More anecdotal evidence that ALAC ripping is broken

[audioengr]

http://forum.audiogon.com/cgi-bin/fr.pl?ddgtl&1234444992

 

Those that are ripping thousands of CD's might think twice about using iTunes to rip Apple Lossless, even with error correction on.

 

Steve N.

Empirical Audio

 

[AV-OCD]

I spent a fair amount of time comparing my former $7000 Classe CD player to the digital output of my Mac Mini playing Apple Lossless files of the same CD. While they did sound slightly different, I would not call one dramatically better or worse. If anything, I slightly preferred the sound of the ALAC files tough the Mac Mini. That Classe player was a beautiful machine and quite capable of making beautiful music, so to even have a $800 Mac playing compressed files compare to it was enough for me to sell off the player.

 

BTW - Where is the original anecdotal evidence? It seems like only a few people in the thread you point to feel there is a problem. Where are the others?

 

[Fuel]

Unfortunately there is no control point in this experiment that people have been doing therefore I cannot agree with you that ALAC is broken, only that ripping via EAC and ripping via iTunes create subjectively different results. Please don't spread internet rumors; it's crap like this that ruins our hobby.

 

 

 

[arin]

Here we are again!

 

As I told in one of my post in another forum on CA I spent quite a time comparing ALAC, WAV and Aiff files ripped with iTunes and EAC, and I don't mean by ear only but measuring with professional equipment and software and they always resulted to be IDENTICAL!

 

Many audiophiles keep on judging by ear only and they also keep on believing "rumors" especially when the rumors are alarming or weird: I just read on a very well respected hi-fi magazine that at CES most companies chose to demonstrate their gear with a computer or an iPod+Wadia just to be sure their precious CDs were not stolen!

They couldn't just accept the fact that a computer or worse (for them!) an iPod could sound the same if not better than a CD player.

 

Recently I recorded the digital output of several CD players (from 100$ up to 8000$) and compare the files with the ALAC I got from the same CDs with iTunes: again what I got in the end was bit-perfect identical outputs.

 

So please, stop spreading the rumors and listen to some good music.

Please!

 

Ciao!

 

Arin

 

[audioengr]

I dont expect everyone to hear this. The professional reviewers that have done the experiment have verified the same result. It's not a rumor.

 

It's just like Wow and Flutter in a Turntable or Jitter in a digital system. If you system has a lot of other sibilance, you probably will not hear any of these. It's the difference between Hi-Fi and real music.

 

Steve N.

Empirical Audio

 

[tfarney]

I apologize in advance for not reading the thread; I don't have time to wade through an Audiogon discussion. Can someone give me the Cliff notes? Was this "experiment" conducted scientifically, with any kind of control in place? Was there any evidence, other than anecdotal and opinion, presented by the participating audiophiles and professional reviewers? Were there any measurable differences documented? Were they within the known audible range? Were they verified by scientifically-conducted listening tests?

 

No? Until any of that happens, it is reasonable to assume that this is like so many other "experiments" in the audiophile community that have confirmed the audible superiority of the inaudible. And what my ears hear is more valid than the results of this experiment. I'm sorry to be this blunt, but there has been so much snake oil endorsed by the Audiogon community - and professional reviewers - that I will trust my ears, the data, even the broad, general opinion of the audiophile unwashed, far ahead of their "expertise."

 

And Steve, we hear you, but your last line of defense against whatever we hear that varies from your position is that our systems aren't good enough to reveal the dark subtleties you are fencing. You are not ignored; you render yourself irrelevant with your own argument. I mean no offense. You may be right that if I would only make the financial sacrifices necessary to have a system in my home that exceeds that of most audiophile magazine editors, most professional mastering engineers, etc, I would hear what you hear. I'll never know. And your position will never be tested.

 

Tim

 

[BEEMB]

 

I'm being lazy too, and am responding based on my thoughts on ALAC.

 

As a PC owner I WAS using iTunes to play back my music via the Airport Express. After some time I found I wasn't enjoying my music as I should.

 

My reason for taking the iTunes route was that, having owned a Mac, it seemed easy to simply import the rather large library and go from there.

 

However, I gave Media Monkey a try for a good few hours the other night (not given it much of a chance before). Having been told by Ash that the DAC in my ADM9.1s reclock the digital signal, I was happy to output digitally from my soundcard (using certain drivers) via ASIO (all ensuring bit perfect output) and found an improvement in the sound quality that, on some tracks, was pretty considerable. I'm enjoying my music.

 

My understanding is that iTunes, whilst getting around Window's Audio problems, converts your audio to ALAC before sending it to the Airport. On tracks that have a lot going on I found a difference between this option and Media Monkey.

 

So my point is, ALAC is not for me. I'm now selling the Airport. MM all the way the for me.

 

 

Matt.

 

[Fuel]

You could have just said 'potential subjective difference in some systems WAV vs ALAC; please perform an experiment to find what works best for you', please don't say 'ALAC is broken'; that is just pushing out more internet myths.

 

As a engineer I would have thought you would gone to the trouble of collecting some experimental evidence to back up your position.

 

 

 

[DanRubin]

If you listen to a variety of computer audio implementations, such as USB, Toslink, various SPDIF converters, etc., you quickly come to realize that bit-perfect does not tell the whole story. At least that's been my experience. If you look at the Sonic Studio/Amarra product, it's clear that they are saying that the playback software is critical and makes a huge difference.

 

Steve is only reporting what he's hearing from his customers. Why is he obligated to prove anything to anyone?

 

[Fuel]

I completely agree with the fact that playback software may make a difference. What I object to is the use of the language 'ALAC is broken' backed up with no proof.

 

[BEEMB]

 

I guess it could be Media Monkey producing the differences i am hearing.

 

Now, if Peter of XX High End could get that kind of library management into his software things would be interesting.

 

Peter???

 

Matt.

 

[glt]

Software bugs are deterministic, even the ones that "cannot be reproduced". Not being able to reproduce a bug means the testing is not good enough. If there is a bug in the encoder, then there should be at least one test case that will show that the result is not bit perfect, and if you show this to Apple, then they will put the developers to work and fix the bug.

 

 

 

[DanRubin]

...that bit perfection is not the issue, that the sonic anomalies are due to other factors?

 

[audioengr]

Matt - I doubt if ALAC is the problem with your setup. It's the AirPort Express. A device with very high jitter. This is why I invented the Pace-Car.

 

One of the customers of mine that reported the ALAC anomaly uses the AirPort Express with a Pace-Car. It even happens when using WiFi. I'm too busy to do these experiements. I just posted this as a public service.

 

Steve N.

Empirical Audio

 

[Fuel]

Could well be! Like in speaker design it has only become apparent over years and years of testing, real world experience and scientific observation that things like off-axis response are important to consider in design.

 

Maybe with computer audio there could be system phenomena that are influencing what people hear, such as: ALAC decoding requires higher CPU utilization than straight WAV playback, increasing PS supply current, increasing coupling of noise back into common system PS, therefore impacting analog amplification stages down the line.

 

It is only with hypothesis formulation and testing will we get to the bottom of this, and therefore have more idea of which factors need to be controlled to get superlative reproduction. Otherwise we are shooting in the dark.

 

[glt]

I have my iTunes on a Thinkpad T30 (probably 5-year old laptop). I decided to send ALAC and AIFF to Airport Express and observe the performance meter. In both cases, they were virtually the same topping at around 25% CPU utilization (it is an up and down cyclical pattern). In fact the high mark of CPU utilization had nothing to do with playing music because I also saw a similar pattern without playing any music. Thus I concluded that decoding music is a trivial task to any modern microprocessor (in this case encoding AIFF to ALAC to be sent to AE vs sending ALAC to AE)

 

I've also done some experiments with PS "noise". I have some tracks that are totally silent. When I play them (send them to AE) with the volume fully cranked up, I still hear nothing. My conclusion: Most likely I have no noise issues through the power supplies

 

[BEEMB]

 

Steve,

 

Thanks. But it's no longer in use anyway.

 

Mind you, should jitter be an issue. Ashley tells me that my DAC reclocks the signal and elimates jitter. So .. .. . . .?

 

 

 

Matt.

 

[tfarney]

I'm not in the business of doubting anyone's word, but I'm afraid I have to doubt perceptions when people are hearing audible differences where there is no reason for one to exist. Bit perfect is just what it describes, and it is all zeros and ones (or on/off) until conversion to analog. A DAC is incapable of reading noise, unless, I suppose, it is loud enough to be heard as "on," and then it would create an obvious error, not the kinds of aesthetic (analog, hifi...) differences that are described in these discussions.

 

By all means, listen to both and if you hear a difference choose what sounds best to you. But if you want others to believe there is credibility in that choice, yes, the burden of proof is on your side of the debate. It is easy to demonstrate that bit perfect is bit perfect, that exactly the same information is going to the DAC from both file formats. It's the other side of the discussion that is insistent, but without evidence.

 

Tim

 

[DanRubin]

I think we are mostly in agreement that bit perfection is readily achievable. The question is whether there are other variables in computer audio that might contribute to meaningful sonic differences between different hardware/software solutions.

 

As to the claim that this or that DAC re-clocks the signal so jitter is a non-issue, read Charles Hansen's white paper on the Ayre website to see why this is not -- or, at least, may not be -- true.

 

[markr]

... white paper there Dan. (Here: http://www.ayre.com/PDF/Ayre_MP_White_Paper.pdf ). While Mr. Hansen's white paper is certainly pointing more toward furthering the cause of Ayre's CDP product line in my opinion, CDP's after all do have DACs in them. That (DACs) is one of the things that computer audio folks, like us here at CA, are interested in. I don't know how much this has to do with this thread, though I suppose different hardware (DACs in particular) could have something to do with how lossless audio formats such as the ALAC "anecdotal 'evidence' " being discussed here are audibly perceived by various individuals with different DACs.

 

Very interesting paper indeed. Especially since I use an RME DAC (among others), and RME not only reclocks at the output in order to reduce jitter, but also uses an FPGA chip to do their DSP (digital filtering in particular for this discussion) instead of an actual DSP chip. Gotta go ask around the RME board now to see what sort of filtering is being used in the current firmware now....

 

On lossless audio codecs changing the sound: I can certainly hear slight differences when playing back *some* lossless files. Both ALAC and FLAC exhibit this to my ears. I cannot however, hear differences between my original CD raw audio rips and AIFF files that I have restored BACK from lossless files that I have heard these 'shortcomings' on. These same files returned to AIFF are bit for bit perfect. Weird, but true for me. I just won't mess with lossless or lossy compression if I can avoid it. I will readily admit that others don't experience any issues with lossless, and good on them.

 

I'd always recommend comparing several songs in raw (WAV,AIFF, etc.) audio formats with their lossless or lossy versions before deciding on which to use as your 'go to' format for your home server anyway. I have to believe that many of us have several different formats of audio stored on our HD's though. Just enjoy your music!

 

 

 

[DanRubin]

Sorry, I should have supplied a link. I was thinking of this white paper: http://www.ayre.com/pdf/Ayre_USB_DAC_White_Paper.pdf

 

[markr]

I think the info at the link I mistakenly put up is more germane to this thread though. This one is an advert mostly, though they do mention using Gordon Rankin's asynchronous USB (great), their own new 'super clock - placed where it should be' (yep, but should it be placed before or after the filter and which did they do?) and also the MP digital filter that is in the paper that I referenced earlier.

 

Getting back to this thread - I think the filter is the more likely culprit (if there IS one) here.

 

 

 

 

 

[tfarney]

"You may have missed the point

 

I think we are mostly in agreement that bit perfection is readily achievable. The question is whether there are other variables in computer audio that might contribute to meaningful sonic differences between different hardware/software solutions. "

 

Well it's possible that I'm wrong, as I've certainly been wrong before, but I don't think I missed the point. Software operates in the digital realm. It's final output, to the DAC, is digital -- zeros and ones...on/off. If they match the zeros and ones on the original file, it is a bit perfect. Identical DACs, reclockers and subsequent analog audio hardware should do identical things with the signal, regardless of the computer software or hardware involved.

 

I know that some smart people say that software "sounds." But I have yet to see any of them provide a rational explanation for how that is possible.

 

Tim

 

[The Computer Audiophile]

"I know that some smart people say that software "sounds." But I have yet to see any of them provide a rational explanation for how that is possible."

 

Hi Tim - Please take this in the spirit in which it's intended. I respect your opinions 100% and I enjoy reading them. However, in my opinion the one thing going against your very logical and down-to-earth opinions is your insistence that absence of evidence is evidence of absence. Everyone has their own opinions on the "sound of software" issue and I don't think anyone really has bullet proof evidence supporting these opinions. For now I think we are stuck using our ears and the subjectivity that goes along with this practice.

 

Again, I like 99% of your reasoning and think you are the voice of the logical common man among the many strong voices of those suffering from audiophila nervosa. Please keep posting in your usual fashion!

 

 

From Wikipedia

 

The argument from ignorance, also known as argumentum ad ignorantiam ("appeal to ignorance"), argument by lack of imagination, or negative evidence, is a logical fallacy in which it is claimed that a premise is true only because it has not been proven false or is false only because it has not been proven true.

 

The argument from personal incredulity, also known as argument from personal belief or argument from personal conviction, refers to an assertion that because one personally finds a premise unlikely or unbelievable, the premise can be assumed to be false, or alternatively that another preferred but unproven premise is true instead.

 

[audioengr]

This thread has gotten off-topic again.

 

We are not claiming that lossless CODECs change the sound and it has nothing to do with jitter, other than these things are more obvious in low-jitter systems. The claim is this:

 

The iTunes ripper that converts a CD to ALAC is somehow flawed.

 

If a .wav file ripped using EAC is brought into iTunes and then changed to ALAC, it sounds better than the same track ripped directly from the CD to ALAC.

 

Steve N.

Empirical Audio