XXHighEnd

Hi,

I’m a little bit late for the following question.
Did anyone compare the (with dBpoweramp) regenerated WAV file from a m4a with the ORIGINAL WAV from a CD.
The sizes of both WAV files have nearly identical values – but the SQ?

Thanx
Joachim

I made a comparison and the result can save us some money, at least with the newer releases/records on iTunes.
I bought the Fourplay album "Esprit De Four" at Itunes and as a CD. The Itunes m4a copy was converted with the internal Itunes converter from dBpoweramp, and the CD was ripped via dBpoweramp, latest version.

You see both tracks loaded in iZotope RX 3, and the graphic shows no difference.
The SQ was nearly identical except in the end of the track the cymbals on the Itunes copy were lacking in resolution.

Joachim

Hi there Joachim,

I don't know what makes you think that you can compare the quality of tracks (rips etc.) by such super rough means of pictures, but ... you can't.

Still a somewhat more trained eye can see that these two versions clearly are not the same and they should not sound the same; whether really perceivable is something else.

I don't know how that conversion can fail and I also don't know how your rip can have failed - similar as that I don't know whether both versions were not the same to begin with. So :

Your rip is more compressed or the conversion couldn't deal with the soft limiting applied to the track in order (and it made less limiting of it);
The easiest for me is to think that both tracks were not the same to begin with.

I am sure that if we look inside of the tracks (drill down to the real data) both are all over different.

Regards,
Peter

Hi Peter,
The CD was ripped with a Plextor PX-L890UE as I'm used to.

When we are looking to the peaks of the upper channel, it looks nearly identical. Both copies shows a frequency cut a 20KHz, and that should not be at the Itunes copy, it should be at lower frequencies, or not?

I made cuts with two different resolutions at the marked peak of the upper channel. See CD picture above.

Joachim

A question first Joachim :


Why ?
(I just don't understand)

I thought that AAC-LC limits higher frequencies, but that is not the case.

I checked the rip again it is identical with AccurateRip.

Hi Joachim,

I thought that AAC was lossy compression ? I am surprised...

Alain

http://www.applelossless.com/

Hello,

the tracks are bought from iTunes and they claim to use AAC-LC.
I was very astonished that the newer recordings at iTunes are nearly identical in their graphs (WAV) with the CD rippings  since the beginning of 2013.

There should be a difference, but it seems .......... .

http://www.iis.fraunhofer.de/en/bf/amm/produkte/audiocodec/audiocodecs/aaclc.html

best

Joachim

from Peter's link:


Maybe they do lossless in the Apple Store after all...
Otherwise checksums cannot be the same (accurate rip)

regards, Coen

But that was about Joachim's own rip of the CD version. Not sure how to check against AccurateRip with a download.
And then *still* it doesn't tell much because 10 versions of an album can exist and they will (normally) all be in the AccurateRip database ...

Peter

check!

That would be the most surprising: itunes store and CD from the same source!

regards, Coen

I never looked into iTunes and I guess I never will (it's like a virus on your PC). But now I don't know ... Is it maybe commonly known that the M4A from iTunes should be normal CD quality but it is not ?

At least I can think that it's a dirty trick to present M4A as CD quality while it just as well can be lossy (like MP3 (AAC)) and that maybe nothing is going to detect that really; I mean I too can easily save an MP3 (for example) after it has been expanded to WAV and save it as WAV. Nothing would detect the difference except possibly your ears and spectral analysis.

Peter

And the checksum...

Alain

Now that checksum will tell you if they are completely identical. If you're only one sample off (even one byte) than the checksum will not be the same. This is the difficulty, you cannot know how your rip is comparable to the presumably unripped ' cloud'  version. They have to be from the same source AND have the same start and end samples. So no padding or truncing allowed...otherwise you'll end up with an unequal checksum even if the music is 100% the same.

I'd say only the CD ripping process gives reproducible results unless there is some smart checksum that is able to match the actual music data with a reference regardless of padding/truncation.

On a more DIY track you could align the two datasets on peak values in a (hex) file or music editor (pick a channel and search for nice extremes) than you can observe if the dataseries before and after the peak are exactly the same. Of course with the absolute sample magnitude. If not: they will not be equal enough to sound the same!

regards, Coen

Another graph from a CD and an ITunes track (256kbps). The WAV of the iTunes record was created with the internal device.
It is here very well to see, that the frequency was cut reduced to well below 20KHz.

Peters remark about the "Virus program" is ............ .

Joachim

Joachim,

I am glad you can see/confirm that "well below 20KHz". But I don't see it or don't know where to look. :sorry:

Regards,
Peter

In the middle of the .............. .

I must be thick ...
So you like riddles eh ? Too difficult for me.