, so where's the fraud?
If you want to kid yourself that you're getting hi rez files when you download from HDTracks, don't let me burst your bubble, I ceased giving them my money a few years back. I suspect that Pono Music is exactly the same. It seems to be forgotten that to remaster even the most popular of the back catalog is a mammoth task, taking years. Does anybody really think the labels are going to do that? Real simplistic to think that's gonna happen.

 

[Roly1650]

I should add, that the Chesky Records albums are superb recordings, but they're 16/44 whatever HDTracks might claim to the contrary. How do I know? I can down/up sample and the files null, I can look at the spectrum and see there's nothing above about 21kHz. which isn't also about 100 dB down.

 

[KeithEmo]

That's not what I'm claiming at all. What I can state with 100% certainty, having used my brain and done my homework, that none of the 5 or 6 albums I've downloaded from HDTracks are what they claim they are. They are no better than 16/44, while claiming to be 24/96 or 24/192. The excuse is that they are at the mercy of the record companies and only pass on what they are given, but the 2 Chesky Records I downloaded were the same. Chesky owns HDTracks, so who's kidding who? Upsampled is what it is.
Can they be had for fraud, which is what it is? I think a good lawyer would argue that they are 24/96/192, so where's the fraud?
If you want to kid yourself that you're getting hi rez files when you download from HDTracks, don't let me burst your bubble, I ceased giving them my money a few years back. I suspect that Pono Music is exactly the same. It seems to be forgotten that to remaster even the most popular of the back catalog is a mammoth task, taking years. Does anybody really think the labels are going to do that? Real simplistic to think that's gonna happen.

 
I would say that, if you can PROVE that the albums you were sold really were NOT converted from analog to digital at the sample rate they claim, then you would have a case.
 
However, it sound to me like you're confusing "remastering" and "remixing" and "re-converting".
 
I agree with you that remixing an entire album is often a major project. The 24/192k re-releases of the Grateful Dead studio albums were re-mixed and restored. (There is a two or three page description of everything that was done to remaster the various albums included in the set on HDTracks website.) And, as a result, they do in fact sound VERY different.
 
When the majority of "high-res remasters" are created, they simply take the master tapes that were converted at 16/44k to make the original CDs and perform the conversion process again at 24/192k to produce a new 24/192k version. Assuming that you have access to the master tapes, and a proper A/D converter, this process shouldn't take more than a few hours. So, if an album is "simply remastered to 24/192k", then the only differences you should specifically expect to hear are those associated with the audible differences between the sample rates (which many people claim not to hear at all).
 
Unfortunately, the word "remastering" is used to mean lots of different things. At this stage, the engineer may in fact also adjust the mix, or the EQ, or perform other "acts of restoration" - but those are a separate thing, and should be described. However, unless any specific claims to the contrary are made, what you should expect is simply a new conversion, starting from the same original master tape, and ending up at a new digital version at a higher sample rate.
 
Note that this is NOT the same as "upsampling" - which means actually taking the 44k digital version and converting it to a higher sample rate file directly. When an album is remastered, the conversion is performed at a higher sample rate, and contains more information at higher frequencies - which may or may not be audible - and may or may not consist of anything more than high frequency tape noise from the master tape.

 

[sonitus mirus]

Not sure how the original reel-to-reel tapes used in the Grateful Dead's studio recordings could magically make use of 24-bit audio.  16 bits was already a substantial overkill.  I've heard both the HD files and the lossy versions, and they both sound excellent.  I can't hear a difference, but I'm sure it is my inferior ears or inexpensive equipment I use. 

 

[Roly1650]

T

I would say that, if you can PROVE that the albums you were sold really were NOT converted from analog to digital at the sample rate they claim, then you would have a case.

However, it sound to me like you're confusing "remastering" and "remixing" and "re-converting".

I agree with you that remixing an entire album is often a major project. The 24/192k re-releases of the Grateful Dead studio albums were re-mixed and restored. (There is a two or three page description of everything that was done to remaster the various albums included in the set on HDTracks website.) And, as a result, they do in fact sound VERY different.

When the majority of "high-res remasters" are created, they simply take the master tapes that were converted at 16/44k to make the original CDs and perform the conversion process again at 24/192k to produce a new 24/192k version. Assuming that you have access to the master tapes, and a proper A/D converter, this process shouldn't take more than a few hours. So, if an album is "simply remastered to 24/192k", then the only differences you should specifically expect to hear are those associated with the audible differences between the sample rates (which many people claim not to hear at all).

Unfortunately, the word "remastering" is used to mean lots of different things. At this stage, the engineer may in fact also adjust the mix, or the EQ, or perform other "acts of restoration" - but those are a separate thing, and should be described. However, unless any specific claims to the contrary are made, what you should expect is simply a new conversion, starting from the same original master tape, and ending up at a new digital version at a higher sample rate.

Note that this is NOT the same as "upsampling" - which means actually taking the 44k digital version and converting it to a higher sample rate file directly. When an album is remastered, the conversion is performed at a higher sample rate, and contains more information at higher frequencies - which may or may not be audible - and may or may not consist of anything more than high frequency tape noise from the master tape.

Not in the least bit confused, reel in the condescension a tad......

Your "a few hours each" would equate, if your lucky, to 1,000 albums/person/year, which sounds like a drop in the bucket. The few data points I have, as opposed to ancedotal blah, blah, suggests that the labels are upsampling the cd master in the majority of cases. The thought of them employing rooms full of people re-recording, in real time, from the original tape is dreaming, imo. Their attitude is likely to be, "we did this crap already for the cd, just upsample for now, we can come back to it if needs be".

Sure, they'll provide the funding for a remaster of the "important" back catalog, if they see a chance of commercial success. At that stage, dependant on access to the multi-track master, an engineer can decide to become artist or the artists themselves can have an epithany and want to change things. But the key is access to the original master, some artists have retained ownership of these and denied use by third parties, some may not exist anymore and others are too far gone.

 

[KeithEmo]

  Not sure how the original reel-to-reel tapes used in the Grateful Dead's studio recordings could magically make use of 24-bit audio.  16 bits was already a substantial overkill.  I've heard both the HD files and the lossy versions, and they both sound excellent.  I can't hear a difference, but I'm sure it is my inferior ears or inexpensive equipment I use. 

 
I have no idea if the 24 bits makes any difference or not; and, for that matter, I haven't down-sampled them to 16/44k to see if that would hurt either. However, in this latest iteration, the re-masters have also been completely re-mixed, and, at least according to the literature, they actually adjusted the time stability, and did some other significant restoration on some of the albums as well. The re-mixing is pretty obvious and, at least to my ears, all for the better (which I don't always think is the case). In short, I think the improvements they made justified the price of buying the set again, so I'm not really concerned whether any of that improvement was due to it's being issued at 24/192k or not.
 
As for lossy files..... I've been around through the invention and evolution of MP3 files, and I've done the whole "let's see which encoder and which setting works best with this song" routine, and I simply don't have the time to waste. (Because of the way MP3 encoding works, even though the decoding process is standard, different encoders are allowed to make different decisions about what to discard, which means that, at the same exact settings, the output of two different encoders often sounds different. This means that, if you're determined to use lossy compression, it makes sense to encode each song with several different encoders and then decide which version you prefer - and, with luck, one of them won't sound noticeably worse than the original. Nowadays, storage space is cheap, and my time isn't, so it's not worth it to me to take the time to find out if I could store some of my music in a little less space.)
 
Just for the record, I will say that I absolutely have bought 24/192k albums that didn't sound any better than the 16/44k original (I'm sure they were simply a re-conversion at a higher sample rate rather than a re-mix; and, either the sample rate really doesn't matter, or that particular master tape simply wasn't good enough for it to make a difference). However, this has always been the situation with "remasters" and "reissues" (some are better, some worse, and many not noticeably different), so I don't see this as specific to "high-res" reissues.

 

[KeithEmo]

You could be right - although, especially in this particular discussion, "anecdotal evidence" seems to be frowned upon.
 
As someone who does understand the actual technology, I also have trouble with a lot of the anecdotal claims themselves. For example, many of the early studio mastering tape recorders simply didn't have frequency response past 20 kHz, so there won't be anything on the master tapes above there. If that's the case, then I can convert that tape to digital at 16/44k, or at 24/192k, and probably not end up with any more information (other than a more accurate rendition of the ultrasonic tape hiss and, perhaps, a cleaner recording of the residual bias frequency). However, while this suggests that encoding that particular tape at anything above 16/44k is "pointless and useless", it still doesn't mean that it wasn't done. There is some relatively simple software that can determine with pretty good accuracy, by looking for specific anomalies, whether an uncompressed audio file was "sourced" from an MP3 file. However, reliably determining whether a high-res file was really upsampled from a CD, or was simply made from a master tape that was of limited quality to begin with, is a little bit more difficult.
 
Several years ago, the "high-res remaster" of a certain popular album was found to be an upsampling of an ordinary CD (I don't know how it was determined for sure). There was a proverbial "big stink" about it, apologies were made, along with the claim that it was "an error", and a new proper version was provided to everyone who had bought the bogus one. And I'm sure that it damaged sales figures for the companies who carried that label. I would have to suspect that, if any of the major labels was provably and consistently really selling simple upsampled copies, someone would have found out, made accusations, and published a major expose on the subject.
 
Of course, another possibly controversial area is where a source, which could include an original 16/44k file, is "fixed" and then re-encoded at 24/96k or 24/192k. One example of this is the current version of Dolby's latest "professional encoder" - which is very commonly used to encode the audio tracks on Blu-Ray discs. One of the options this encoder offers is to re-encode content that was originally at 48k, using special filters to "repair" some of the ringing caused by the digital filters used in the original encoding process (they use digital processing to shift pre-ringing into mathematically equivalent, but less audible, post-ringing). Interestingly, the output of the encoder with this option enabled MUST be taken at 24/96k (and they specify that the benefits of the improvement will be lost if you reduce that back to 48k). In the context of this discussion, however, what you have is a 96k version of a file, which began life as a 48k file, but has been processed in such a way that the result, which is claimed by Dolby to sound significantly better, can only be stored at 96k or higher. In other words, at least according to Dolby, you have an "upsampled copy" which is better than the original - because of the improvement their processing has wrought. (So, at least according to Dolby labs, an upsampled CD really could be different - and better - than the original.)  
 
You can find white papers about it on Dolby labs' website here:
http://www.dolby.com/us/en/technologies/dolby-truehd-encoder-white-paper.pdf
http://www.dolby.com/us/en/technologies/dolby-truehd-lossless-audio-performance-white-paper.pdf
 
 
Quote:

T
Not in the least bit confused, reel in the condescension a tad......

Your "a few hours each" would equate, if your lucky, to 1,000 albums/person/year, which sounds like a drop in the bucket. The few data points I have, as opposed to ancedotal blah, blah, suggests that the labels are upsampling the cd master in the majority of cases. The thought of them employing rooms full of people re-recording, in real time, from the original tape is dreaming, imo. Their attitude is likely to be, "we did this crap already for the cd, just upsample for now, we can come back to it if needs be".

Sure, they'll provide the funding for a remaster of the "important" back catalog, if they see a chance of commercial success. At that stage, dependant on access to the multi-track master, an engineer can decide to become artist or the artists themselves can have an epithany and want to change things. But the key is access to the original master, some artists have retained ownership of these and denied use by third parties, some may not exist anymore and others are too far gone.

 

[sonitus mirus]

   
I have no idea if the 24 bits makes any difference or not; and, for that matter, I haven't down-sampled them to 16/44k to see if that would hurt either. However, in this latest iteration, the re-masters have also been completely re-mixed, and, at least according to the literature, they actually adjusted the time stability, and did some other significant restoration on some of the albums as well. The re-mixing is pretty obvious and, at least to my ears, all for the better (which I don't always think is the case). In short, I think the improvements they made justified the price of buying the set again, so I'm not really concerned whether any of that improvement was due to it's being issued at 24/192k or not.
 
As for lossy files..... I've been around through the invention and evolution of MP3 files, and I've done the whole "let's see which encoder and which setting works best with this song" routine, and I simply don't have the time to waste. (Because of the way MP3 encoding works, even though the decoding process is standard, different encoders are allowed to make different decisions about what to discard, which means that, at the same exact settings, the output of two different encoders often sounds different. This means that, if you're determined to use lossy compression, it makes sense to encode each song with several different encoders and then decide which version you prefer - and, with luck, one of them won't sound noticeably worse than the original. Nowadays, storage space is cheap, and my time isn't, so it's not worth it to me to take the time to find out if I could store some of my music in a little less space.)
 
Just for the record, I will say that I absolutely have bought 24/192k albums that didn't sound any better than the 16/44k original (I'm sure they were simply a re-conversion at a higher sample rate rather than a re-mix; and, either the sample rate really doesn't matter, or that particular master tape simply wasn't good enough for it to make a difference). However, this has always been the situation with "remasters" and "reissues" (some are better, some worse, and many not noticeably different), so I don't see this as specific to "high-res" reissues.

 
I'm not really struggling with the lossy conversion process at all.  I use the Lame encoder and have the command line parameters set at a quality level that I have, to this point, been unable to hear a difference in an ABX test with any music CD that I have ripped or HD file that I have converted.  My listening goal is to try and have a library that includes practically everything.  I want the equivalent of a huge Tower Records store filled with millions of albums at my fingertips.  I get this with the music streaming services, and I don't believe I am sacrificing sound quality for the convenience.  I've done a great deal of testing to prove this to myself.  In this process, I have also become quite skeptical of the HD audio industry and those that rave about it without also providing any type of verification other than suggesting that they trust their own ears.
 
I was able to test the HD (24/96) version against the streaming MP3 files that Google Music subscription service uses.  (Lame version 3.98 320 CBR)
 
I wanted to make sure I was not missing anything, and I could not hear any difference.  If there is a difference, it is more subtle than any differences found in one of the Philips Golden Ear challenge tests, which I can pass, albeit with considerable effort.

 

[KeithEmo]

I think the problem is that audiophiles tend to like to exaggerate.... the differences we're talking about are much smaller than, say, the differences between two different brands of speakers.
 
Back when MP3 files were new, I compared the output of various MP3 decoders to the original (which was a CD; this was before high-res files). My conclusion was that, if I used a reasonable level - like 256 VBR or 320k CBR, on about 80% of 85% of them I didn't hear any difference at all; on another 10% or 15% I could hear a slight difference, if I listened carefully, but it wasn't huge; and, on a few specific recordings, the MP3 versions sounded somewhere between very different and absolutely awful.
 
Which makes perfect sense if you realize that the MP3 format was designed based on the concept of "psychoacoustic masking" - which is a fancy way of saying that, if there's a loud sound, it keeps you from hearing the details of quieter sounds that are near it in frequency and time, so you can store those quieter sounds with much less resolution and probably not notice the difference. And the way the encoder decides what information it's safe to store at lower resolution is based on a whole bunch of research about exactly how this masking effect works. As a result, it really does work well with most music (because most music "fits the model"); however, with certain specific pieces of music which don't fit the model, it doesn't work well. (It's easy enough to deliberately construct a test signal where the difference will be obvious.)
 
Also, since this model is based partly on what different people do and do not notice, different people react differently to the result. I personally find MP3 versions of vocals to sound very close to the original, but that cymbals often sound very wrong. Other people say that it tends to make the attack transient on piano notes sound odd, but I don't tend to notice that very much. And most people find that "electronic music", made up of pure tones, often gets altered quite a bit by the process. The effects also vary based on the specific frequencies present in the content, and on how long they're present, (so you may end up with a recording that sounds perfect - except for a few "odd spots" where there are obvious artifacts).
 
 
 
 

   
I'm not really struggling with the lossy conversion process at all.  I use the Lame encoder and have the command line parameters set at a quality level that I have, to this point, been unable to hear a difference in an ABX test with any music CD that I have ripped or HD file that I have converted.  My listening goal is to try and have a library that includes practically everything.  I want the equivalent of a huge Tower Records store filled with millions of albums at my fingertips.  I get this with the music streaming services, and I don't believe I am sacrificing sound quality for the convenience.  I've done a great deal of testing to prove this to myself.  In this process, I have also become quite skeptical of the HD audio industry and those that rave about it without also providing any type of verification other than suggesting that they trust their own ears.
 
I was able to test the HD (24/96) version against the streaming MP3 files that Google Music subscription service uses.  (Lame version 3.98 320 CBR)
 
I wanted to make sure I was not missing anything, and I could not hear any difference.  If there is a difference, it is more subtle than any differences found in one of the Philips Golden Ear challenge tests, which I can pass, albeit with considerable effort.

 

[sonitus mirus]

 

 
Yes, your description of MP3 comparisons match what I typically find, though I am only able to hear these differences at much lower bitrates. I also listen at very low volume levels for extended periods of time, rather than a shorter, louder session, so I probably can get away with lossy formats easier than most.  I generally turn up the volume level when I attempt to ABX, but I have a difficult time hearing artifacts, regardless. 

 

[RRod]

I wanted to make sure I was not missing anything, and I could not hear any difference.  If there is a difference, it is more subtle than any differences found in one of the Philips Golden Ear challenge tests, which I can pass, albeit with considerable effort.

 
Among the tests in the Golden Ear challenge, the frequency band and mp3 compression tests are often cited as the hardest. And the compression test doesn't go beyond 192 if memory serves. (edit: it goes up to 160 only).

 
Also, since this model is based partly on what different people do and do not notice, different people react differently to the result. I personally find MP3 versions of vocals to sound very close to the original, but that cymbals often sound very wrong. Other people say that it tends to make the attack transient on piano notes sound odd, but I don't tend to notice that very much. And most people find that "electronic music", made up of pure tones, often gets altered quite a bit by the process. The effects also vary based on the specific frequencies present in the content, and on how long they're present, (so you may end up with a recording that sounds perfect - except for a few "odd spots" where there are obvious artifacts).
 

 
I think "very wrong" and "obvious artifacts" are quite a strong word choices unless you've tested things blindly. As I said above, the Golden Ears test is pretty hard, and indeed I've read many people focusing in on the transients to eek (eek, mind you) out the differences at bit rates lower than 320kbps.

 

[Bmullock]

Why do we assume that our ears are the only receptors to sound. Other parts of the body may accept higher frequencies and process them for the brain. Back to double bling crossover studies of a reasonable statistical sample if you want to get anywhere near the truth and remove selling hype!