AIFF or Apple Lossless

[The Computer Audiophile]

Hi Izzznet - I disagree with you. 1+1=2 because we have all agreed it equals two as a way to simplify many things we do.

 

The Earth was once considered flat and those who didn't have the knowledge to prove it were ostracized. If the round-earthers just took the word of those who said, "The Earth is flat and there is no possible way it can be otherwise" it would have stifled all progress into the matter.

 

Bringing this back to audio and lossless v. AIFF. When people I know who listen to music for a living and master much of the music we all listen to say they can identify the differences between the file formats I tend to believe them. Then, when differences are measured between these file formats during playback I take this more seriously. Even if the measured differences are not audible to many people I still think it discredits the arguments of those who say there can be no differences between lossless and AIFF. If these formats are truly identical then there can be absolutely no measurable difference.

 

I don't think the onus to prove anything is on one side or the other. If we are all looking for an answer we must consider all sides and try to prove and disprove both of them. For example, those who discount all possibility that there can be a sonic difference between lossless and AIFF should conduct the tests and supply measurements that show there is no difference. This would be wonderful and be very helpful for all involved. If we had a collection of data points people could make up their own minds based on these data points and their own experience.

 

I think your reference to the dark ages is appropriate to both extremes of this issue. People the extremes reflect the dark ages by discounting the possibility that something other than what they know could be real. Those who say there has to be a difference based on what they hear and shut out all other data are equivalent to those who look at a lossless v. AIFF on paper and conclude there cannot be a sonic difference and shut out all research into anything else. Both of these extreme views don't help anyone including those who hold the opinions.

 

Both sides of this issue are losing ground by using current methods of discussion. If anyone is truly interested in finding an answer, or if anyone is 100% positive we already have an answer, they would put forth information for all of us to read and refuse to stoop to a 3rd grade level of discussion.

 

[The Computer Audiophile]

krabapple said "And our host labels *me* sophomoric (/rolleyes)"

 

 

I stand by my assessment of your comments and have yet to see any data the would lead me to believe otherwise. Here is the definition I use.

 

Pronunciation - säf-'mOr-ik

Adjective

sophomoric

1. of, relating to, or characteristic of a sophomore

2. conceited and overconfident of knowledge but poorly informed and immature

 

http://en.wiktionary.org/wiki/sophomoric

 

 

[flatmap]

HarryHWombat, nice posting and this pretty much sums up what I was thinking when Apple Lossless first became available. Well no, actually you're saying it much better than I was thinking it.

 

I would never have given it another thought except that there were a bunch of files (all classical as I recall) which just never sounded very good in my system. As I said elsewhere, I had taken to just skipping over these when they came up in the playlist; they just didn't sound worth playing. I came to use my computer audio setup mostly for pop and jazz -- and played classical records in the evening on my turntable. Ah well, I thought that was just a matter of getting a DAC. Or something.

 

However after reading something about AIFF and ALAC on the internet, I converted some of the problem files and they became quite listenable -- even enjoyable. So it was a big difference for me.

 

Maybe this all comes down to having an old computer. This is a 5 year old machine, an outdated processor, a slow system bus, slower HDD, and 512MB of RAM. Yes, I know, somewhat laughable... but as I grow older I've come instinctively to have more compassion for outdated equipment. :-)

 

Anyway, trying to figure out what's going on, my best guess is that file copy, floating point benchmarks, frame rate throughput and etc. aren't necessarily sufficient measures of quality for audio production.

 

For example, in the file copy you mention, although it's dispatched quite quickly, there's no guarantee (and no need) for it to execute uniformly in time. If the process converts half the file in a nanosecond, it goes into a wait state for half a second, and then finishes the other half in a remaining second -- well no problem. We still get the resulting copy done in less than 2 seconds. And like you say it's probably happening a lot faster. But if the process generating audio gets banked into a holding pattern, what is the the dac supposed to do? Maybe it has a good buffering scheme set up?

 

Anyway I have no expertise here, but I guess something like that must be happening. So until my ship comes in, I converted to AIFF... and I'm very much thinking about investing in seriously more RAM.

 

[izzznet]

Chris said: "1+1=2 because we have all agreed it equals two as a way to simplify many things we do".

 

So according to your logic, if you can persuade everyone that the world is flat, and we all agree with you that it's flat, the earth will be flat again.

 

No, the truth is, we'll have returned to the dark ages, having chosen to become a deluded herd of idiots.

 

On that note I'm out of here. I wish you well in hearing all the differences you choose to hear in whichever identical series of bits you choose to compare.

 

Izzznet.

 

ps. I'm not even going to bother with the "I know special people and they ..." paragraph (i.e. the 3rd one). Except to say that's it's merely a dogmatic device intended to distract the recipient from the fact that the person speaking has absolulely no proof / evidence whatsoever, suggesting instead they are an "authority by association" and thus deserving of special treatment whereby any demands for evidence or proof can and should be waived.

 

This seems to be a proof free zone.

So, happy yapping!

 

[markr]

... and I wonder - really wonder. Some here have said that they cannot hear this difference between flat PCM and Lossless at all. Fine. Some have said that they hear it some of the time but definitely not all of the time. Fine too. Then there are those that say they hear it all of the time. (C'mon, really? That isn't my experience, but I'll admit that it is possible.....) ... It seems to me that there are enough differences in hardware configurations, software (encoding and decoding/playback) to make a real mess of any causal analysis of the end effect of what each camp could be hearing. I don't know what the answer or solution to any of this might be because there are many many variables here. Too much is left out of the equations here for a real analysis to be made, I believe. We must be more precise in what we say.

 

Looking at the thread on Hydrogenaudio that was begun by AV_OCD from here at CA - and continues now because this thread continues ( http://www.hydrogenaudio.org/forums/index.php?showtopic=70950&st=25 ), I see that there is some fragmentation and some questioning of the reality of actual proof that the playing back (implementation) of lossless is truly related to the comparison of (non audio, just file comparison) of lossless files vs. PCM. While this is a hard thing at HA, this is as we here think it should be at this point. We aren't debating or arguing the data comparisons, we are debating the audio comparisons. This is not going to be an easy thing to resolve. Everyone, HA members included, please try to be civil and circumspect in what you say on each others forums.

 

I personally applaud the efforts that hydrogenaudio folks make in the areas that they make them (these are manyfold, people...). I find it unfortunate however that some individual members of HA come here with derision and abuse in mind and action. When things such as what we are discussing here happen, I would recommend a civil and logical response over an emotional one every time. .. EVERY time. I join with Chris and Poo in hoping that we may find a dialogue between us that will be immune from such tactics.

 

regards to both communities - markr

 

PS: Hey krabapple, why did you find it more important ( I'll give you the benefit of the doubt this time) to quote me at HA on my saying that you made me feel like I was baby sitting the nephews, than saying that I 100% believe that for storage, downloading and conversion to PCM, I found lossless transcoded to PCM to be bit perfect when comparing to original PCM? That was kind of strange to me.

 

[HarryHWombat]

Codifus hi:

 

I thought that the AIFF (or WAV, for that matter) would cause increased CPU usage because itunes has to convert it to ALAC then send it over the network. With ALAC files, it just transports it over like a truck, no processing.

 

Good point - I must think more clearly before replying! Actually, this is an interesting case as there is then no such thing as native AIFF replay over wireless if it converts to ALAC first.

 

Flatmap hi:

 

Maybe this all comes down to having an old computer. This is a 5 year old machine, an outdated processor, a slow system bus, slower HDD, and 512MB of RAM. Yes, I know, somewhat laughable... but as I grow older I've come instinctively to have more compassion for outdated equipment

 

RAM is extremely important. Modern OS X will not really operate at all effectively with 512MB. I had 1GB and (I think because of the external DROBO but I have no proof) and it crawled. I upp'ed to 4GB and it flies. Additionally, the old "G" processor machines are not very good at integer arithmetic (as I remember) so if any of the coding or decoding involves a lot of integer arithmetic this may be an issue.

 

I still stress the point that unless errors accumulate in some way in a DAC, for a file to sound different there must be consistent errors throughout the whole file. I don't really know anything about drivers or their policies but one can imagine an audio driver having a "prioritise getting something out in time even if it is wrong rather than waiting for the next bit" in which case I can imagine it sending the last 16 bit word to the DAC and then discarding the correct one when it arrives. If that makes any sense. As I said, I don't know and sometimes speculation as mine is above is more dangerous than staying silent. On my computer with the speeds I see I cannot see anything that could go wrong that would cause any of these effects.

 

If this was running on a BBC Micro, Commodore PET or Spectrum then .. maybe

 

[Gag Halfrunt]

There seem to be a lot of drive-bys on this thread. I guess the guys at HA are having a good laugh at our non-scientific expense. The format seems the same in each case: Drop by, make some statement that challenges the perceptions of the forum, watch the reaction, then post some high-handed 'goodbye' note. Irrespective of the force of the argument raised by these people, it hardly benefits their overall argument.

 

"I'm right, I have science on my side... ner, ner, ne nyer ner!" is never going to find you friends and tends to harden the authodoxies of those who take a more subjective stance. That being said, the over-reaction (for example, 1+1=2 is consensus-based) is just that, an over-reaction. It's a weak defense that leaves you open to grotesque conclusions (if 1+1+2 because we all agree on it, is this forum posting sent by applied science, or by wishing really hard?).

 

Thing is, I do understand the arguments made by the hard science lot. Trouble is, although their argument may work on an intellectual level, it doesn't on an emotional one. Here's why; I remember the double-blind ABX testing of PASC (Precision Adaptive Sub-Band Coding), used to demonstrate the advantages of the short-lived Digital Compact Cassette. I also remember similar tests of ATRAC for MiniDisc. In those tests, PASC and un-compressed PCM off a CD were impossible to differentiate, as were original ATRAC and CD.

 

Why do these two tests pose fundamental problems for me? Because both have subsequently been downgraded as newer formats emerged. PASC has limitations in the sub-band coding mechanism that would rank it alongside 96kbps MP3... and yet in the early 1990s it was 'identical'. ATRAC-1 was announced with a slew of peer-reviewed papers (using double-blind ABX tests) claiming how it was transparent in ABX tests, yet the improved ATRAC-3 was found to be less transparent than Ogg Vorbis and more.

 

The PASC and ATRAC issues serve to reinforce my watching brief on blind tests. Back in the 1990s, these things were 'transparent'. Today they aren't. What does that say for double-blind ABX tests? I am saying this from a position of being 'test agnostic' - it's not that I want to discredit blind-testing, I genuinely want to find out what's going on, and why.

 

There have been a history of blind and double-blind tests in audio, dating back to Edison. Granted, some of these tests have been closer to PT Barnum than good scientific method, but some haven't; in all fairness, I haven't been able to find the paper (not an AES member and $20 per paper gets a bit pricey), but I've been repeatedly told of robust double-blind ABX tests conducted in the early 1960s by Cambridge University, the Audio Engineering Society and Quad Electroacoustics that managed to determine that a open reel tape recorder, valve pre-power and a pair of electrostatic loudspeakers were functionally identical to the original live sound.

 

So, while I don't dismiss the concept of double-blind ABX tests at all, I'm not exactly sure how robust they really are. Could it be that what we can't hear today, we can hear tomorrow? I would hate to adopt a 'transparent' methodology now only to discover it wasn't really transparent in a couple of years time. As disc space is cheaper than sleepless nights, I'll go with the path of least worry.

 

And I would love to post a similar point on Hydrogenaudio, but it won't have me as a member. I wonder why?

 

[John]

Well put Mr. Halfrunt,

 

Imho, the drive by's are not only sophomoric but have become soporific. I can't be bothered taking the bait.

 

I share your concern about abx protocols in this area. I have found that the greater satisfaction of one format over another come over a long period of listening. For better or worse I tend to prefer vinyl very marginally, even to hi-res digital but I can't point my finger at exactly why in a quick A:B comparison. I just find that I listen to more music, for longer, when I play from preferred formats. So, I have drifted into those preferences long term due the vaunted and vilified 'emotional impact' factor. An abx test that captured the long term satisfaction of various formats and systems in a selection of environments is imaginable but, I fancy, completely impractical and unaffordable to conduct.

 

As for AIFF vs ALAC, I don't know if I can tell the difference because I never tried to compare. I don't much care. But with disk space as cheap as chips I'll stick with AIFF because life's to short to worry about the smaller files that ALAC would give me. (As long as nobody claims that lossless actually sounds better of course.)

 

- John.

 

[Codifus]

What type of G4 is it? 400, 500, 867 Mhz? Also, are you running itunes 7 or 8?

 

The reason I ask is because I had an interesting experience with my 1st music server. It was a Powermac G3 at 350 Mhz and iTunes 6. Apple introduced 256 kbps downloads in the store and forced me to upgrade to itunes 7. Itunes 7 on a G3 sounded awful. The music played but something just didn't seem right. I then found out that the iTunes system requirements for itunes 7 was a minimum G4 500 Mhz. I upgraded my hardware to a G4 733 and all was good again.

 

On a slightly different note, perhaps this issue with ALAC and AIFF could be made more apparent with weaker hardware. Exxagerate for effect. I don't have any of the old hardware, but if someone could put itunes on a G3, playback lossless and the equivalent AIFF files while watching cpu activity, net usage etc, the problem might make itself more apparent on a slower machine.

 

One last thought on this, I find it strange that Apple specified that iTunes 7 needed a minimum of a G4 500 Mhz to work properly, but now iTunes 8 specifies a G3, G4 or G5 processor as well as intel, of course, and does not specify cpu speed at all. So now itunes 8 was written with more efficient code such that it could happily run a G3 whereas itunes 7 could not? I find that hard to believe. Apple's software team must be writing the code for the intel, then adjusting that code appropriately to run on PowerPC G5s and maybe G4.

As with any company, I can't imagine their software team to have the resources to write a program and have it optimized for every hardware platform. Why not then make itunes 8 run a a Motorola 68040, Macintosh Quadra 950, and Mac OS system 7.1?

 

 

CD

 

[Wavelength]

Steve,

 

A few things...

 

When going WiFi to an Apple TV or AirPort express, the difference is obvious to me and some of my customers. This has nothing to do with the clock or jitter from the PC.

 

These are two radically different devices. The Express uses the PCM2705's SPDIF output and as a DAC. That means the AE controller converts to USB then outputs to the PCM2705.

 

The AppleTV has a lot higher grade processor and 24/96 output capabilities.

 

It is simple to understand the differences between the two. Also all AirTunes only transfers at 44.1/16 as far as the protocol and what format it's sent in... I don't know. I have the source code here somewhere I just don't have the time too look at that.

 

Either way these are both pretty poor performers if you are considering high end.

 

As far as the ALAC iTunes ripper on the PC... well I don't know. I tell everyone that on the PC you should use J River I think it's a better program. If you are invested in iTunes then make sure you are using Vista as it makes a huge difference. Annoying... yes, better sound, yes.

 

Thanks

Gordon

 

[glt]

If I remember correctly from my experiments, ATV only supports 44/48 when playing audio...

 

[soooowhat]

 

Markr,

 

Thanks for posting the link to the HA thread.

 

As someone who doesn't fall into either camp = obj or subj = I found the responses on that thread rather hilarious.

 

In short, they haven't done any ABX analysis on the issue about which they show so much certainty - AIFF versus lossless.

 

Indeed, there seems to be no way that they can do so, if I am interpreting what I read correctly.

 

BUT, that doesn't seem to sway them in the least - the argument seems to be - there is no THEORETICAL basis for playback of lossless files sounding different than playback of AIFF files, therefore we DON"T NEED NO STINKIN' ABX to prove our point. "We've proven it 'scientifically' without even need to listen" - they seem to be saying.

 

IOW, when all else fails, even the most 'objectivist' group (I use that term loosely) relies on subjectivity (i.e., their own opinions and an agreement to agree with each other that ABX is not warranted).

 

Most don't even seem to appreciate, or mind, this discrepancy. Only a couple of posters really seem to get the quandary and their response is to suggest that they modify the charter of the group to effectively address situations in which (they all agree that) ABX is NOT required to prove their point. :-0

 

As I said, quite hilarious.

 

 

 

 

 

 

 

[markr]

.. welcome to Computer Audiophile!

 

I thought the link would prove interesting to some. Speaking of proof though - we still don't have anything but 'ears' on the thing yet. It is up in the air as I pointed out and the methodology to be used to get anywhere with the thing is unclear. This is further complicated by the fact that actually getting a good grip on what is going on, isn't going to make anyone any money by doing so. That could make getting a real answer quite slow, especially in these tight-budgeted times .....

 

You know though, I could plow through a couple of threads on CA here and get similarly 'hilarious' examples of thought or testing processes. SO ...... let us let logic, civility and circumspection be the rule in our communiques. For instance, I don't really find it hilarious that this is being so painful for members at HA. I more find it gratifying that they themselves see that they don't know it all. I myself am more from the school of "the more I know, the more I find that I don't know". How else to learn?

 

Thanks for the thanks.

 

markr

 

[BEEMB]

 

Gordon,

 

Your comment regarding iTunes on XP is not strictly true. It depends on the soundcard, drivers and certain conditions. The K-Mixer can achieve a bit perfect output when certain conditions are met and I believe you can find confirmation of that within these very pages. As soon as the K-Mixer alters the sound then you're in trouble.

 

Lossless vs AIFF .. This will go on and on until somebody produces some results. Anyone going to volunteer?

The stereophile link I posted earlier compared one form of lossless, FLAC, to uncompressed and showed there was no difference.

 

Somebody try ALAC vs AIFF please ...

 

Matt.

 

 

[soooowhat]

 

 

Markr,

 

thanks for your welcome, altho I've been lurking for a couple of months or so.

 

Yes, I'm sure there are views that either camp will find hilarious, even here at CA, although Chris does a really good job, from what I've read, of trying to minimize the vitriol.

 

Clearly, those with agendas and/or attitudes need not apply. Worse still, when the 'agendas' are more like religious dogma, as was pointed out earlier by another poster.

 

As for proof, well, don't hold your breath. As for me, I'm not looking for proof, just practical, pragmatic insight that I can apply myself. Regarding AIFF versus lossless, that would seem to fall neatly into either of two categories - if add'l HD space is relatively cheap (for you) and/or you're a 'rather be safe than sorry' temperament, then rip to AIFF. Otherwise, use lossless.

 

I will say this about my experience with the obj vs subj debate - in my experience, it's the objectivists who get overly worked up when someone has/shares an opinion different than theirs. They believe that their arguments should over-ride the opinions of others, when in fact, even with all their ABX they cannot disprove the possibility that others can hear things which they could not. IOW, they cannot prove their arguments with the same veracity that they attribute to them.

 

Given that they will never be able to convince subjectivists to go through the trouble to apply their objective 'tests' to prove themselves (subj) wrong, they will NEVER be able to prove that it is impossible for them to be wrong - short of testing the entire universe. And I think this is a source of great frustration to them, frankly. It seems to me, they want to be seen as infallible. Ain't gonna happen.

 

Like I said, don't hold your breath waiting for proof. It's never gonna come.

 

"When in the company of mad men, keep your own counsel."

(note: I can't remember the attribution, or the exact wording, of this quote).

 

 

 

 

 

 

 

[soooowhat]

 

Matt,

 

Don't hold your breath. From what I read at HA, they don't even see the need to ABX it over there, and they seemingly test everything. And those that considered it, don't think it's possible.

 

And then, those that don't use ABX tests are not likely to change their stripes just for this debate.

 

And besides, what would you gain by having those who don't see the need to test it actually testing it?

There would be no integrity in the sense that you'd have people doing the testing with NO motivation to hear the difference. Care to predict the results?

 

soooo what

 

[BEEMB]

 

Soooowhat,

 

I think both sides make relevant points in this discussion. However, I don't believe that such a discussion cannot be proved to be correct either one way or the other.

 

To avoid even think about it I do rip my music uncompressed but download a fair number of AAC's from iTunes. If it's something I really like I may then buy the CD album to rip. (AAC's of course are proven to differ from uncompressed though hearing that is probably difficult at higher bit rates).

 

But, it can all be measured and proven one way or the other. I'm just a little too lazy right now to have a crack at it. < posts ... sit's back ... plays next uncompressed track >

 

[markr]

Correct me if I'm wrong, but Gordon was talking about VISTA and iTunes, not XP. There is no "kmixer" in VISTA. I know that there are still resampling issues with VISTA and audio, but no kmixer anymore.

 

Re: ALAC vs. AIFF - As soooowhat said: don't hold your breath. That sentiment is what I've been saying all along here, but folks want an answer now ( I don't feel that need ) - don't know how else to say it other than: This couldn't be a priority to folks doing business, and we won't get the start of an answer without those with the means to measure having the time to spend resources on it. The best we could do from here is to do more ABX testing. I'd encourage anyone with questions about it to do their own. I don't need convincing that something is going on. You, on the other hand could find no difference. I trust you Matt, so I don't think that ABX will get us anywhere. So...... we seem to be back to measurement. (I'm NOT pushing here, Gordon).

 

Patience is a virtue,

markr

 

[BEEMB]

 

Mark, Gordon said "make sure you use Vista with iTunes", which I read to mean "Vista over XP", hence my comments. In the right setup you can have iTunes sounding great on XP too. (I don't use it but have conducted blind tests (too few) against MediaMonkey with ASIO). Using ASIO removes the need for any speculation over whether its (k-Mixer) bit perfect or not, ... just iTunes is so nice to use.

 

RE the comparison - science and measurements are what I'm actually after instead of ABXing. I think stats speak for themselves. So it will be interesting to see what anyone comes up with but am 99.99% sure measurements will prove there to be no difference - - - though if there is a difference that's cool by me too.

 

Currently listening to some rather good music, stuff I've not heard before: The Notwist - Neon Golden. Rather good stuff. Spent my entire evening hunting down new stuff. (new to me)

 

Matt.

 

 

[markr]

I didn't think it was possible to use ASIO with iTunes. I too prefer the iTunes interface.

 

I just downloaded the HDTracks 24/96 sampler. That's where I'm heading right now - off to listen to (gasp!) FLAC on the Mac.

 

markr

 

[soooowhat]

 

matt,

 

if it's science you're after, then you're all set.

The 'scientific' argument is that bits are bits and so long as the lossless file can be re-created bit-for-bit then there can be no difference.

 

That seems to be the rationale that the folks on HA use for not even bothering to use ABX.

 

As for statistics, I'm reminded of the oft quoted phrase 'there are lies, damn lies and statistics'.

caveat emptor if you're basing anything solely on statistics.

 

as for your belief that it can be proven to be ONLY one way or the other, perhaps you mean to say it can be proven to your satisfaction one way or the other. To prove that playback via lossless files can NOT POSSIBLY make a difference (as compared to playback of AIFF files) would seem to be an order of magnitude more complex than simply proving that a file is bit perfect after lossless compression and then un-compression.

 

hint: CANNOT POSSIBLY does not mean, e.g., in all cases in which ABX was used, or even, for the majority of instances, it means that it can never happen....which is very difficult to prove. I don't see the need to prove that, BUT, to your point, that is what is required if you want to prove it can only be one way and not another.

 

AFAIC, until/unless it can be proven that it cannot possibly make a difference, I'll be ripping to AIFF, as the penalty for ripping to AIFF is not that significant.

 

 

soooowhat

 

 

 

[Axon]

I'll take the blame for linking HA to here, although I don't feel *that* sorry about doing so. I am flattered that krab mistook me for somebody else, though. I would like to make some comments that I hope are productive.

 

I've put a lot of thought into this, and I hope people here find much to be learned in the HA topic paralleling this one: http://www.hydrogenaudio.org/forums/index.php?showtopic=70950&st=0 .

 

I'd like to try to dismiss the belief that this debate is all about ABX testing. It's not. Fundamentally, it is about the fallibility of perception, and how one communicates with others under that fallibility. Quite simply, if one's hearing is infallible, subjectively hearing a difference between FLAC and WAV means that a difference must exist, and one can communicate with others based on that. But if it's fallible, there are a multitude of reasons why such a subjective difference can be chalked up to placebo or what you ate for breakfast or what have you - and thus, rational, meaningful communication of audio perception requires either numeric measurements (objective testing), or statistical validity (ABX testing). Otherwise, why should anybody believe what you have to say? Your belief on the subject, more than anything else here, will dictate your belief on FLAC vs WAV, power cables, and virtually everything else in audio.

 

Everybody in this debate quite obviously lines up on one side or the other on the fallibility of hearing, and it's pretty obvious where I stand on this. I won't labor that point too much, since most here have already made up their minds, except to say there is a Olympus Mons-size mountain of evidence in fields of psychology and observational science to not take your perceptions for granted, nor those of others, and especially not those of "experts" on the subject - self-proclaimed or not. Whether or not ABX testing has been used in the past to sell a pig in a poke is besides the point. Even if it is, naked subjective testing is worse.

 

The other facet of this debate is that, quite simply, there's nowhere for a meaningful difference between FLAC or WAV to hide in a computer playback system. Thus the claims of "impossibility" for this to happen, vs speculations on the nature of the difference, where it can be if it can't be in the datastream, etc. Again, I make a basic outline of a proof as to why a meaningful difference is not possible in the HA thread. As a degreed electronic and computer engineer, it's enough to convince me (but I've been known to be wrong sometimes). And I haven't seen a hypothesis in this thread that passes muster.

 

soooowhat, I would reread my comments on HA as to why it's not so good of an idea to ABX this (a lot of HA people still do not understand this). Quite simply, if you withhold the assumption that "bits is bits" in order to construct an ABX test to compare the sound of FLAC to WAV, my opinion is that you will never come up with an ABX test that truly does away with that assumption. So that, in the eyes of the naysayers, it will never truly be a valid test. In such an event - especially in light of the complete absence of a valid hypothesis of effect in the first place - I don't think it's fair to try to ABX it. At least, not unless extremely specific constraints are made as to the scope of the test and its meaning, ie testing only the audibility of changing decoding algorithms, and extrapolating from there.

 

Note, I am in opposition with a few HA people on that general point.

 

The overall cost of being "wrong" on this particular issue is virtually nil, at least if you are running iTunes and can tag AIFFs and WAVs just fine. But this is in large part a proxy war for all sorts of other topics in audio, and if some guy tells looks down on me for not listening without (insert product here that claims to affect what is supposed to be a bit-perfect operation), you can bet your ass I'll be cutting him down to size.

 

[Axon]

ignore (adding reply as comment to thread rather than n-deep reply)

 

[markr]

I don't expect that you'll be hanging around though. That was not a Drive By - That was a literate and engaging visit. Thank you sir.

 

regards,

markr

 

[Axon]

My forum time is spread thin enough as it is, and I've directed more than my fair share of gros mots at some members of this forum. And there really aren't that many topics I'm interested in nowadays. Still I'll try to pop in if I see something interesting.

 

Don't look at me much different from krab; after all I posted the original link, and if I wasn't in as good of a mood, I would have almost as much schadenfreude as him right now But this HA-vs-the-world/"hi HA, what do you think of these idiots on this other forum?" thing is not exactly new. If I had any more time I'd write a f*cking playbook on the subject. And that is kind of unfortunate because, well, it radicalizes people. And keeps them from agreeing. ... With me.