Re-Amping Tracks in Logic has me Worried - Defying Laws of Physics?

Question

Level 1

30 points

Re-Amping Tracks in Logic has me Worried - Defying Laws of Physics?

I recently got into re-amping guitars, basses, keys, etc. and the results are fantastic. Its really an incredible technique thats given me more power and flexibility on all of my projects, both live and in the studio. However, while re-amping bass tracks today I discovered something that just doesn't seem right at all.

I'm routing the original DI'd bass track out of Logic, into an ADAT hd24XR via lightpipe. I send the analog output from the XR into my re-amping box. The output of the re-amping box goes back into the XR for A/D conversion, then back into Logic via lightpipe.

Heres the problem: When I zoom in on the original DI bass track and the new re-amped track, the re-amped track looks like it happens BEFORE the original track. Every waveform is clearly slightly ahead of the original signal. I have no idea how this happened, and its making me a little worried about all of the overdubs I've been doing in Logic. Someone must know whats going on, thanks in advance for the help!

Powerbook G4 1.67 ; Core Duo Mini, Mac OS X (10.4.6), 2 gigs RAM, ADAT HD24XR, 3 Presonus Firepods, LaCie 160 gb FW

Posted on Feb 8, 2007 3:17 PM

Reply

Answer 1

iSchwartz

Level 5

4,321 points

Feb 8, 2007 3:50 PM in response to over-man

First you need to do a loopback test on your system to determine the correct "recording delay" value. Details of how to do that are found in numerous threads here on the forum (do a search for "loopback").

There's even a chance that your recording delay value is set to a positive value, which would explain a few things about what you're experiencing. So that's the first thing to check. Look in Prefs > Audio > Drivers. If there is a positive value, or a value of zero, that would be an immediate indication that these values are wrong and need to be set via a loopback test.

Reply

Answer 2

over-man Author

Level 1

30 points

Feb 8, 2007 3:56 PM in response to iSchwartz

iSchwartz,

You're the man, thanks for the quick response. My recording delay value is, and has always been set at 0.

I'll admit I haven't done a loopback test before, but how in the world could the audio come before the original audio that triggered it, especially when the recording delay is at 0?

Reply

Answer 3

Feb 8, 2007 3:57 PM in response to over-man

Sounds like you have plugins on some of your tracks, and PDC is on. IIRC, Logic delays tracks to compensate for other tracks that have latency. This would have the effect of your new recording being early when compared to existing tracks. When routing out to external hardware, I turn PDC off and in any event I avoid using latency inducing plug ins until I've finished tracking.

There's an apple article about this (which I can't find at the moment) that explains all this in detail. Try searching the Apple knowledge base for 'delay compensation.'

Hope this helps.

Reply

Answer 4

over-man Author

Level 1

30 points

Feb 8, 2007 4:04 PM in response to John Alcock

John,

I do have PDC set to ALL, and channel eqs on some of the tracks. I'll try switching it off to see what impact that has, along with running the loopback test. Thanks guys.

Reply

Answer 5

iSchwartz

Level 5

4,321 points

Feb 8, 2007 4:16 PM in response to over-man

iSchwartz,

You're the man, thanks for the quick response. My
recording delay value is, and has always been set at
0.

: - )

I'll admit I haven't done a loopback test before, but
how in the world could the audio come before the
original audio that triggered it, especially when the
recording delay is at 0?

Was anyone in your family ever involved in the production of The Twilight Zone?

: - )

I've just given this some thought and it made my brain hurt. I'm not sure how that could happen...

Reply

Answer 6

over-man Author

Level 1

30 points

Feb 8, 2007 4:17 PM in response to over-man

When I disable PDC and all plug-ins, the tracks actually move even further ahead of the original, & the first dub w/ PDC on. Not by much, but a little still. About to try the loopback test.

Reply

Answer 7

over-man Author

Level 1

30 points

Feb 8, 2007 5:06 PM in response to over-man

Well, this is bizarre. I did a loopback test with a snare sample, sent to the external D/A convertor, straight into the A/D convertor (both on the XR), recorded back into Logic. The closest I could come to having the files phase out was by setting the recording delay to +48, but they don't cancel out completely. I tried other variations, but w/ +48 the waveforms line up almost exactly, and it seems to have the most cancellation.

Are they not completely cancelling due to the A/D/A conversion?

Also, how common is it to have a +48 setting for recording delay, and how does this impact my regular tracking or overdubbing?

I'm very confused, and its a little disconcerting that I can't make them cancel completely.

Reply

Answer 8

iSchwartz

Level 5

4,321 points

Feb 8, 2007 5:19 PM in response to over-man

This doesn't make sense. It's extremely rare for the recording delay to have to be set to a positive value.

And you should be able to get the sounds you use for your loopback test to cancel perfectly, resulting in sheer silence.

Did you do this test without any plug-ins on any channels, or, with PDC turned off completely?

Reply

Answer 9

over-man Author

Level 1

30 points

Feb 8, 2007 6:11 PM in response to iSchwartz

iS,

I must have inadvertently discovered the secret to time travel, some sort of hole in the quantum structure of our universe.

I made sure PDC was totally off, and I was in a blank session w/ no plug-ins, only a little snare sample. Good times.

Reply

Answer 10

iSchwartz

Level 5

4,321 points

Feb 8, 2007 6:20 PM in response to over-man

I've got to see this for myself! If you're willing...

...and With recording delay set at zero...

Take a CD track and import it into your arrange. It can be anything, but preferably something with drums or sharp transients so that peaks are easily seen.

Do the loopback record. (Make sure your levels going out are 0 dB.)

Zooom waaaaaaaaaaaaaaaaay in to the very beginning of the waveforms as shown in the arrange window and take a screen shot. Then post it. If you don't have the facility to post pics, write again and I'll give you an email address you can send them to.

-=iS=-

Reply

Answer 11

over-man Author

Level 1

30 points

Feb 8, 2007 6:24 PM in response to iSchwartz

iS,

I just left the studio for the day, fed up with more techno bs... but, I'll post a pic tomorrow morning if I get a chance before my session. I zoomed in on the arrange page all the way, and even used the ctrl+mouse scroll to zoom in even further. This let me line the start of each waveform up on the very edge of the screen, and they are most definitely offset. Pics on the way tomorrow...

Reply

Answer 12

Feb 8, 2007 8:14 PM in response to over-man

i'm not sure i understand what the big deal is. does it sound ok? who cares what it looks like?

Reply

Answer 13

iSchwartz

Level 5

4,321 points

Feb 8, 2007 8:44 PM in response to over-man

I think what's happening here is that for some reason your system (FW driver software, who knows...) thinks that anything that's recorded needs to be advanced by some 48 samples. The only clue I can come up with re an explanation (which may be totally entirely completely wrong) is that you're using FW interfaces. Would you know if they, like other FW-based interfaces, are succeptible to a variable amount of latency each time you boot (I think this is a thing with Apple FW drivers)? Or are the FW drivers some kind of proprietary RME thing that will avoid that anomaly?

Anyway, I'd be curious to know the results of your loopback test after two subsequent restarts. I think it's necessary to rule out any FW-based driver latency weirdness first.

Best,

-=iS=-

Reply

Answer 14

over-man Author

Level 1

30 points

Feb 8, 2007 9:17 PM in response to iSchwartz

iS,

I was just going through my original post and realized that I forgot to fully describe the setup, I left something out. The lightpipe interface between my mini and XR is....here it comes, that dirty word....M-Audio (cringe) Lightbridge. After all the problems I had with Firepods and firewire, I can't believe I didn't think of this sooner.

I think you are right on, I'll contact M-Audio tomorrow. They are notorious for bad drivers, the product is new, and its firewire.

Why two subsequent restarts? I've heard mention of the firewire issue before, but I don't know the nitty gritty details.

Reply

Answer 15

iSchwartz

Level 5

4,321 points

Feb 8, 2007 10:19 PM in response to over-man

o-man,

As I understand it (someone please correct me if I'm wrong) Apple FW drivers are notoriously flawed in that each time you start your computer, a new value for the FW driver's latency value (delay needed for processing) is generated. So my idea was to do a loopback test twice (one after each restart) to see how much -- if any -- variability there might be in setting your recording delay between the two restarts.

In any event, I understand that the amount of latency, once set at bootup, doesn't change. So once you calculate the proper value for basic recording delay -- whatever it might be, we can then set about trying to figure out the additional advance/delay needed to get your re-amped tracks to be in time.

Whew!

Reply