Nigel asked you if you could tell the difference in a double-blind test. You avoided answering that- twice! He drew the conclusion- as did I- that you didn't answer because it would require admitting that you wouldn't be able to tell the difference in a true double-blind.

The orange analogy is flawed. An orange is food- food has nutritional value (or lack of) and thus there is more to consider than just the pleasure (i.e. sense) of eating it.

Music- or rather, music reproduction- is *all* about how it's perceived and experienced by the listener. That's it. It's a means to an end. That's not to trivialise it, or make it shallow- arguably the opposite, as music is purely and only about how it is received in peoples' minds. If one genuinely can't perceive the difference in reproduction, then in effect, there *is* no difference.

Of course, other factors come into play as well; for some people, putting on a record and reading the sleeve while they listen may hold more appeal (and enhance pleasure- or even perceived quality) more than an otherwise identical reproduction music on its own might.

That's a perfectly legitimate feeling, it just doesn't prove anything about the sound quality in itself.

Similarly, someone might enjoy uncompressed music more than compressed simply because they *know* it's uncompressed- even if they'd be unable to tell the difference in a double-blind test. But the latter implies that if we were able to secretly swap the uncompressed source for the compressed one, then that listener would be unable to tell the difference.

Audiophiles have a notorious reputation for buying into overpriced pseudoscientific guff and coming up with excuses against double-blind testing that would require them to admit (as much- in fact, probably more- to themselves as to anyone else) that they couldn't tell the difference and their preference wasn't founded in "rational" absolute sound-quality.

Let me emphasise- I'm *not* claiming that sound quality and listening pleasure can't be improved by replacing a crappy, low-quality system with one that's audibly better. Of course they can. I'm saying that *if* an audiophile can't tell the difference in a properly-conducted double-blind test between two systems, then it's meaningless to argue that one has higher sound quality, regardless of any "scientific" rationale. Scientific instruments measure sound waves and the like, but they don't "listen" to music- which is its only (and noble) purpose- only people do that.

Despite the "128mbps MP3" having become the whipping boy for bad, over-compressed audio since the Napster era, ascribing this to the bitrate isn't entirely accurate.

The low sound quality of a lot of those early 128mbps MP3s had as much to do with the crudeness of some of the encoders that were around in the late-90s. I have some MP3s I encoded myself using a LAME-derived encoder (*) at 128mbps, and whether or not they're "HiFi" quality (probably not), they're still miles better.

It's also possible that some of those files were transcoded and/or re-ripped, which tends to magnify compression arifacts exponentially.

(*) Generally considered to be probably the best of its era

What do you think I have avoided!?

There is no need for any test to show you that the spectrum is the same shape - your pc can do that,
So what is the blind test and how would it measure the listener's emotional reaction to the music. Would it capture their tears drop by drop? Would it be wired to their nervous system?
I don/'t think you would eat the Orange I handed to you of which 80% had been removed without leaving any sign . I think you would be curious about the content of the missing 80%. But you seem to have no curiosity about the much smaller file size you are listening to even though - I am going to guess - you are unable to explain what has been removed.

You do realise of course that nothing like 80% of the information has been lost in the MP3 file especially when recorded at the higher rates.

Are you sure?
Adele Skyfall:
Flac: 28.07 mb
Wav: Probably around 56 mb
Mp3 (320 kbps): 11.43mb

I think you avoided answering the question regarding the double-blind test- that's what I already said.

Having read the following, I now realise this is because you didn't know what it meant, rather than avoiding the question...


A double blind setup is one that ensures an experiment is scientifically rigorous and free from bias.

It's ideal (e.g.) for comparing whether one sound source is better than the other and/or indistinguishable as you're *only* judging by the audio quality and nothing else.

Human beings are notoriously subject to bias, e.g. it's been shown that they prefer cola drunk out of a Coca Cola can to a generic one, even if the content is identical. Similarly, if one thinks CDs sound better than 320kbps compressed audio, then you're likely to be biased towards the one you know is the CD.

On the other hand, if you- and other audiophiles- can't tell the two apart when you (and the people running the experiment) don't know which is which, that proves there's no audible difference.


That's not what it's intended to measure. We were talking about sound quality and whether the "missing bits" you were aggrieved at not getting affected the sound quality.


That's because it'd be something I'd never seen or heard of before and couldn't figure out how it had been done, nor what the implications would be for me and my health. On the other hand...


...I already know broadly how sound compression techniques based on psychoaccoustic masking and frequency domain principles work, and have done for a long time. So I have a pretty good idea what the compression it entails and no need to be curious every time I listen to an MP3.

Also, I already explained why the orange analogy was flawed.


Well, I *do* know how it works, so there. I'm not about to spend hours writing you a whole blooming article just to prove I am "able to explain" what has been removed, though.

This might be a start, if you're interested.

I never get hung up on this. I can choose to listen to 45s from a choice of two vinyl jukeboxes, LPs, cassettes and CDs, via my vintage tuner/amp. mp3 rips and even YouTube videos downloaded as a 64k mp3 on it too, via an iPod. Though you have to pick and choose between the varying quality of some of the uploads.
The first track on this is one of those, given that this is just a small digital camera that will take short videos, I find the quality on some as this, considering it's just an iPod, is in my opinion, excellent, particularly the stereo separation. The second track is a rip from a CD.

http://www.youtube.com/watch?v=06jp__aN0gE

If it really is completely undetectable, why should I care?

Then it is of no use. We are not discussing sound! We are discussing music. The problem is not what you know. It is what you don't know....
Do you agree that Music has an impact on mood?
You do? Fantastic!
All that we need now is a test that will measure the impact of the two files on Mood/Emotion/Well Being.

I think you would fail the blind test with the two oranges because you would be unable to detect the missing 80% from taste and texture alone. But if you went on a mock orange diet (not impossible?) you might find yourself horribly vitamin deficient in 6 months.

And what would I be deficient of by listening exclusively to 320 kbps Mp3s? Come up with a proper test and maybe we could answer that.

Sound is the medium of music. The sensation of music is the perception of that sound.


Oh, is that "all"?

Hint; actually designing such an experiment to work in a watertight and reliable manner is likely to be bloody hard.

Actually carrying out such an experiment (which would inevitably have to study people over a longish period of time outside the lab) would probably require a *lot* of work.

Please let us know the results when it's complete.


Given that these are hypothetical "80% invisibly missing" oranges (*) whose qualities you are making up on the fly, speculating on what I might do is rather meaningless! You could have said "I think that the oranges will evaporate after six days due to their lack of mass", and I couldn't argue against that

(*) Not as widely available as satsumas


Er, yes. I believe I already raised this point when I said that "The orange analogy is flawed. An orange is food- food has nutritional value (or lack of) and thus there is more to consider than just the pleasure (i.e. sense) of eating it."

This contrasts with music where the sensation *is* the entire (and only) reason for its existence.


Why would *I* be expected to know? *You're* the one suggesting that there might be side-effects of listening to MP3s instead of uncompressed music! (Assuming you're actually being serious )


*You* come up with a proper test since it was your theory!

Oh my god, this is getting rather silly now

I did not expect you to cave in as quickly as that.


This is amusing. The analogy is flawed becasuse you say it is. If you are answering on behalf of Nigel, (or you are his sock puppet) please deal with the point about romance in post #21 . Don't tell me it's "another flawed analogy". . Just humour me.

You just don't get it do you?
We've done the double blind test. The listener was unable to identify the wav files from the 320kbps files. This is clear proof that the impact that music has on humans is not impacted - in any way - by the bit rate. Let's pack up and go home!

"But....he clearly heard the audio artifacts when we started out with a bit rate of......"

Nope. Not Listening! la la la lal la la!


We've done the double blind test with the oranges. Kodaz cannot discern any difference whatsoever in taste or texture. Lets put him on the Faux Orange Diet!
Apologise to Kodaz for secretly swapping the oranges - He needs to understand that all this was "for his own good".


I am going to assume that you do agree that music has an important impact on mood. But you can't think of a way to measure this. What a shame.

Er, nope. The study was *your* suggestion for the purpose of investigating *your* original notion that compressed MP3s have an effect on mood. That'd make it your responsibility, not mine.

As I said, such a study wouldn't be the piece of cake you clearly expect if one was to do it properly, but I'm sure you're up to the challenge! *cough*


Nope. I didn't ask you to take my word for it- I gave my reasoning...

"An orange is food- food has nutritional value (or lack of) and thus there is more to consider than just the pleasure (i.e. sense) of eating it. This contrasts with music where the sensation *is* the entire (and only) reason for its existence."

That's the third time I've said that by the way. Let me know if you need me to cut-and-paste that for you a fourth.


I'm not, and I'm not, so there.


I genuinely have no idea what you think you're trying to say here. Your already bizarre "missing matter oranges" analogy seems to have gone off the deep end...!


You got that one right- I've no idea what on earth you're on about any more...


Buried in this confused strawman misrepresentation and/or misunderstanding of my position, it seems that your argument is...

Because there's a clear audible difference between a low-bitrate compressed file and uncompressed, then there *must* be a relevant difference between the high-bitrate one and the uncompressed file, even if an audiophile can't tell them apart. But if that difference isn't perceptible in something whose only function is perception, what is the relevance of that supposed difference?!


Oh, I'm quite sure it's doable. Just that it would take a lot, *lot* more work to design and carry out than you clearly realised when you dismissed it as "all that we need now".

That's like saying one can turn a 7-year-old Vauxhall Corsa into the fastest car on earth, "all" you need to do it make it go 150mph faster.

And, as I said, I'm not sure why you think the burden of this should be on my shoulders anyway- your half-baked ideas, your responsibility to prove them!

You obviously don't understand compression. Just because the file is 20% of the size doesn't mean it only has 20% of the information.

Exactly! In fact, compression need not lose *any* information at all - not even information undetectable by human ears. FLAC for example has 100% of the information at 60% of the file size. I wonder how Landis would explain that one?

Overdone dynamic compression harms quality more than high bitrate mp3 compression anyway, IMO, and that happens on CDs as well as mp3s.

The issue of speakers is VERY important to this discussion. I am fortunate to own a 5.1 KEF surround sound Home Cinema system, routed through an Onkyo receiver.

I can report that it is capable of splitting MP3s of music that is 30 or 40 years old, and using it's electronic brain to work out which speaker the various components of the song should come from.Obviously, this is from stuff that was never recorded to be heard this way. The result is staggeringly fine.

Is it better with more up to date MP3s? Yes, most of the time, but not always. As for CDs and other supposedly superior media, personally I find that CD's are often a little too crisp and toppy. Sure, it's incredibly clean, but at the price of warmth and feeling. Therefore, most of my music listening is from MP3's streamed over a NAS drive, and I have no complaints.

Off topic here but can anyone tell me which I should be using when importing music to iTunes and then onto my ipod touch. AAC or Lossless or what ever ?. Thanks

But what is my motivaton for such a study! I am already listening at best available bitrate. The only person with something to gain is one who is listening to a self chosen lower bitrate. You/Nigel seem to be strong advocates for a "no difference" test which takes no account of mood. I am just pointing out the flaws.

I see. I don't need to glance at the top of the page to check which forum this is. You have called me a Strawman, and made up an argument which you now attribute to me. I guess the third element to this will be another patronising link to wiki?

And I did not come onto the thread to promote my own test or any other test. I came onto the thread to comment on the introduction of double blind tests (to the discussion) by amother forum member.

You may well be correct. I have only pointed out the difference in the file sizes as a basis for discussion.

FLAC is unbelievably impressive. I could not even start to explain how that could be!

The choice is yours really.

Apple lossless will give better quality on your computer and will play on the ipod, but probably not on other devices.
AAC or MP3 are much more widely supported format, but are lossy.

But you'll really only be able to tell the difference on a Hi-Fi or with a good set of headphones on your ipod, but even then it's arguable. As can be seen from this thread.

By saying that you implied that the lower bitrate was lower quality... that's what the listening test would be designed to find out.


It's still genuinely not clear what you think you're "pointing out" at all!

Were you suggesting that

(a) Mood affects perception of sound,
(b) Reduced bitrate sound affects mood, even if one can't consciously tell the difference,
(c) Something else


No, I accused you of making a "strawman" (i.e. intentional misrepresentation) of *my* position with that bizarre dialogue. The link I gave explained what "strawman" meant, and you still misunderstood.


That was my genuine best guess at what argument you were (badly) trying to make, hence "it seems your argument is".

If I misinterpreted your ramblings I apologise, but in that case I have no idea what the **** you *were* trying to say there!


It's not my fault that you don't understand (or lack the initiative to even look up) terms like "double blind" that are common in discussions of music quality.


You admitted that you didn't know what the term "double blind" meant, but replied to a question about it anyway?

Ah - That's interesting, becasue I don't recall admitting any such thing. I do seem to recall asking you to explain the test in your own words - and to explain to me how it would measure the aspects which are important to me...


Let's go back to the basics of what we are discussing.

I questioned the validity of a test which took no account of Mood/Emotion/Well Being. (Yes - I think that music impacts on these things over a long period. It would be interesting to know how bit rate affects this)

You responded by telling me that such a test would be lengthy and difficult.

If that is the closest to acknowledgement that we are going to get - I am more than happy with that.





ps - I almost forgot. You invented this particular strawman misrepresentation! Remember now?
Just unbelievable....

You said "So what is the blind test". I took this to mean you didn't know what a double-blind test was and answered accordingly

If you were actually asking me about the setup, well, you didn't clarify that.


That's because it would be; "Mood/Emotion/Well Being" are things that would be affected- and have to be measured- over the long term, outside the lab and are far less straightforward to gauge than sound quality comparison tests. Such an experiment would have to be rigorously designed to be scientifically watertight.

And when you say that "I think that music impacts on these things over a long period", remember that we were discussing compressed (but high-bitrate) audio versus uncompressed audio, so the implied issue is do *imperceptible* differences in quality have a subconscious effect on mood over the long term?

Bear in mind that the "uncompressed" CDs you like are still 16-bit, 44kHz representations of sound, and some argue not as high quality as (e.g.) 24-bit, 96kHz. In fact, since- in theory- sampling directly at the higher resolution/sampling-rate then discarding bits to downsample it to 16-bit/44kHz should give the same result as sampling directly at 16-bit/44kHz, then the latter could be considered "lossy compressed" relative to the higher-bitrate source, or analogue original.


No, I don't. Please quote or link to where this supposedly took place.

Now - Let's remove the part that you invented based on the snippet of dialogue from the imaginary double-blind test.
What's left?

I wouldn't say I was "warming to you" but I am not particularly unhappy about this paragraph either.