itunes vs foobar convert

[hanaluluelle]

itunes seems to release songs at cvbr 256

how does this compare to the quality of an aac rip of an original cd assuming my settings are the same?

[hello_hello]

Depending on who you ask, AAC becomes transparent at much lower bitrates than 256kbps (for stereo).

In theory, for a given bitrate, the quality order for encoding methods would be: Constant Bitrate, Average Bitrate, Constrained VBR and True VBR.

True VBR is much like x264's CRF encoding, in that you pick a quality and the file size/bitrate will be whatever it needs to be. For TVBR, foobar2000 defaults to Q64 (the quality settings change in fairly large increments). I generally use Q91 myself. According to fb2k, that'll result in (very) roughly, 192kbps for a stereo track. Q109 would be somewhere in the vicinity of 256kbps.

Another advantage of TVBR is you don't need to adjust the bitrate according to the number of channels (assuming you also encode multichannel audio). You'd simply continue to use your preferred quality setting.

[stonesfan99]

If you are encoding the rip in foobar2000 to AAC (Apple) and using VBR Q 109: 256kbps (*) for the bitrate, this should produce files that are the exact same quality as if you bought it on iTunes. My question is if you have the CD, why not use Apple Lossless?

[KarMa]

It takes a special individual with quality audio equipment and exceptional ears to be able to reliably discern between 192kbit lameMP3 and a FLAC, in an ABX test. I am not one of those special individuals, if they even exist. So considering that AAC Nero and AAC Apple codecs are better than lameMP3 at low bitrates, I think it's fair to say that either one of these AAC codecs should be far and away transparent to you @256kbit. foobar2000 supports encoding with both Nero and Apple codecs with Apple considered to be slightly better in low bitrate tests.

You can do ABX tests with a ABX plugin for foobar2000, to do your own easy comparisons between different audio encodings.

[stonesfan99]

It takes a special individual with quality audio equipment and exceptional ears to be able to reliably discern between 192kbit lameMP3 and a FLAC, in an ABX test. I am not one of those special individuals, if they even exist. So considering that AAC Nero and AAC Apple codecs are better than lameMP3 at low bitrates, I think it's fair to say that either one of these AAC codecs should be far and away transparent to you @256kbit. foobar2000 supports encoding with both Nero and Apple codecs with Apple considered to be slightly better in low bitrate tests.

You can do ABX tests with a ABX plugin for foobar2000, to do your own easy comparisons between different audio encodings.

I am one of those people who does not hear the difference between 192kbps LAME MP3 and lossless. But if he has the CD, why not buy a large hard drive and make the initial rip into Apple Lossless for archival? (I only say Apple Lossless because it sounds like he is working with iTunes). And I wouldn't say the difference between lossless and even 320kbps CBR LAME MP3 is that difficult to hear - Head-fi Blind Test - FLAC vs 320 MP3.

[KarMa]

I am one of those people who does not hear the difference between 192kbps LAME MP3 and lossless. But if he has the CD, why not buy a large hard drive and make the initial rip into Apple Lossless for archival? (I only say Apple Lossless because it sounds like he is working with iTunes).

I'm not against going lossless if OP wants to go that route. Just have to also consider compatibility with whatever device OP has. Playing audio on the computer (and any cellphone with VLC installed) is easy and can play just about anything. But things like car stereos and most portable "MP3 players" might not play lossless (besides simple .WAV). For archival purposes lossless is certainly a great option.

And I wouldn't say the difference between lossless and even 320kbps CBR LAME MP3 is that difficult to hear - Head-fi Blind Test - FLAC vs 320 MP3.

But you just said you can't hear the difference between 192kbit lameMP3 and lossless.

Anyway with the forum thread you posted, I'm not entirely sure what you are pointing out but the OP in that thread posted an errored MP3 originally (lame decoder issue apparently) making it easy to spot the difference between the two files. But they seemed to have fixed it half way down the first page. It's also too bad that this is an old thread dating back before the ABX plugin was updated with encrypted log hashes and hashes for the files being tested included in the log. Making it super simply to just edit the text to fit your liking, and no one would know. There is just no way of knowing if they were being honest.

But lets say the same tests happened today with the newer version of ABX testing in foobar2000, with encrypted hashes. While I'm going to guess it's pretty difficult to tamper with the ABX logs and have it pass the signaturecheck website. It's still easy to use an outside spectrum analyzer like Audacity to look at the spectrum of the audio playing during your test. Which is what I did when I took this test using the audio samples provided in that thread (head-fi.org).

foo_abx 2.0.1 report
foobar2000 v1.3.10
2018-03-30 13:18:48

File A: A.wav
SHA1: 4dbea6f3394e72c06814b0e5c674d6e927fcb373
Gain adjustment: -2.43 dB
File B: B.wav
SHA1: f56de0da32d5f8a28ab02aba3a5c5bc220923c0e
Gain adjustment: -2.43 dB

Output:
DS : Speakers (Realtek High Definition Audio)
Crossfading: NO

13:18:48 : Test started.
13:19:36 : 01/01
13:20:08 : 02/02
13:20:33 : 03/03
13:21:16 : 04/04
13:22:08 : 05/05
13:22:47 : 06/06
13:23:12 : 07/07
13:24:05 : 08/08
13:24:29 : 09/09
13:24:47 : 10/10
13:25:12 : 11/11
13:25:31 : 12/12
13:26:04 : 13/13
13:26:34 : 14/14
13:27:10 : 15/15
13:27:55 : 16/16
13:28:51 : 17/17
13:29:14 : 18/18
13:29:49 : 19/19
13:30:31 : 20/20
13:30:31 : Test finished.

 ---------- 
Total: 20/20
Probability that you were guessing: 0.0%

 -- signature -- 
caacb9f06baac2c206f04e11eae4fa010d85e426

Behold, my fool's gold ears. The MP3 version had a low pass filter applied to it, cutting off frequencies above ~20khz while the lossless source went all the way up to ~21.5khz. Making it easy to pick out which was which during testing with Audacity. Sadly this makes me suspicious of any claims of being able to tell a difference between lossless and high bitrate MP3. I'd probably only believe someone if they actually had to go to a location and have a 3rd party administer a 320kbit MP3 vs lossless test.

[hello_hello]

And I wouldn't say the difference between lossless and even 320kbps CBR LAME MP3 is that difficult to hear - Head-fi Blind Test - FLAC vs 320 MP3.

The golden-eared types at Hydrogenaudio don't seem to agree.
Recommended LAME encoder settings.

[Marsia Mariner]

It takes a special individual with quality audio equipment and exceptional ears to be able to reliably discern between 192kbit lameMP3 and a FLAC, in an ABX test. I am not one of those special individuals, if they even exist.

Yes, they exist.

Exclusively CD. I can't listen to mp3 even at the higher bit rates. I can hear artifacts that many cannot. I can't watch any HD TVs. That's just the way my senses are. I can tell if a car has LED tail lights that are not d.c. powered. I can even tell you the duty cycle. Very distracting while driving.

[KarMa]

.... if they even exist.

Yes, they exist.

Exclusively CD. I can't listen to mp3 even at the higher bit rates. I can hear artifacts that many cannot. I can't watch any HD TVs. That's just the way my senses are. I can tell if a car has LED tail lights that are not d.c. powered. I can even tell you the duty cycle. Very distracting while driving.

I'm not sure who this user is or why you are using this decade old comment as evidence, considering they don't state the codec used (lameMP3?) and nothing more than vague MP3 bitrates. They don't even say if these were blind tests. Finally, you are quoting someone who said they could not tell the difference between PCM (CD Original) and ADPCM.

If you encode mp3 at the maximum bitrate they, IMHO, you might as well use ADPCM.

No, ADPCM is demonstrably dramatically worse than highbitrate mp3.

The problem with such a hard claim is that mp3 relies on psychoacoustics and no one model of the human auditory perceptual system can possibly represent every listener of mp3 encoded material.

Personally (and, hence, subjectively), I could not distinguish ADPCM vs PCM for the same material whereas I can here the difference between mp3 and PCM. My PCM sources are CDs and older/non-mainstream at that - i.e., not mastered so hot as to be worthless IMHO. Of course, I no longer have a need for ADPCM.

I would have to agree with 2Bdecided on this. With ADPCM only starting to do better against 128kbps or lower MP3s, in my tests today using the IMA ADPCM encoder with FFMPEG. Though the artifacts between lameMP3 and ADPCM are like apples to oranges, with ADPCM giving me a minor headache. Considering ADPCM has a constant bitrate of around 354kbps for 44.1khz content, 192-320kbps lameMP3 blows it out of the water.

[Marsia Mariner]

I'm not sure who this user is or why you are using this decade old comment as evidence,

JohnnyMalaria was a name known to all and respected by many here on Videohelp.
He was from the U.K. (if I remember correctly) and he was (¿still is?) a software developer as well.
I quoted him because what he said did impress me and therefore remained in the conscious level of my memory banks.
I started visiting Videohelp in 2006 and became a registered user in 2007.

considering they don't state the codec used (lameMP3?) and nothing more than vague MP3 bitrates. They don't even say if these were blind tests. Finally, you are quoting someone who said they could not tell the difference between PCM (CD Original) and ADPCM.

*yawns*

I would have to agree with 2Bdecided on this. With ADPCM only starting to do better against 128kbps or lower MP3s, in my tests today using the IMA ADPCM encoder with FFMPEG. Though the artifacts between lameMP3 and ADPCM are like apples to oranges, with ADPCM giving me a minor headache. Considering ADPCM has a constant bitrate of around 354kbps for 44.1khz content, 192-320kbps lameMP3 blows it out of the water.

You meant 352.8 kbps at 44.1 kHz at 16-bits, I presume.
Also: it seems you don't know very-much of what you're talking about. It's not only a matter of required bitrates.
Clearly you cannot understand and therefore cannot accept, for example, that MP2 IS better than MP3.

*plonk*

[hanaluluelle]

space is an issue for me unfortunately so no can do with your lossless encodes and hard drive suggestion.

so most of you guys are basically saying that cvbr 256 is already transparent to the original so i wouldn't tell the difference between what itunes is selling and foobar aac encodes anyway? are you implying that itunes's source is cd?

[Cornucopia]

iTunes' sources should be from the original studios' masters. Probably equivalent to the CDs, but there are exceptions.

AAC @ 256kbps, whether cbr or vbr, is virtually transparent to the great majority of people. At that rate, whether encoded by iTunes or encoded by foobar, they should both be transparent and equivalent, even if not identical (this assumes 1 Generation each, and both using proper settings, and no pre-filtering).
The (efficiency/quality) differences between the 2 codec implementations should not be apparent at the highest bitrates, only at middle and/or lower bitrates.

Scott

[KarMa]

are you implying that itunes's source is cd?

Most likely or maybe even a higher quality source. But I don't work for Apple and am just guessing.