Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Peter Weihe]

marcel wrote on Wed, 12 January 2011 03:35

Has anybody reliably eliminated the possibility that the interface (not physical, but software) between PTHD and CoreAudio is to blame for this stuff?  Will, say, GarageBand or Cubase produce a similar result?

A further test with Cubase and Digi Core would help to find out.
Has anybody access to that combination?

[Peter Weihe]

marcel wrote on Wed, 12 January 2011 03:38

Noah: Not to derail, but what's with the +/-22ms of latency in the Logic session?

arconaut wrote on Wed, 12 January 2011 05:00

I'm not sure, maybe it's a buffer size setting needing adjustment in Digi Core Audio?  Looks like it might be maxed out at 2048? Maybe the latency could be improved, but would it affect polarity?

With my HD/Logic setup latency is roughly about 2 msec higher with Logic than with Pro Tools with the lowest buffer size setting for Logic.

As the polarity switch was observed in three different systems now, your's and my HD
setup und the HD Native setup of my friend each with different buffer size settings
I believe that it has no influence on the problem.

[Peter Weihe]

KB_S1 wrote on Tue, 11 January 2011 16:49

I am working on an album just now that has been mostly live takes of drums, bass and guitar. Overdubs of more guitar, lead vocals and other things. It is now going to the producers home studio for the twinkly details but he is running logic. It may then come back to me on the PT HD rig. It will be interesting to see if it causes any problems.

Probably not as long as he's keeping it in the box.
As soon as he would leave the digital world, sent a signal out through the DACs in order to process it with analogue gear and record it back through ADCs it would be polarity reversed.
However that would probably only be the case if he used the Digi Core driver.

We have not found out yet, whether the same would happen with a Logic/Symphony
system. But that is hard to believe.

[Peter Weihe]

Spindrift wrote on Tue, 11 January 2011 01:00

Forgive my ignorance:  What are the real-world ramifications of this if it turns out to be the absolute case? I know there is equipment out there in use that is wired Pin 3 HOT (Tascam DA-30 DAT deck for example) that some engineers don't know about...What is the end result? If a record is mixed to a DA-30 with Pin 3 still wired HOT, when a kick drum pounds, will the end-listener's woofer have negative excursion vs. positive? Or is it usually caught at mastering? Is it ever really a problem?

There is a AES paper by Milton and Greiner about a number of experiments about the audibility of acoustic polarity:

You can find it here:

http://www.aes.org/e-lib/browse.cfm?elib=6950

"AES E-Library
Observations on the Audibility of Acoustic Polarity

A number of experiments are described which show that absolute acoustic polarity is clearly audible in certain select cases of reproduced sound from acoustical instruments. The nature of the audible differences and the characteristics of the temporal signals which lend themselves to audibility are described. A large double blind listening experiment using varied musical program material is described as well.

Authors: Greiner, Richard A.; Melton, Douglas E.
Affiliation: University of Wisconsin Electroacoustics Laboratory, Madison, WI
JAES Volume 42 Issue 4 pp. 245-253; April 1994 Import into BibTeX

Click to purchase paper or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!

This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.

Learn more about the AES E-Library

E-Library Location: (CD aes5)   /jrnl8997/1994/8531.pdf"

[arconaut]

I tried to chase this down a bit more today but was really short on time. But I'll recap what I did do - created an audio file again using Signal Generator. I played it out of Pro Tools 8 and patched the output of the 192IO directly into a tie line to another room.

In the second room, I created a Pro Tools session and recorded the waveform being fed from the other room. Polarity remained intact. This was into a 96IO using PT8.

Returned to the "source" room and imported my waveform file into iTunes, which plays through the 192IO via DigiCore Audio. Then recorded this file to a second track in the other room. Again, polarity remained intact.

Then... I ran out of time. Next was to launch Logic on the receiving end and see if there was any difference there and after that to have Logic as the source via CoreAudio and Pro Tools recording once more. Eventually, the culprit will make itself known.

[Peter Weihe]

arconaut wrote on Thu, 13 January 2011 02:01

I tried to chase this down a bit more today but was really short on time. But I'll recap what I did do - created an audio file again using Signal Generator. I played it out of Pro Tools 8 and patched the output of the 192IO directly into a tie line to another room. In the second room, I created a Pro Tools session and recorded the waveform being fed from the other room. Polarity remained intact. This was into a 96IO using PT8. Returned to the "source" room and imported my waveform file into iTunes, which plays through the 192IO via DigiCore Audio. Then recorded this file to a second track in the other room. Again, polarity remained intact. Then... I ran out of time. Next was to launch Logic on the receiving end and see if there was any difference there and after that to have Logic as the source via CoreAudio and Pro Tools recording once more. Eventually, the culprit will make itself known.

Wow Noah, thank you for your effort!

That's a brilliant test method and you are right, the culprit will have to leave his hiding.

[Peter Weihe]

Sorry, Double post.

[arconaut]

Further snooping, here are a bunch of images. I generated a waveform and patched from the 192 IO output 1 into a tie line. Then I did the following things:

1. Recorded the waveform into ProTools on the 2nd system first using Pro Tools for source playback, then using iTunes via DigiCoreAudio for playback.

2. Recorded the waveform into Logic on the 2nd system, again using ProTools and iTunes for playback.

3. Recorded the waveform into ProTools again, this time using Logic and iTunes for the source.

Here is a shot of the original waveform:

[arconaut]

Here is a shot of ProTools, showing polarity remaining intact whether playback comes from ProTools (from source system) or from iTunes/CoreAudio:

[arconaut]

Next we have Logic as our recording platform, and you can see that using both PT and iTunes as the source, the polarity is now flipped.

[arconaut]

One more - this is recording into PT while using Logic and iTunes for source playback. Again, polarity is ok.



So the issue seems to be on input. But still, we don't really know if the issue is with Logic or with CoreAudio. For this, we'd need to eliminate CoreAudio.

But that's all the time I can dedicate to this for now!

Noah

[Peter Weihe]

arconaut wrote on Thu, 13 January 2011 17:45

One more - this is recording into PT while using Logic and iTunes for source playback. Again, polarity is ok. So the issue seems to be on input. But still, we don't really know if the issue is with Logic or with CoreAudio. For this, we'd need to eliminate CoreAudio. But that's all the time I can dedicate to this for now! Noah

Hi Noah, thanks again for your thorough test.
I can confirm the result.

I imported the sine wave from my test into I-tunes on a second Mac.

1. I recorded the output from the second Mac into Pro Tools on my Studio Mac and the polarity stayed intact.

2. I recorded the output from the second Mac into Logic ( with Digi Core) and the
polarity was switched.

You are right. The polarity Switch happens on the way in.

I asked some friend to perform the same test with a Logic/Symphony system.

3. Christoph played a sine wave from Logic trough the Symphony system with Apogee 16x DA converters and recorded it back into Logic with Apogee 16x AD.

The polarity stayed intact.

4. Then he imported the sine wave into I-Tunes and played it back through the onboard Apple analogue output.
He recorded it back into Logic through the Apogee 16x AD and.......

The polarity was switched!!!!

That means that with Christoph's Logic/Symphony setup the polarity is switched on the way in.

But it also means that the polarity is again switched on the way out!!!!!!

It seems to switch polarity twice.

That means that everything he records will be polarity reversed when being played back on a Pro Tools system.

That might be one of the reasons why a mix played back with Logic/ Symphony can sound different when played back with I-Tunes through the same converters.

I have asked another friend to perform the test with another sound-card and Logic.

[arconaut]

Well now my head hurts.

One possibility - not to cast doubt upon Christoph's results - is that his wiring is backwards? If he had pins 2 & 3 mixed up both on input and output, then wouldn't it appear that polarity was correct when feeding the system into itself, yet it was actually flipping the polarity twice? If that were the case, then it could be Logic is fine and Digi Core Audio's input causing the issue.

Otherwise, we now have a few more variables.

[arconaut]

I think the issue is Digi CoreAudio flipping the polarity on input. I just ran a test using Garage Band and had the same thing happen. There are some discussions in various places on the web about the output flipping polarity a few years back, and more recently others have found the same issue on input.


Edited to add:

http://duc.avid.com/showthread.php?t=262613

[Peter Weihe]

arconaut wrote on Thu, 13 January 2011 20:11

Well now my head hurts. One possibility - not to cast doubt upon Christoph's results - is that his wiring is backwards? If he had pins 2 & 3 mixed up both on input and output, then wouldn't it appear that polarity was correct when feeding the system into itself, yet it was actually flipping the polarity twice? If that were the case, then it could be Logic is fine and Digi Core Audio's input causing the issue. Otherwise, we now have a few more variables.

Yes, we have to check it again in another studio.
I have already asked another friend to repeat the test with his Logic/Symphony Setup.