AIFF vs ALAC with iTunes?

Dear Guy,

Obviously my thoughts concerning a PC/MAC have been a question of moving forward from an existing position. One of initally using an ancient [now retired] PC tower with USB out only. Thus I got a suitable USB DAC, and so the new PC was equipped with this in mind! My pervious experiments with optical and co-ax spidf disinclined me to go back to it after what USB brought. Not least that the clocking is within the DAC as it is seen by the computer as an off board sound-card, and the on board one is bypassed. One clock is better than two trying hard to synchonise. 

The ultimate USB audio implementation of today is the asynchronous system adopted by such [at the budget end] by Arcam. Rega do not yet have this, and consequently have optimised their DAC for SPIDF via optical and co-ax, whereas Arcam optimised for asynch USB. I did the demo on the Arcam, and liked it very much. In fact it is a something that is supposed to be suitable for use with a MAC as well as a PC, but I remember what Earwicker [EW of this Parish, but more or less lapsed these days] told me some time ago. PC is harder to set up for good audio, but if properly set up is not less fine than MAC is by default. Just that PC is cheaper, and offers the chance to go wrong!

In reality, with budget replay, all I do is discuss with those who know better, learn, optimise, and then see if I like it. If I do then that does for me.

I am open minded, but the budget is not limitless. 

With the ESLs and the Nait 5i-2 being such a blessed partnership, I know that part is not going to change, and it is quite fine enough to reveal changes and improvement at the source end. 

I'd say that ESLs have a way to make such as the SBLs sound a little crude., and yet those are regarded as fine enough for the Olive Naim topline. 

The front end is in flux as things improve, and I have no carved in granite view on exactly what will come in the future!

ATB from George

Dear Zinger,

I certainly did not do an examination of the CPU usage, but as reported above merely was struck - inadvertently as it goes and therefore not a coloured view, but one unforced by expectation - by a change for the better in the enjoyability of the replay. Nothing scientific, and as you will see reading the early posts, not immediately obviously or even definable by me via short A/B style swapping over.

It will involve some small cost and efforts in file type conversion over the next couple of months. I am not calling this day and night, but only something worth the effort.

I remember almost ten years ago reporting here that playing a Nero rip off the computer had a way of sounding better than playing the same CD in the computer. I was laughed at for this observation at the time, but I only reported what I found.

Sometimes life is stranger than fiction.

And it is quite free for anyone to discount - on scientific grounds - what I have found from time to time!

ATB from George

Dear George

It is quite possible the setup has optimised as much as is possible. I actually prefer XP for audio replay though it doesn't have as many settings availabe as Win 7.

In reality you are clearly getting a sound you are very happy with and thats what matters.

I do think moving to a Mac mini when funds allow would bring a performance enhancement. For one thing it is possible to use 'Bit perfect' to replay out of i-tunes on a Mac as Guy has mentioned, which does an excellent job. 

Of course if you do go for a Mac Mini your rips in AIFF should be ideal and you will have the option of an optical digital out which will widen your DAC options as well.

In the meantime I am sure you are enjoying your music.

regards

Geoff

Dear Geoff,

Thanks for your post.

Indeed, I have never had more pleasure from replay than in the last few years. I am quite open minded about exactly where the source/front end/computer actually goes. Having found such high quality from Radio Three on the internet, and iTunes ticking the boxes as well, I am at a point where for the time being, it will be finessing, rather than radical change. But in a few years maybe something quite different will come.

What is sure is that the amplifier and speaker situation is utterly sorted. 

All the best from George

Dear George,

Although the issues you raised at the outset have been sifted, I thought I might add a couple of observations. First, FLAC streamed to my UnitiQute does not sound as open to me as does an AIFF file of the same track ripped from a CD. My son, who is into computers in ways that I am not, thought I must be wrong. He pointed out (correctly) that an AIFF file can be compressed to FLAC (or ALAC) and then expanded to a file that matches the original AIFF exactly. Of course, that presumes no errors. However, we did the experiment and could not find one error across many iterations of compression and expansion. (The computer did the file comparisons, not my ears.) So, I conceded that point. He then suggested that he test me, by playing back the same tracks without my knowing which was AIFF and which FLAC. He bet me I could not tell. I won. (In all cases, files were streamed from a LaCie NAS across a wireless network to an Apple AirportExpress that is in turn physically wired by to the Ethernet input on the UnitiQute.)

With only the best,

Steve

He then suggested that he test me, by playing back the same tracks without my knowing which was AIFF and which FLAC. He bet me I could not tell. I won. 

Dear Steve,

I am finding that the difference is a little like the difference in CD playback between a CDS series machine and a CDX series one ...

Just as much energy, but just that little bit less forwardness. An ease and flow in projection. Nothing major in the short run, but valuable in lengthy and concentrated listening to a great piece of music. 

When I bought my long gone CDS2, I did a comparison with the then current CDX. No doubt the CDX was immediately more impressive because it was more forward and exciting. In terms of musical detail then there was nothing in it. 

My comment was that I'd prefer to take my excitement from the music than the replay.

It's subtle, but once discovered, impossible to ignore in terms of raising the game at only a small additional cost with my existing hardware. 

ATB from George

Steve, if your son is into computers, get him to think about power and bus decoupling in low level hardware design, and settle time consideration in low level drivers, then he might get an appreciation of why FLAC in the real world causes different side effects to WAV/AIFF.

We ate talking here about the differences between Computer Science and Computer Engineering. Designing and implementing a CPU from first principles makes you very aware of these issues, and makes anyone saying well its only 1 and 0s sound ridiculously naive.  

Hi George,

here is another piece of evidence for AIFF > ALAC:

http://www.avguide.com/forums/apple-lossless-or-not

This has convinced me in favour of AIFF, but I guess the difference in sound quality is not proportional to the difference in file size (well, I just bought a 512 GB SSD for my MacBookPro, so this should suffice for a while).

Best,

Stefan

Interesting Steve

However, in this case the UQ is doing the rendering and Naim systems seem not play compressed formats as well as uncompressed ones. I still believe that if a computer with a powerful CPU is doing the rendering and filling up a buffer in a Naim DAC with exactly the same data irrespective of the source file format then there is no difference in SQ between ALAC. FLAC, AIFF and WAV. Surely, if a DAC performs digital to analogue conversion on data in a memory buffer it is unconcerned what format the original file was in ... I cannot see how it would ever know. 

This is one reason I see no point in my buying the NDS as it is optimised for WAV - one format I avoid because it simply doesn't work for me. If I converted my files to AIFF I'd need more disc space than I currently have. Fortunately, I can't hear any of the reported differences in the way I have it. 

With my UQ in the office it gets FLAC to play. Perhaps AIFF or WAV would sound better, but it sounds OK with FLAC and I've no way to transcode on the fly. 

All the best, Guy 

Stefan - I would not call that evidence, it seems like somebody's opinion.

Why does he call FLAC .. "so called lossless" ... FLAC is lossless whatever the opinion on SQ. 

I'm still totally unconvinced a FLAC, ALAC or AIFF or WAV file sounds any different through my system

>  I just bought a 512 GB SSD for my MacBookPro

A very nice addition ... I really like booting from SSD as it runs like a rocket compared with the old HDD. I wish SSD was cheap enough for us to be able to buy 2TB disks - perhaps one day. 

All the best, Guy 

I am sorry to have only just seen this.

Thanks to Stefan Vogt, for the link and Guy for making the quotation from Stefan.

I read through the link and it seems that I am not the first to notice a difference, and the difference noted was the same. 

With a so-called "lossless" format like FLAC (or Apple Lossless, etc.), it is possible to restore the original, bit-for-bit (as proven down to the sample level in a null test) BUT this is true ONLY when the extraction is performed outside of real-time, i.e. NOT while listening but as a separate process performed by the appropriate software.

On the other hand, when extracted in real-time, i.e. while listening, there is a sonic price to pay as the result sounds harder and brighter than the original, un-reduced, linear PCM file. This is not a "night-and-day" difference and I can imagine some listeners on some systems not detecting it. Nonetheless, the difference is quite audible and on our system, I was able (and every one of my listening partners were able) to identify the reduced file within the first few seconds of play. This was in a blind, direct comparison with the original file. Everyone got it 100% of the time, with different types of music and playing the files in different software applications (some of which are pro level apps I use in my work).

I have spoken with some colleagues about this and their experience has been the same. It is much the same as the case with the best sample rate conversion algorithms: when used off-line (i.e. not while listening), they can be utterly transparent; when used in real-time (i.e. while listening), the sound gets brighter and harder. (I hear exactly the same characteristics with DACs and CD players that apply real-time sample rate conversion, and so, I avoid them and don't recommend them.)

It is strange, but among CD players, the only ones that I have found that did not seem "slightly" bright, and on some recordings actually more edgy than right, was the CDS2 and CDS3. I am not familiar enough with the CD 555 to say on that!

As I noted earlier, that sense is of slightly more distanced presentation, yet with all the detail naturally proportioned and clear, because there is not the slight brightness, which naturally makes the replayed music sound slightly nearer, but with additional detail, of course. 

Of course it is nice to find someone agreeing with one, especially when they said it first in an unknown source!

Mind you it is gratifying that the Nait 5i-2 and ESL 57s are good enough to show what apparently does not show on all replay systems!

ATB from George

James and others

it's true that itunes will do this but my experience (with a few files as an experiment) is that it will result in the original ALAC and the AIFF files being retained (as mentioned elsewhere). Sorry if I've missed the solution but, is there a way to easily delete the ALAC file at the time or after conversion without going through all the files, please.

If you have a proper folder/file structure, which I am aware is not the case, when one is relying on iTunes, manual deletion shouldn't be a problem. AFAIK there is no automated command to do this under iTunes.

In the music list - add another column with bit rate. Set the order so you can isolate the ALAC files and then select and delete.

James

I actually selected 'kind' in view options then clicked the header to separate them out.

Excellent - thanks for the hint/tip.

Mick

I am more than 45 pro cent  the way through converting ALAC to AIFF in ITunes.

I am doing it piece-meal, because my hard disc is not large enough to hold all the AIFF files, if the ALAC were still there.

Simply select about 100 or 200 track and convert. When they are done the selected tracks are still high-lighted in blue. Delete these and carry on and keep calm.

Is that called a work round?

ATB from George

Now 80 per cent or more done. [Converting ALAC t to AIFF].

One more evening, though not Friday evening  will see the task completed, and the results are surprising [ly, ... amasingly fine], so please excuse a review of the differences over the weekend.

I have no hesitation in saying that is you have a high resolution replay system, then the difference is all good, and if you use iTunes, then use AIFF, not ALAC.

ATB from George

The conversion is complete  and the space used has risen from ALAC's 157 Gbit, to AIFF's 345 Gbit, including the deletion of less than one half per cent of the contents as I went along. The effective capacity is 466 Gbit on the Hard Drive, so sufficient room for safety. Any significant extra music would really need to be on a large dedicated HD, I think

I'll write something about the musical effect of this change. And it is a music change that is to me surprisingly significant!

ATB from George

I'd still be flabbergasted if it made any difference .... just can't get my head around this digital stuff 'cos ALAC and AIFF contain the same music .... they must do or you couldn't convert from one to other and back again and again. Still what we hear is what we hear and so if you prefer it after the change then that is good news ....  as you imply, George, it is all about whether the music sounds more enjoyable. 

Now you've done all the converting and deleting then you may want to de-fragment your disk (I think it is an option on your PC) as this will help iTunes find and recover the music with a bit less effort (disk crunching) - after lots of file conversions and deletions your disk is likely to fragment (or at least it always used to be like that on Windows) and you can quickly fix this for optimum results. 

During my holiday there was lots of music - mostly pretty drab stuff with a band attempting to emulate the sound of Frank Sinatra - you can imagine how much I enjoy that (not) - there were two exceptions: one night a concert band played orchestral music transcribed for a concert band and another night a string quartet from Kiev played popular classics including Bach and I have to say both were like a breath of fresh air. 

In spite of what Guy and others say, I do find that there is an incremental quality increase using AIFF instead of ALAC with iTunes. As this is on my sole music system, then I decided that for labour alone and no actual cost this increment of improvement made the effort of conversion worth the effort. 

In practice the conversion to AIFF produced sonically identical results to a direct new AIFF rip from the original CD in question, so I was released from the wish to re-rip all the discs.

While the sonic effect is slightly less bright in most cases -though not all, and I'll go over that later - the effect of the change does nothing to reduce the detail - rather the reverse, details are occasionally revealed that were suspected as being there! But the sense of the space within which the performance becomes more apparent, and the actual performers - in most cases, which I'll deal with later - are more naturally distanced.

This sounds like a sort of Hifi artifact  but it is not, because the acoustic of the hall always affects the way musicians play. In a dry acoustic [with short and quiet resonance], then the players will make up for the lack of resonance by playing their articulation longer with less of a gap between the ends of staccato notes and the start of the next next ...

It helps clarify why a performance is being played this way or that, and aids the replay in getting out of the way and allowing for easier concentration of the music.

Some examples of the musical effects of this change.

On close recordings and small scale chamber music recordings, there is no sense of the performers moving further from the microphone [and back from the front of the speakers, as it were], but rather a sense of there being as immediate as one could wish. A real beneficiary of this is Ella Fitzgerald in those estimable Verve recordings from the 1950s, which sounds as fresh as a daisy, but now without a hint of false sibilance. With chamber music such as Reginald Kell playing Brahms Clarinet Quintet [with either the Fine Arts Quartet in the late 1950s, or the Busch Quartet in the late 1930s] the result is that his most beautiful and expressively varied Clarinet playing is even more clearly nuanced, and involving. It has an even more visceral quality.

The Solo and Concertant Piano benefits from this more natural quality just as much as the human voice or the solo clarinet! Annie Fischer's fiery Beethoven Sonata Cycle is revealed as being both fiery and subtle! Multi-faceted ...

On larger productions, the results are seemingly slightly different, but no less pleasing. The great period of stereo recording at EMI from 1955 till the mid 1960s produced replay that has always been teeming with detail,, and naturally balanced at a distance. This change brings two improvements. Firstly the actual perspective of distance is clarified, and yet the details emerge even more naturally proportioned so that with respect to these recordings, there is the sense of more expressive nuance than one ever suspected got past the microphone before! As this covers most of Otto Klemperer's recordings, naturally I am very pleased with this development!The musical balances are even more acutely lucid, for example when a flute doubles the violin line, and then a second joins in ... and so on!

Where a large Choir is present, the rich weight of several hundred voices in concert is quite remarkably pleasing. Often this causes replay allsorts of difficulties, but it is something that my replay is now excelling at. This is quite a surprise! But along with the piano and solo female voice it is among the most difficult of styles to replay without strain. Very pleasing.

On 78 transfers, though the tonal effect is slightly less fiery, there is the sense of an even more visceral connection to the ancient performances relayed! I was listening to Vaclac Talich's estimable late 1930s Abbey Road recording done with the Czech Philharmonic Orchestra of Dvorak's Eighth Symphony, and though it clearly in the "wall of sound" style of recording with no sense of depth within the recording itself, it is revealed as being better balanced between the sections of the orchestra than most modern recordings of the same music, and this performance breathes passion like no other in my experience. What must London audiences have made of this performance when given in the old Queens Hall in Langham Place [next to BBC Broadcasting House before it was bombed away in 1942] the same week as the recording was made at HMV whilst the orchestra toured Great Britain? Yet the recording is shown as a technical triumph of its time, and the music making becomes that much more involving with the change. 

On modern digital recordings, one realizes that the criticism of digital methods compared to analogue vinyl discs, is actually a question of the way the recordings themselves are presented by the replay. In the old days I bought a CDS 2 player [ex-dem] and said at the time that I thought is was the first CD player that no longer sounded digital. I though it was as different as chalk and cheese from the CDX [first generation], and was more than content that the additional cost was worth it. I would say that whilst the difference is less stark  the AIFF presentation tends toward the natural and visceral presentation of the CDS 2, whilst the ALAC sits closer to the CDX, which was more forward and more tiring over time. Less natural and more intrusive.

As Simon notes on page one of this thread, once one has noted the improvement of AIFF over ALAC, it is hard to ignore. For my own perspective on this, I was firmly with Guido that it was a nonsense  but as my earlier posts show, I accidentally discovered it! There was no wishful thinking. Rather the reverse, I was thinking what a pain it was to discover it! But once the itch has "itched" then one must scratch it!

I have found not one single case where the musical aspects contained in any recording have not benefitted from this change, sometimes subtle, and sometimes profoundly significant!

AIFF has it over ALAC at the same kind of level as a CDS series player has it over the CDX series. Now if you prefer the CDX series, perhaps you should stick with ALAC!

ATB from George

PS: The most incredible improvement occurred on a needle drop transfer of a late 1950s LP of the Milt Jackson Quartet that a friend brought me! There is a Xylophone that has the potential to rip your ears off! Now what sounded before like very marginal tracking of the LP is revealed as being a wonderful presentation of the large amount of overtone upper partials of the Xylophone  If goes from being marginal [or on another day unlistenable] to now being a lovely primitive recording of some highly enjoyable music making!

George, very insightful write up. If you can find the time, I would be interested to hear your views of the differences between AIFF and FLAC on some of your favourite recordings.

I am not as eloquent as your good self, but I simply find FLAC tends to sound sterile and mechanical to WAV which tends to be more visceral and organic, as well strangely more distinctive in the timing and beat of the music.

Simon

Hello Simon

Please could you google "fordgtlover Here is the difference between WAV and FLAC" ... you may find it fascinating ... it is test that the fordgtlover conducted and got some interesting results. Not saying it proves anything - it is only an experiment using one track, I just thought you may find it interesting. 

All the best, Guy 

Dear Simon,

I have not had a chance to try out FLAC versus WAV. When I first did experiments with computer audio, I tried EAC for ripping WAV, and Media Monkey for replay. It was too awkward for me, and I almost gave up.

I then had the idea that I might experiment with iTunes in-spite of it having a bad reputation here on the forum. I found my results did not correspond with the received wisdom of the day here, and subsequently ripped tracks from about 600 CDs. All in ALAC.

But I am now very pleased that I have converted to AIFF.

Julian H may say something about this, as I discussed this with him, and he said what I had found paralleled his experience of FLAC being inferior to WAV.

I hope he posts his experience of it, as it is important to share these things. 

I would have ripped to AIFF in the first place if I'd known what I do now!

ATB from George