First of all, apologies if this seems like a lame post, or I'm trolling, or if this has been discussed to death. I don't hang out on this forum that much, and I'm actually just hoping for a simple answer...

It's been a while now since I encoded any CDs, and my collection has grown significantly since then so I'm hoping to re-encode my entire collection sometime soon. My question is, is mp3 dead, and if so which of the formats seems the most promising at this moment?

I'm not bothered at all about encoding speed, and I've started using Foobar2K (which I believe can handle just about any file format), and I don't mind large files (although the smaller the better obviously); the only thing I really care about is that my files are completely and totally transparent to the original CD, and that I won't have to re-encode them all when "standards rise" and it turns out the file format I was using isn't actually transparent!

In a nutshell: I want the best there is, but I don't see any point in lossless if lossy sounds identical.

Until now, I've always used mp3 (EAC with LAME's alt-preset-extreme setting). I was always given the impression (from reading these forums) that transparency is perfectly achievable with mp3s, once you get above 200-250kbps(ish), and that the alt-preset-extreme setting would give me 99.9% transparent files (maybe the odd moment every thousand tracks may sound slightly off, but for the most part I would never notice any difference). I was also under the imporession that most people could never really hear any difference between a LAME mp3 and the original CD.

However, after talking with a group of self-proclaimed audiophiles in another forum, I've been told that this is completely untrue and in fact it's quite easy for an audiophile to tell any mp3 (even alt-preset-extreme) apart from the original CD. Their reason is this: normally when a CD player plays the inaudible frequencies that mp3 cuts out, these frequencies somehow 'interact' with the audible frequencies to change them, which results is a feeling of 'depth' and 'space' and 3D that you never get from an mp3. According to the guys on this other forum, if you listen to a CD, you can generally hear a lot of information about where the musicians are placed, and you feel a greater sense of 'depth' in the room, but with an mp3, this information is gone and it all sounds 'flat'.

So, what I must ask is: are these guys right? Is mp3 flawed in this respect? Isn't LAME tested by thousands of people? How could they miss this?

And if mp3 is flawed, what is the best alternative file type? How do you recommend I encode my large CD collection in the most future-proof way?

>>in fact it's quite easy for an audiophile to tell any mp3 (even alt-preset-
>>extreme) apart from the original CD.
If you are unsure, test for yourself, it is rather easy
But I doubt you will be able to hear any difference.

EDIT: Oh, forgot, and use headphones for the test - it easier to hear differences with headphones than with big speakers... But you can try speakers too, if you wish.

Well, personally I have never been able to tell the difference between any LAME mp3 and the original. But I'm not an audiophile.

But just because I can't hear the difference doesn't mean I'm happy having files that sound bad to other people. Not just because I play my music for my guests, but also because the idea that my files are 'imperfect' bothers me! (Very irrational, I know!)

Whoever said that was pretty much just feeding you crap. Its probably best you ignore him.



Sounds like there is no problem then.

Also, I'm curious what makes one an audiophile. Is it the perpetual inability to ever be satisified with audio or the hanging around on the internet trying to make other people not be satisfied?

LOL! Nicely put.

I would be surprised if MP3 "dies" anytime soon - done well, it is more than good enough for the vast majority of listeners. Even another codec that sounds slightly more transparant doesn't really solve any problems that anyone would have with mp3. Another codec would have to be significantly better sounding (not much room for that for most listeners) or significantly smaller file size (possible) and be non-DRM (unlikely) to be compelling over mp3. Long live mp3. Have my whole collection in mp3; backed up to FLAC in case I need to transcode sometime in the future.

[

Audiophiles are like the annoying little trolls that like to make up statements and pull claims out of their ass without any scientific evidence to back up their statements, whether or not they agree with it.



Vorbis fits that criterion . Everyone should be free to use, whatever they like that's why there is choice out there

Vorbis certainly does not meet those criteria.

Thanks for the replies, guys. This is putting my mind totally at ease about mp3s (sigh of releif). I'm quite glad, actually, because there are a few CDs that I've encoded and I don't have any more, so they now exist only in mp3 form.

Amazing how many liars there are out there on the Internet who will swear to something that is completely untrue. I should go back in that forum and tell them to stop spreading their lies!

I think I'll carry on using EAC/LAME alt-preset-extreme, as I always did, and apologise to my mp3s for ever doubting them.

lame -b8

It'll give you super small file sizes.

I thought the minimum bitrate in mp3 was 32? In the histiogram, that is the lowest shown bitrate.

It most certainly does, for some meanings of "significantly," for some people.

Transparency does not change - either the encoded file is transparent to you or it is not.

Depends on the MPEG version (and therefore sampling rate): http://www.id3.org/mp3frame.html

I have no doubt about MP3 being transparent to me or to anyone else if given a decent bitrate (--preset standard or even extreme), but I feel more comfortable archiving lossless rips anyway, just in case a better lossy codec gets invented, either providing increased compression and still reaching full transparency, or simply requiring less ressources for decoding, for instance. My MP3's would still be transparent, but I would want to switch anyway for reasons other than audio quality. And given the hassle of ripping audio CD's (or even worse, whatever DRM infected format the industry is plotting to feed us with), I think it's much less painful to rip them once and keep them archived in a format that's much easier to deal with.

I read lots of audiophiles giving opinions, I never read them backing anything up with evidence.

Somewhere along the line the idea that "everyone has a right to an opinion" got distorted into "everyone has a right to an opinion, and all opinions are equally valid". That is the idea that informs audiophiles. They have no interest in evidence or empirical data. Being an audiophile is just a game of people throwing around different opinions. Sort of like a group of university academics talking about post-modernism.


This is just a standard debating line. A lot of audiophiles denigrate CD for this very same reason! Somehow only vinyl LPs offer the right sound stage and "warmth" (whatever that means).

Other lines of argument include "but all people hear things differently", the level of difference they propose would make the difference between the hearing of two average humans the same as the difference between a human and a cricket.

Implicit in their line of argument is a beleif that psychoacoustics is witchcraft, and that everyone has unique hearing. It is a pissing contest based around who is better at finding flaws in things. Instead of understanding that provided someone's hearing isn't severly damaged, most people's hearing abilites are about the same.



Audiophiles: Often prefer analog to digital, usally prefer collecting hifi equipment rather than music, usually posses certain dogmas against new audio technologies, and refuse to ever test whether their dogmas are supported by evidence.

It's sort of like a religion when you think about it.

While the explanation is bollocks there can be problems with some encoding methods for MP3 and stereo separation which can lead to a lack of "depth". However you just avoid those methods (basically avoid Intensity Stereo).

Anyone to consider the OP.... I'd recommend that you rip your collection one last time using EAC and FLAC as the codec. You can then transcode to any codec you like. Personally I think that a FLAC archive combined with MP3 for portable use is the perfect combination.

As for the quotes about transparency requiring 220kbps+ bit rates... clearly some people don't understand what the listening tests are telling us.

As for Audiophile.... I'm an audiophile and so is everyone here. The childish denegration of the term does us all a disservice. Audiophile "A person having an ardent interest in stereo or high-fidelity sound reproduction.". Audio=sound Phile=lover


Fairy

OK, sorry, no offence meant. I am got a bit fired up due to things like this: Ok, Lossless is lossless

I consider myself a musicphile. I'm more interested in music. All HiFi (and I see compression technologies as part of that - a controversial stance!) are designed to serve the reproduction of music, they are tools. I personally don't buy equipment based on apperance. I'm not interested in collecting hifi equipment, but I do cherish my CDs, and more importantly, the music content on them.

musicphile is a made up word that means audiophile. Don't let the idiots take ownership of words :-) I love music and part of that means that I have selected equipment (quite low end separates) that make a world of difference over my old "all in one system".

I really don't see how people can't come out of the closet and say "I am an audiophile but I have never thought a passage of music sounds like a naked girl running through thistles in spring" or "I am an audiophile and I can't tell the difference between an APS and the original source material in a blind test".

By the way I'd avoid that forum those guys have been smoking weed while sitting on poles.

I know that thread! (The second post is mine. )

One of the great things about the HA forums is TOS#8, which states (in part) that "All members that put forth a statement concerning subjective sound quality, must -- to the best of their ability -- provide objective support for their claims." This eliminates the groundless blustering that we see so often in other forums. It's too bad that the Hoffman forums don't have such a stipulation; I've noticed a lot of misinformation about audio codecs over there.

But as for you, ShowsOn, if you want to "future-proof" your audio files, you should really look into lossless archiving (as previously stated). With your files in FLAC, WavPack, Monkey's, or any other lossless format, you can easily encode to any lossy codec that comes along. (Personally, I use FLAC and WavPack at home, and OggVorbis on the portable.)

Hmm... after reading through that Steve Hoffman thread, it seems there are plenty of people who genuinely can tell the difference between a LAME mp3 and the original CD.

So... that would contradict what a lot of people are saying here. Surely there can't be that many liars?

As long as there are hard drives based on platters for mp3 players we will never loose our mp3 format. To many peoples music collections are mp3 based. the new lame is also very fast and I am happy with it

Read the forum rules and what was posted in the 13th post of that thread:


This forum it seems is the anti-hydrogenaudio site. You can make completely baseless claims, and by the rules of the forum not be challenged in any scientific way. You can ask someone if they used abx, but you can't deny their claims based on that. To each their own I suppose. Those who want scientific discussion to make informed decisions about audio compression and things of that nature should use hydrogenaudio.org.

I wouldn't call these people liars, I would just say that the majority of them did not prove that they could tell the difference, they just said they did. I generally don't make it a habit to believe claims of that nature unless someone has proof.

Uh, no, Vorbis does not meet this criteria. I'm talking about:
Significantly better sounding: for 95% of the listeners out there MP3 and AAC are just fine and pretty much equal. There isn't really any room in sound quality for the average listener to completely change over to on the basis of sound quality - I'm talking about the need for the difference on the order of LP to CD. All of the codecs are at least good enough. I don't believe that if you took a reasonably well encoded mp3 and a vorbis file that your average listener would hear a significant difference or care about it if they did.

Significantly smaller file size: no way vorbis is compelling here. Maybe 10~20% smaller at best, not what the average person would call significant. To really be compelling to change vs. mp3 in my opinion you would need to be an order of magnitude smaller - say 10% the size of an mp3 at roughly the same quality for people to change codecs on the basis of file size.

Non DRM: yup, Vorbis got that.

Don't get me wrong, love my Vorbis, but it is not significantly better than MP3. Its close to mp3, maybe slightly better in some ways, but people don't migrate to a new format for close. It has to be significantly better. Think LP to CD, VHS to DVD, etc. AAC might even have a shot despite the DRM just because it looks like the only format with enough support for legal downloads as the moment; but if given a choice I'd be surprised if most people wouldn't rather download an MP3 without DRM instead of the AAC.

I still don't see any codec that challenges MP3 for market acceptance, but I'm open to suggestions.

Hey, just because it doesn't meet your criteria doesn't mean it doesn't meet mine. -q3 Vorbis sounded better than 128kbps LAME for me, and for me, the usual decrease in filesize and the increase in quality is very significant.

So for some people, Vorbis does offer significantly higher quality at significantly lower filesize, and it's a viable alternative. You can't deny that.

Well it depends on the bitrate... but even the ones who think they can tell the difference between 320kbps and CD, they probably genuinely think they can tell, sure. That doesn't make them liars, they're probably just imagining it. If you're not blind-testing when listening for a difference, and you expect to hear one, you probably will. At least that was my experience before I tried blind testing (and the experience of other people I know too). It's quite humbling to realise you were fooling yourself the whole time - it's like the placebo effect gone crazy. For example people spend ridiculous amounts of money on expensive cables and are sure they hear a difference (as did I), and then you put yourself in controlled conditions and find out you were completely imagining it.

see for example:
http://www.hydrogenaudio.org/forums/index....topic=33951&hl=
and
http://www.hydrogenaudio.org/forums/index....topic=29989&hl=
and this thread for a laugh/groan:
http://www.hydrogenaudio.org/forums/index....topic=37485&hl=

Regardless of any genuinely perceivable differences, the same thing is bound to happen with mp3 vs CD or anything else where there's is a preconcieved expectation about differences there might be there to hear. This is why double-blind testing is a must. You don't have to be stupid or a liar to hear a difference that isn't there.

Removed, inappropriate post.

If an encoding artifact is audible to one's ear, one will hear it regardless of the equipment.

And of course no one tests lossless, because it is "lossless", i.e., it has no loss against the original.

Keytotime: Have *you* ever listened to MP3s on a good system? Could *you* prove you can hear a difference in a double-blind ABX test with such equipment?

If you can't answer yes to both questions, your claims have no merit. If you *can* answer yes to both questions, could you show us your test results, please?

It is possible that someone can hear a difference between a lossy encoding and an original. Lossy is lossy after all.

A big "However", is that I have seen virtually no proofs on that thread, except possibly OcdMan. Then again scientific results would be much better.

Back to your original question. I would suggest you go for lossless, because from time to time you will face the dilemma on re-encoding your whole album, whether to another codec or to the same, but updated codec. Re-ripping and re-encoding is much more time consuming than simply converting from your lossless rip. This way, you only need to rip your precious CDs once (and you can keep the CDs in a safe place), and the lossless is treated as a backup. When you need to get lossy copies, e.g. for your DAP, you can convert them from your lossless backup. I even use DVD+Rs to backup the lossless rips, such that I can hold several lossless CD rips on a single DVD.

The point of quoting the hardware tests is that there is a known, proven psychological phenomenon, the 'audio placebo effect' or whatever you want to call it, which fools people into thinking they hear differences when they expect to hear differences (see especially the second link I gave, which talks more about how certain people were about the percieved difference). This result-distorting phenomena is in the brain, not in the hardware or software. So it will be present when making any comparision, including mp3 vs cd comparisons, audio cables vs starter plugs or anything. That's all I was saying. The fact that there are many people who are certain they hear a difference when in fact they don't shows that we're systemically over-anxious detectors of perceptual differences. (this is probably for good evolutionary reasons: false positives when listening for a leopard approaching in the dark are not as costly as false negatives).

So I wasn't saying that there are not audible differences, especially between sources and lossy encodings at low-mid bitrates (and I understand the difference between lossless and lossy). But we should all expect lots of false positives from others and ourselves, and we need something to filter this systemic error out from the veridical information. That's what double-blind testing is for. If you're sure you can hear a difference you're probably right, but probably isn't good enough and the default position in the absence of testing should be skepticism.

Plus there are a lot of false theories and expectations out there about sound quality that still sound perfectly plausible. I have no idea about what quality of wire it takes to carry a clear electrical signal - that's pretty advanced physics. Perhaps starter plugs/cables will sound crap, perhaps not (see the quote in the linked page about using a power extension lead for top-of-the-range speakers...). I'm with you and won't be using them any time soon, but it'd still be interested to see the test results.

Most of what counts as discussion is "groundless blustering". If they had to back things up with evidence then their 'discussions' would ground to a halt.

I usually just post to the music forum. I couldn't care less if someone gives an opinion in there that I don't agree with. But when a piece of crap gets repeated as gospel concerning compression, then I can't help but speak up.



I rip everything to FLAC and convert using Foobar & iTunesEncode.exe to iTunes AAC for my iPod.

Lastly, I have visited this forum for a long time, but I don't post that much. I feel there are a lot of people in this forum that know a lot more about the topics discussed that mean. So instead of me just posting a heap of rubbish, I prefer to read through threads to learn what the consensus is. This differs to other forums I know where everyone is encouraged to just jump in with whatever idea they dreamt up the night before.

http://www.hydrogenaudio.org/forums/index....96&#entry312296

With 128 kbps LAME mp3 versus 80 kbps Vorbis, the reduction in bitrate was nearly 38% in size with quality "almost as good" as LAME across various genres. This marked improvement would even be more so if we were comparing fhG or other, less advanced mp3 codecs. I would really like to know if "95% of the listeners" could differentiate between the aforementioned codecs and the associated bitrates...

I would agree with kjoonlee in saying this is significant as well. Hell, this is the only reason I chose the new Ipod Nano 4GB for my birthday this weekend (thank you, Rockbox devs ). My trusty Iaudio I5 will be going to a needy family member.

From the same thread (on the results discussion):


Then again: not significant.

Use whatever you want, though.

I use MP3 just because it plays everywhere, everytime (no need for hacked firmrware, no matter how good)

Well, I'm not sure anyone's done any 96 kbps Ogg vs. 128 kbps tests, which is still 25% smaller and most likely perceptually indistinguishable to the majority (if 80 is not).

When extrapolated, the savings add up, either way.

Also, mp3 may play everywhere but in very few places does it play gaplessly (thank you, Rockbox. ).

Yea! -q 2 with Vorbis all the day ... well I did some ABX and the results are usually 10/16 to 14/16... but when played on my setup at full blast... plus 1 bottle of vodka, some tequilla shots, copious mugs og beer...

Suffice to say I didn't notice the artifacts

On more serious occasions... the low-frequency thumping of bus engines (which I ride everyday) perfectly covers up any artifacts

And yes the size savings do add up real nicely

So this ABX featured -q 2 versus Lame128...well, maybe I should prepare to eat crow now. Have you tried -q 1 versus Lame128?

So what do you define as a good system. I'm more than happy with MP3s using --alt-preset standard on my system. Good KEF speakers, good Denon amp, decent quality interconnects/speaker wire etc.

Anyone here heard of What Hi-Fi? magazine in the UK? They like to call themselves 'audiophiles' and can apparently tell the difference between mains plugs! Wow they must have amazing hearing

I spent a weekend at a friends house last year discussing the merits of MP3's. He was dismissive, having never really listened to them and being an audiophile himself. We spent the weekend listening to music I had burned onto CD's through his Naim Audio system and Dynaudio speakers, about £15,000 of kit in all and comparing them with the originals. This was a subjective and informal listening test but neither of us were able to spot any difference between the sound of the discs, and what my friend didn't know (until I told him later) was that my CD's had been burned from MP3's encoded using -alt extreme setting in LAME. Needless to say though, it all sounded stunning, and since that day, MP3's have been more than good enough for me.

I used to read What Hi-Fi? out of pure smugness just to count all of the factual errors they made and the wrong advice they gave to the poor saps who wrote in. I still have a flick through it in newsagents now and again. Not only is it still full of biased (depending on which manufacturers have spent the most on advertising with them this month) subjective drivel, it now reads like it's been written by a bunch of 12 year olds.

Same here, I keep meaning to write to them and point them here and ask why they never produce any links to information and resources.

Recently they had an issue about using your computer for music. Not once did they mention VBR or provide you with any resources to find out more about things like LAME.

Such a pity that so many believe the rubbish. That's why I started reading Home Cinema Choice for that side of stuff. At least they provide technical measurements for everything so there's a somewhat more scientific basis for comparing things.

It's already been said, but the solution is to archive in a lossless format and transcode if/when required. This should provide future proofing as the lossless stuff can be re-encoded to better lossless formats as they become available (without building up potential step by step degradation).
The only other thing that i would add is that people seem to be underestimating the placebo effect. The fact is, whether it can be experimentally or otherwise proven that your ear cannot tell the difference in a blind test, this does not change the fact that some people will still subconciously invent a difference when listening to a music encoded in a lossy codec. As such I think it is perfectly reasonable for those people who do not wish to percieve a difference to take steps to avoid it. It has been shown in medical studies that those people who have been given placebo treatments (including surgical procedures) and seen improvements, have not reverted to a state of worse health after being informed that they were in a placebo group. As such it may not help anyone to tell them or even demonstrate to them that they cannot really tell the difference.
Please don't treat this as flamebait, it makes sense to me

Oh and I have come up with a name for the condition too:

Audiophilus Placebitis

Sounds scientific doesn't it

Well said!

Hmm, ok i'll bite.

It's known that expectation can directly affect experience in all areas of sensory perception. A quick google turned up the following paper on the role of expectation in designing auditory spaces.

http://scom.hud.ac.uk/scompc2/pdf/pchuengIHCI.pdf

In particular section 3.1 deals with imagined experience. I'm not presenting this as fact or direct evidence, just that it is an area that is being studied. Further, the fact that so many people claim to hear differences or artifacts in encoded formats that can be shown to be imagined (as is the consensus - that i am indeed accepting) is evidence of a psychological effect (which also seems to be the consensus).

Now my hypothesis (which i think you consider "a heap of rubbish") is that simply pointing out the fact (or proving it through blind tests) that the perceived differences people are attributing to lossy encoding are in fact imagined may not result in those people no longer experiencing the differences under non test conditions. My reasons for thinking that this may be an over simplification are that this perception is a learned response - perhaps derived from the mere use of the word lossy in this context which has learned negative connotations. This learned response will then be backed up by experential evidence on the part of the listener over time (all be it tainted experience). This kind of learning can be hard to overturn.

My conclusion (I know i've skipped the experimentation stage, so maybe i shouldn't be drawing a conclusion, but here goes) is that it is reasonable for those that feel lossy encoding will result in a poorer audio experience to encode at a level they are comfortable with (maybe even losslessly) in order to avoid psychological effects. Whereas as trying to unlearn their conditioning may (and you will notice that throughout i tend to use "may" rather than "would") prove to be more difficult if not futile.

However I will concede that if you have a task to perform that requires you to remove the taint of expectation from experience then the effort involved in doing so would be worthwhile.

What prompted me to make this hypothesis is the continual over-heated argument that goes on between those that say they can tell a difference and those that point out that it can be proved that they cannot. In these cases the proof never accounts for the inherent "faultiness" of the human brain (a major component in the system that is being checked) - in fact the blind test deliberately removes the effect of expectation. Unfortunately under normal conditions expectation is always present and has a number of other effects too.

One thing that i think may help would be stop using the term "lossy" then newcomers to the field may not fall into that expectation trap (I know its use is strictly correct and does not refer to perception per se). However I still think there will be a tendency for newcomers to start encoding music at low bit rates in which differences may be real and then experiment and find that as they improve bit rate the quality of the recordings increase - the problem here is that they will build up expectation that whenever they increase bit rate, better quality ensues, thus they will likely be unable to tell accurately at which point quality stopped improving and will tend to overshoot (another hypothesis). So this problem is possibly doomed to recur forever in each generation of music lovers.

oops, some may think that i've gone way off topic here. So here is something to bring things back on track.
If we accept that expectation is a major factor in perception and that expectation changes over time then this backs up the approach of archiving audio losslessly so that when expectation changes we don't have to waste too much time searching out "facts" that can be argued endlessly - but instead transcode to whatever format fits our expectations best.
The process of receiving recommendations though is nice as it reinforces our expectations resulting in greater comfort/confidence in our perception.

Wow, they were very interesting, thanks!

I still don't quite understand that "clever little clock" thing. Everyone was talking about it like it was a real advert, but there's no way it could be genuine. The advert even says "your clock may look different to the one in the photo... but don't worry, that's normal". It has to be a joke, surely?



But surely no intelligent person would ever claim to be able to tell the difference between two sound sources without blind testing them? That would just be plain idiotic.

I'm going to post what I was told in another forum, just so you can see it...

I said:

I was told:


and



Comments?

The same places that Vorbis plays, MP3 plays gaplessly, mind you.

And in those places that don't, it is no big deal. And I do listen to DJ mixes, Pink Floyd, live albums and the like.

I agree.

I used to base my encoding decisions in what was said here and there, before I took a hard look (erm, listen) and saw what was transparent to me. It sits around 140k for most modern encoders, so I went with MP3, which will play anywhere.

I do have lossless backups of everything, just in case. Tough I doubt that save for my computer be stolen (as it was last december), I will need to transcode again.

This is the "people listen differently when they aren't performing a critical listening test" theory. I've heard it before, but don't see any evidence to back it up.

Is perception really a learned response? I think that point is debatable on its own. I think the perceptual capabilites of two healthy humans is about the same, and is determined more by the physical capabilities (and limitations) of the senses rather than due to a person being cultured into hearing and seeing in a particular way.

DIGRESSION: The idea that the educational and cultural influences on humans determine how they perceive is a "cultural relativist" argument. Unfortunately still popular in post-modernist Humanities deparments. The post-modernist view is that "reality is an inherently cultured (learnt) experience". This is a load of crap, and research has shown that even primitive people, living on different continents drew pictures, or made musical instruments in very similar ways. This is because humans share the same (or at least very, very similar) physiology, hence we perceive the world in very similar ways, irrespective of different social, educational or cultural influences. When primitive people desired to represent the world they saw, they used very similar, or even universal techiques. African, American-Indian, or Australian Aboriginal cave paintings share many similarities. Such as the basic principle that "lines imply object edges".

As I understand it, you are proposing that the prejudice of lossy compression will impair a listening judgement when not listening under controlled conditions.

This is quite a convenient way of saying "that ABX testing may apply to you, but it doesn't apply to me". Which is another way of saying "I'm unique! I have golden ears! I can pick flaws in equipment better than anyone else!"

Alternatively, conceding that lossy compression can sound transparent is a way of admitting "well really my hearing is average, it is about the same as everyone else’s, I'm not unique, I don't have golden ears, I can't pick differences in equipment as well as I said I could, I was wrong, psychoacoustic principles actually work!"

Who wants to admit being average? Of course people want to say they are unique and brilliant. (Often when it isn't true)

But you have to remember that we are discussing modifications of the signal that testing could show are actually inaudible - not perceivable. They are imagined as an attempt to psychologically justify an inherent prejudice ("all lossy compressed exhibits flaws") and then just assumed as being present, even when there are no perceivable flaws. Isn't that by definition a placebo effect?

I don't think you've described anything more than a standard placebo effect. Or rather a nocebo effect. The assumption that some 'remedy' will actually make things worse. ABX testing accounts for either a bias for or against an encoder by making you pass judgement unaware of which file is compressed.

I agree that anyone is free to encode with any encoder and bitrate they want. But I don't think that makes their opinion equally valid to a broad consensus that is supported by empirical evidence. One opinion supported with some evidence is worth more than a million opinions made via prejudice, speculation, hunches, free association, or random thinking.

I'm not sure those expectations will actually degrade the sound. I don't think if someone thinks that the sound is degraded that the sound actually becomes degraded.

This is just a "peace of mind" argument. Well, I guess I do that myself. For my iPod I use iTunes AAC 160 Kbps VBR. I could probably use 128 Kbps VBR and rarely pick differences. However, I have all my music on my iPod and still have 12 GB free. So I decided to use the setting "one higher". This is a practical decision that works for me. That is different to saying: "all lossy compression is flawed therefore (without any testing) LAME 320 Kbps must be flawed". That is like me saying "I believe the sky is green, therefore without looking at the sky, I can tell you that today the sky is green".

At one level there is a logical flow to the statements, but if you start with a flawed assumption in the first part, then the second part becomes invalid.

I don't think people biased against lossy are against the term "lossy". They are against the idea that you can remove elements from a signal, but they still can't hear the difference. That is an affront to their belief that the purity of the signal is sacrosanct, and that they will be able to perceive ANY modification to it, even if they actually can't.

They are against the whole concept of psychoacoustics, they think it is a grand conspiracy against them, which has the potential to reveal that they aren't actually as good at hearing as they say they are. Psychoacoustics shows you can make use of the limitations of human hearing to compress a signal. To them human hearing, and in particular their hearing, is infallible.

Over shooting to me isn't the biggest problem. If someone encodes at 192 Kbps when 160 Kbps would be transparent, then who cares. Everyone has a right to encode however way they wish.

The things that I do think are damaging are spreading disinformation that doesn't equate with reality. And of course outlandish statements like:

1) Lossy compression can never be indistinguishable from a source file
2) Constant Bit Rate is always provides better quality than Variable Bit Rate
3) Stereo is better than Joint-Stereo

Of course there are lots more, but they all share the same prejudice: lossy compression is an inherently flawed enterprise that ruins every sound it touches.

Well, this may not be a good idea. MP3 isn't a good archiving format, since once you've converted something to MP3, there's no way to 'recover' the original file (and all its data).
This may or may not be iimportant to you. (Suppose a new, better compression format comes along...you wouldn't want to start with an mp3 and then re-compress that. It would almost certainly introduce audible artifacts.)




Not so much 'lies' as sheer -- perhaps willful -- ignorance. I would take everything 'audiophiles' say with a huge grain of salt. I've rarely encountered any who know what they're talking about re: digital or codecs. And many are 'religiously' opposed to scientific methods of testing their claims. Their standards of 'proof' become extraordinarily high, when you suggest something they believe isn't* true, but extraordinarily low for stuff they believe *is* true ("I heard it" is usually sufficient).



Consider using a lossless compression format, like FLAC, for archiving against future loss of the CD. You can always make an mp3 copy of a FLAC file if you need a smaller version of the file.

deleted