Mix sounds drastically different in iTunes?

Hello all,

After bouncing a mix (to 16/44.1), and playing the file within Quick Time and OSX's Finder (by hitting the space bar after selecting the file), the song's mix sounds the same as it should be when compared to how I had it within Logic.

However, when I play the exact same file in iTunes, the balance of the mix is totally different and sounds way off. I have sound enhancer off, and I have the EQ off, so I'm not sure what is causing this peculiar anomaly.

I even tried uploading the same MP3 file to my gmail account and then streamed the file directly off gmail's server via my web browser, and it sounds fine! But something strange is happening in iTunes.

Appreciate any advice!

Hello all,

 

After bouncing a mix (to 16/44.1), and playing the file within Quick Time and OSX's Finder (by hitting the space bar after selecting the file), the song's mix sounds the same as it should be when compared to how I had it within Logic.

 

However, when I play the exact same file in iTunes, the balance of the mix is totally different and sounds way off. I have sound enhancer off, and I have the EQ off, so I'm not sure what is causing this peculiar anomaly.

 

I even tried uploading the same MP3 file to my gmail account and then streamed the file directly off gmail's server via my web browser, and it sounds fine! But something strange is happening in iTunes.

 

Appreciate any advice!

I think you may want to check your Pan Law setting in Logic.

File > Project Settings > Audio

http://documentation.apple.com/en/logicpro/usermanual/index.html#chapter=43%26section=7%26tasks=true

#Edit: Forget that. Not a Logic issue. Didn't read carefully. Don't sue!

Another test I tried was to bypass my ULN-2 interface and just select my Mac Pro's headphone out as the main output. I then just used regular ear phones plugged directly into my computer to listen to the mix and it still has the same problem.

I then just tried to listen to other music of my favorite bands. And I noticed that there were drastic differences in the quality and mix of the songs when compared to QuickTime even with these commercial releases. QT sounded way more dynamic and clear. In iTunes, the mixes don't sound as wide and almost sound a tad grainy and distorted in comparison.

This is quite a revelation because I use iTunes to do a lot of my song referencing in terms of getting ideas for song writing or mixing. If this is an iTunes issue, then I will have to find a better music player to use from now on.

When I play the same file through iTunes and QuickTime they sound very very close but QT is maybe a bit louder, which also makes it seem to have a tad more bottom end.

I'd like to do a test and record their outputs one day and then do a more detailed comparison.

But I wouldn't say that the difference between them is "drastic" by any stretch. I'm still wondering if something else is happening with your system?

Itunes sounds like trash.. I don't think there's anything you can do about it.

There's nothing wrong with iTunes. Just disable all the Sound Enhancer, Sound check and eq settings. Sometimes the iTunes volume level might also be lower than the system one.

I would just use QuickTime player or play from Finder to check out the audio, easier.

I've done all those things.. quicktime always sounds better. I don't even open up itunes to reference sounds.. there may be nothing wrong with it to your ears, but it's always sounded bad to me.

iTunes, Finder and QuickTime player all use the same QuickTime codec and audio playback. Maybe this is something psycho-acoustic?

I've done all those things.. quicktime always sounds better. I don't even open up itunes to reference sounds.. there may be nothing wrong with it to your ears, but it's always sounded bad to me.

 

iTunes, Finder and QuickTime player all use the same QuickTime codec and audio playback. Maybe this is something psycho-acoustic?

Yup, exactly, the reason they all sound the same is simply because they are the same. Psycho-acoustics or pilot error. It's very easy to get iTunes to "do something" to your sound (like auto-normalize, EQ, enhance...) whereas QuickTime and QuickLook are processing-free by default.

Yeah, I forgot.. any sort of criticism of any apple program in these forums is never justified and we are always wrong.

I'll trust my ears.. been waiting for things to change with an update, but it never has, and I've seen plenty of people talking about the same kind of issues with itunes... We'll just say it's the 'pilot error' and leave it at that. What do I know.

I'm glad I found a way to block images.

It's the same codec. Run a loopback phase inversion test on all three "formats" , you'll get silence.

Yeah, I forgot.. any sort of criticism of any apple program in these forums is never justified and we are always wrong.

 

I'll trust my ears.. been waiting for things to change with an update, but it never has, and I've seen plenty of people talking about the same kind of issues with itunes... We'll just say it's the 'pilot error' and leave it at that. What do I know.

We're not discussing opinions here, but facts. iTunes = QuickTime = QuickLook. Same audio engine.

http://en.wikipedia.org/wiki/Itunes

That is a fact that they use the same engine.. fine. But they still sound different even when all sound enhancements and eqs are turned off. There are plenty of other people who have the same sort of issues with itunes. Why would I make this up? heh heh .. I don't really care about this really, I probably shouldn't have interjected because I know how people get all defensive about their software. I don't really care to argue about it, because you're right.. it's opinion and everything is subjective. Still, itunes sounds bad to me.. my solution.. I don't use it.

Not just me folks

http://discussions.apple.com/message.jspa?messageID=10359609

http://forums.macnn.com/82/applications/212286/sound-quality-quicktime-vs-itunes/

That is a fact that they use the same engine.. fine. But they still sound different even when all sound enhancements and eqs are turned off. There are plenty of other people who have the same sort of issues with itunes. Why would I make this up? heh heh .. I don't really care about this really, I probably shouldn't have interjected because I know how people get all defensive about their software. I don't really care to argue about it, because you're right.. it's opinion and everything is subjective. Still, itunes sounds bad to me.. my solution.. I don't use it.

 

Not just me folks

 

http://discussions.apple.com/message.jspa?messageID=10359609

Patient: It hurts when I bend my leg this way.

Doctor: Don't bend it that way.

Phase Cancellation test.

Opinions aren't going to end this polemic. On the other hand, even if this test would inambiguously prove the outputs are the same, you will always hear iTunes sounding like crap. That is what convictions can do to observations.

Moderator's Note:

OK Guys... in case you missed it, I've removed a bunch of the "polemic" from this thread. No more ad hominem posts, please, and stay on topic.

I'm also starting to think there's something off with iTunes.

On the whole, iTunes sounds a little quieter and there is maybe something subtly different about the imaging. I wouldn't go as far as to call iTunes output total crap but it ain't the same as QuickTime. And of course iToo have all my iTunes settings set for non-"enhanced" etc output.

Soooo, I went and did a little test. Recorded the output of both apps through soundflower in an attempt to do a null test. Each app is playing the same file. The null test didn't work - except in a couple of spots, one right at the top and then a few seconds later it breifly nulls.

So I wondered why would it null and then go out of null? What I found was that the iTunes output seems to drift with respect to the QT file. I could see that they weren't lined up when I looked at the middle of the files so I went back to the beginning and lined them up visually at the start and then measured the differences as the song went along.

22 seconds into the song, the two waveforms are out 4 ticks. 90 seconds, out by 12 ticks. By 1 minute 13 seconds it's now out 15 ticks.

I'm beginning to wonder if iTunes is running on the same clock. Can anyone account for what I'm seeing here? Or does my skewed test results have something to do with Soundflower?

Another test..... It looks like they both drift a bit when I'm recording the outputs of iTunes and Quicktime. iTunes just drifts a bit more.

I recorded another pass out of each app. They're all lined up at the top of the files but after a minute, you can see that both blue iTunes tracks have lagged behind the brown QT ones on top.

But none of these line up with each other. Second pass on the QT output put it slightly ahead of it's sibling. Second pass from iTunes runs slightly behind the first one.

This was all running through Soundflower into Logic. Top picture is tune start, bottom about 1 minute and 10 seconds in.

I'm also starting to think there's something off with iTunes.

 

On the whole, iTunes sounds a little quieter and there is maybe something subtly different about the imaging. I wouldn't go as far as to call iTunes output total crap but it ain't the same as QuickTime. And of course iToo have all my iTunes settings set for non-"enhanced" etc output.

 

Soooo, I went and did a little test. Recorded the output of both apps through soundflower in an attempt to do a null test. Each app is playing the same file. The null test didn't work - except in a couple of spots, one right at the top and then a few seconds later it breifly nulls.

 

So I wondered why would it null and then go out of null? What I found was that the iTunes output seems to drift with respect to the QT file. I could see that they weren't lined up when I looked at the middle of the files so I went back to the beginning and lined them up visually at the start and then measured the differences as the song went along.

 

22 seconds into the song, the two waveforms are out 4 ticks. 90 seconds, out by 12 ticks. By 1 minute 13 seconds it's now out 15 ticks.

 

I'm beginning to wonder if iTunes is running on the same clock. Can anyone account for what I'm seeing here? Or does my skewed test results have something to do with Soundflower?

Nice work Camillo.

Well, my first thought is that this drift is pretty big, 15 ticks at 120 bpm with a 48 Khz file is ± 30 ms - it is quite shocking, actually. A few samples, OK. But 1,500...?

I now wonder: is it possible at all to capture two identical files from the same playback method, or is the codec or the clock so "unstable" that that would prodiuce two non-nulling files too? I still think it just doesn't makes sense that iTunes would use it's "own" codec... it would be nice to have an audio-SW developer explain this one...

I said only a test could prove it, and it seems it did. Unless it turns out that this method cannot produce identical files even when capturing the same file from the same player again. So can it?

I'm hoping someone else can run a similar test. Could be my clock source is screwed but it's a pretty decent IO unit - an RME multiface.

Or there might be something else about how I did the test that would cause these kinds of drifts. Any ideas?

Ideally I'd like to be able to start the song simultaneously from both apps but I'm not sure how to do that.

None of this BTW proves that iTunes is doing something seriously bad to the sound, any more than QT is, since they both drift. But it does make it impossible to do a proper null test. That being said, iTunes output, as I said before, does sound subtly different than QT's.

Ok, here goes.

The setup: To rule out any influence of external participants, iffy software and dodgy mini-phone-plugs, I routed all three contenders (Logic, iTunes and QT Player to the same hardware outputs (RME Fireface800). Those hardware outputs are then software patched using Totalmix's foldback feature back into Logic's inputs. All iTunes gimmicks are off, of course.

The test: Record the playback of a WAV file that was initially bounced in Logic through iTunes and through QT Player back into Logic. Align the three tracks (Original, playback through iTunes and playback through QT Player), flip phase on both playbacked tracks and play each against the original track.

The result: Full cancellation from the first sample to the last, for 4.:30 minutes for both iTunes and QT Player.

The next test: Playing a 192k mp3 that was bounced from Logic as Original, through iTunes and through QT Player.

The result: Full cancellation from the first sample to the last, for 4.:30 minutes for both iTunes and QT Player.

And just for a laugh: checking the mp3 through both iTunes and QT Player against the original, uncompressed WAV.

The result: As expected, quiet squishy squashy but otherwise absolutely consistent sounds, from top to bottom, no drift whatsoever.

So, the myth that iTunes does something behind the scenes (besides the obvious, stupid Sound Enhancer stuff) is no more.

The result: iTunes is a completely transparent player, as is QT Player.

Christian

I confirm what Christian says. Did a test with Audio Hijack Pro (iTunes vs Quicktime). They null.

J.

And what have I learned from all this? That my Total Mix can do a foldback and that Jack won't mess with your audio like Soundflower does. Or something like that.

That was Audio Hijack Pro...maybe I should do one with Jack and see how it goes.

J.

Sometimes math cannot determine everything.

You are aware that it's a computer that you're looking at ?

The players sound different.

So you did the test as described ? And it nulled out ? And you still claim they sound different ?

If so, I will never talk to you again. Whether that's a threat or a promise - your choice.

Christian