AIFF to ALAC to AIFF again ?!

You will get the original back intact. It's akin to zipping a data file - when unzipped it's exactly as it was before.

Hope I don't get in trouble for saying this, BUT you probably won't get a definitive answer here either. If you do a search in the Community here, there have been many discussions in the past where tests were done before and after conversions and the files were exactly the same.

So the argument you'll get is:

...but the files before and after conversion are exactly the same

...but even though the file are exactly the same, I can still hear a difference.

But theoretically they will be the same.

Now that's as definitive as an answer you can get 😊

Joe Gramm wrote:

So the argument you'll get is:

...but the files before and after conversion are exactly the same

...but even though the file are exactly the same, I can still hear a difference.

You have to be very cautious about 'I can hear a difference'. When Dolby A noise reduction was first introduced, a leading record critic claimed he could hear degredation on the sound. Dolby invited him to a demo, and played him a tape with Dolby A processing, and another of the same recording without it. He was quite definite that the latter sounded better.

Then they told him that the unprocessed tape had been copied from the Dolby A tape after playback through the processor. He left in a huff, but it rather demonstrates the point about thinking you hear what you expect to hear.

Similar problems arise with expensive speaker cables and other tweaks particularly those that take some time to implement, causing a long delay between samples.

And there was once a recording device which was taken to live demos, where playback of a recording of a live person, and the live person himself, were done behind a curtain to an audience who all swore they couldn't tell the difference.

This was in the 1890s and was the Edison Phonograph.

See, that's why I think if the files can be compared side by side and they are exact, then I would depend on that being true. The brain can do tricky things. There are two many variables to consider with listening tests.

A conversion from a Lossless to say an AAC 320 Kbs would be a better discussion about loss of quality.

I only know back when Apple introduced Apple Lossless, there were many Discussions about quality, conversion, etc. Unfortunately I can't site anything specific. I"ll bet Roger Wilmut1 has some info on tests. He is by far the expert here.

Here's how to test it, if you have the patience.

Make a mono recording in AIFF. Let's call this A.

Make a ALAC copy of it. Then make an AIFF copy of that, and let's call it B.

Open a suitable sound editing program such as Amadeus Pro. Create a stereo track with A as the left track and B as the right track.

Phase-reverse track B. Mono-ize the result. The mono-ized version should be total silence.

You could try it again with AAC or MP3 and listen to the disturbing results.

I did something like this once with FLAC and came out with zero signal.

I just did this with Amadeus II trial version and iTunes 7.5:

Take an AIFF music file, copy one track to a new file (older Amadeus doesn't have monomize feature).

Save new mono file - mono1.

Invert signal on copy of mono file - mono2.

Take first mono file, import into iTunes with ALAC.

Open ALAC file in AmadeusII, copy and paste over (adding them) inverted mono2 file.

- Result: Flat line. The signals cancel.

The line does have slight thickness to it where there was originally some sort of signal (vs. blank audio at very start and end), but I don't know if this is something in Amadeus or what. I tried analyzing that signal and no data are reported. I tried magnigying the sonogram to maximum and nothing shows up. I also took the original mono AIFF and pasted over the inverted original mono AIFF and it produces the same slightly thick line so I'm not going to worry about it in terms of the ALAC being identical to the original. Just to check, I also used iTunes to re-convert the ALAC back to AIFF and pasted over that signal with the same flat-line result.

A useful experiment. This proves with mathematical certainty that the files are the same: and that if anyone claims to hear any differences it's down to external influences such as different jitter or other parameter performance in playback, indigestion, or a wish to hear a difference.

Murphy, If you ever find yourself dealing with better-than-CD-quality tracks, there will definitely be a difference.

Start from a WAV file with 24-bit sample size, convert it to ALAC, and then back to WAV, and all you will have is 16-bit samples.

I just have the old itunes. I looked at XLD which is my preferred ripper and under Apple Lossless conversion settings it includes "same as original" as well as manual samples with up to 92kHz. Presumably the ability is there to sample higher rate files, I just don't know if Apple has actually included this in iTunes.

Limnos, I was using iTunes itself to do the conversion. It has a single setting called "Automatic," which unlike the XLD "same as original," does not promise (and does not deliver) that you will get the same sample size as the original.

By the way, 92 kHz sounds like a sample rate, which is the samples per second. I was talking about sample size, which is the bits per sample. But the point is the same: the conversion to/from "Apple Lossless" was not lossless.

The following line from http://www.computeraudiophile.com/content/XLD-ripping-AIFF-files-Mac with reference to XLD might be of interest:

I just converted a slew of 24bit/88khz and 24bit/96khz albums I purchased from HDTracks to Apple Lossless and the bit rate, sample rate, proper tagging, and album art were all preserved and show up in iTunes perfectly.

I just converted a slew of 24bit/88khz and 24bit/96khz albums I purchased from HDTracks to Apple Lossless and the bit rate, sample rate, proper tagging, and album art were all preserved and show up in iTunes perfectly.

Thanks. That is consistent with my experience. I started with WAV 96/24 and was able to get ALAC 96/24, but the conversion back to WAV was not lossless.

I was using iTunes for the conversion. It is possible XLD would have done better.