Why do apple lossless file sizes vary ??

[Ralf11]

I discovered that some rips from CD into Apple Lossless format vary in size (by 3x sometimes).  In some cases at least, they duplicate files were ripped on the same day, so I don't think I changed any parameters.

 

Any ideas?

[Jud]

8 minutes ago, Ralf11 said:

I discovered that some rips from CD into Apple Lossless format vary in size (by 3x sometimes).  In some cases at least, they duplicate files were ripped on the same day, so I don't think I changed any parameters.

 

Any ideas?

 

Identical sample rates?

 

What software?

 

Edit: I don't think different compression settings would make that large a difference.

[mansr]

Are you saying you ripped the same CD several times and resultant file sizes differ wildly? That would be weird indeed.

[Ralf11]

I don't think they were ripped several times.  The same date stamp makes me think not, and that sample rates and compression settings would be the same.  I didn't check the time stamps tho or other info, so cannot be absolutely certain.

 

I saw this when I was cleaning up iTunes to gain some free space on my long suffering Mac Mini (and stop IClod from harassing me with offers to upload all my secrets onto their cloud).

 

Some files are the same size, other vary a bit, and some sizes vary by 3x.  All were ripped from CDs using iTunes into Apple Lossless.

 

Any way I gained about 16 Gigabooties of space so may be able ot hold out until Apple releases new iMacs.

 

It seems very odd however.

[mansr]

It's perfectly normal for compression ratios to vary between files. Pure tones (and silence) compress extremely well whereas noise is incompressible. Most music (other than Cage's 4:33) falls somewhere in between.

 

If the differently sized files you have are of the same music, something is odd. Otherwise some variation is to be expected.

[Ralf11]

Yes, the exact same track.  Sorry if that wasn't clear.

[rando]

I know where this is going, hitting random with multiples of the same track selected to do a, probably inconclusive non-scientific, blind test.

[mansr]

35 minutes ago, Ralf11 said:

Yes, the exact same track.  Sorry if that wasn't clear.

So you compressed the same track several times with the same settings and got compressed sizes varying by a factor 3? That shouldn't happen. Either you're not describing what you actually did, or your software is horribly broken. I suspect the former.

[Ralf11]

read the OP - I don't know exactly what happened over the course of time I've been using iTunes (2012 IIRC) but the dates were the same

[rando]

What you are likely referring to is the "last modified" date.  Best guess without being there is over the course of who knows how many thousands of iTunes updates it did some processing to their lossless algorithms on your files and things got weird.  

 

I find it odd you haven't investigated any difference in bit rate or run the files through something like spek as a preliminary step.

[kumakuma]

The MediaInfo app ($0.99) from the App Store might be a useful tool for solving this mystery.

 

The Text View gives a ton of detail about any media file:

 

Quote

 

General
Complete name                            : Nicky Parrott - Strictly Confidential/Nicki Parrott - 05. Hallelujah, I Love Him So.m4a
Format                                   : MPEG-4
Format profile                           : Apple audio with iTunes info
Codec ID                                 : M4A  (M4A /mp42/isom)
File size                                : 23.6 MiB
Duration                                 : 4mn 14s
Overall bit rate mode                    : Variable
Overall bit rate                         : 777 Kbps
Album                                    : Strictly Confidential
Album/Performer                          : Rossano Sportiello, Nicki Parrott & Eddie Metz
Part/Position                            : 1
Part/Total                               : 1
Track name                               : Hallelujah, I Love Him So
Track name/Position                      : 5
Track name/Total                         : 15
Performer                                : Nicki Parrott
Recorded date                            : 2016
Encoded date                             : UTC 2018-11-17 05:21:25
Tagged date                              : UTC 2018-11-17 05:21:25
Writing application                      : X Lossless Decoder 20141129, QuickTime 7.7.3
Cover                                    : Yes
Comment                                  : Purchased from 7digital.com

Audio
ID                                       : 1
Format                                   : ALAC
Codec ID                                 : alac
Codec ID/Info                            : Apple Lossless Audio Codec
Duration                                 : 4mn 14s
Duration_LastFrame                       : -88ms
Bit rate mode                            : Variable
Bit rate                                 : 774 Kbps
Channel(s)                               : 2 channels
Sampling rate                            : 44.1 KHz
Bit depth                                : 16 bits
Stream size                              : 23.5 MiB (100%)
Language                                 : English
Encoded date                             : UTC 2018-11-17 05:21:25
Tagged date                              : UTC 2018-11-17 05:21:25

 

 

 

[Ralf11]

Thx - that's a lot more than the Finder's Get Info facility shows

[kumakuma]

4 minutes ago, Ralf11 said:

Thx - that's a lot more than the Finder's Get Info facility shows

 

You're welcome. Be sure to let us know if you find out anything interesting.

[Ralf11]

I will if I find any other duplicate tracks - these were ash canned to gain disk space.

[wgscott]

2 hours ago, Ralf11 said:

Yes, the exact same track.  Sorry if that wasn't clear.

 

Use the afhash command to see if the checksum is identical in the case where a supposedly identical file differs dramatically in size.

 

eg:

% afhash 01\ Dire\ Wolf.m4a

01 Dire Wolf.m4a :
No hash in file.
SHA1 Hash of audio data : 2da2d240245b71592b47734d735e585251c38a42

If they differ, then there is a problem.  If they are the same, then presumably the metadata (probably the album art image(s)) account for the difference.

[tmtomh]

Something doesn't make sense here.

 

I don't wish to insult the OP, but one thing that comes to mind RE the 3x file size is that both Apple Lossless and lossy AAC files have ".m4a" file extensions at the end of their names, and both report as "Apple MPEG-4 audio" in the Finder's "Get Info" window. 

 

So my first guess about the ones that are that radically different in size is that the smaller ones actually are AAC files, not Apple Lossless. If you assume a 256kbps AAC file, then a typical Apple Lossless file - which for most kinds of music is going to be in the 750-950kbps range - would be about 3x as large.

 

For ones that are pretty close in size, then I'd go with @wgscott's guess that it's metadata - album artwork in particular can add several MBs to the file size.

 

@Ralf11, there's really nothing more anyone can say unless or until you provide a good deal more detail.

[Ralf11]

I have been "fixing" the lo res album art recently so that could have been it.  I suppose the art is stored in the 'container' along with other metadata and the music file per se (?)

 

I still have some AAC files on the machine but I don't think any of these were them - I've worked to kill off any AAC files and they are pretty well segregated.

 

I may not have any more details on this as I deleted the files to get disc space.

 

I'll check the laptop to see if there is anything like this going on there.

[Speed Racer]

Something is wonky because the lossless compression algorithm is not random. There is no way it would compress the same track differently from one run to the next. It would be exactly the same every time.

[Ralf11]

Could the Lossless compression algorithm work differently on Windows than on a Mac?  IIRC, some files were transferred over from a Windoz machine when I switched to a mac.

 

Alternatively, different versions of iTunes have been used to rip CDs, so maybe the 2012 iTunes algorithm differs from the 2016,..17 one.

[Ralf11]

OK, on the laptop I found 2 versions/files with different sizes (not 3x, just a few percent).  They both have the same Date modified and Date added...

The times are the same (to the nearest minute).

 

 

[GregWormald]

1 hour ago, Ralf11 said:

I have been "fixing" the lo res album art recently so that could have been it.  I suppose the art is stored in the 'container' along with other metadata and the music file per se (?)

 

I still have some AAC files on the machine but I don't think any of these were them - I've worked to kill off any AAC files and they are pretty well segregated.

 

I may not have any more details on this as I deleted the files to get disc space.

 

I'll check the laptop to see if there is anything like this going on there.

Low and high resolution artwork can vary a lot. And some files may have the artwork embedded in the file and some may use a link to a separately filed album artwork.

 

Greg

[mansr]

5 hours ago, Ralf11 said:

Could the Lossless compression algorithm work differently on Windows than on a Mac?

Minor differences are possible, but nothing of the magnitude you're seeing.

[mansr]

5 hours ago, Ralf11 said:

OK, on the laptop I found 2 versions/files with different sizes (not 3x, just a few percent).  They both have the same Date modified and Date added...

The times are the same (to the nearest minute).

 

That's a 25% difference so it's still quite a lot. If you want to get to the bottom of this, your best option is probably to share the files with a few people and see what they find.

[Jud]

5 hours ago, Ralf11 said:

OK, on the laptop I found 2 versions/files with different sizes (not 3x, just a few percent).  They both have the same Date modified and Date added...

The times are the same (to the nearest minute).

 

 

 

Does this mean there were no files that varied in size by 3x, or that these are different files than you originally described?

[new_media]

First three things I would do:

 

1. Open both files in iTunes and see if the reported bit rate is the same or different.

 

2. Strip off all of the metadata with QuickTime Plus (by exporting to MP4 with audio pass through) to see if the file sizes are still different.

 

3. Export both files to WAV and run checksums to see if the audio data is identical.