EEK!! HELP!! Song mysteriously changes key!?!

[bgrotto]

Last night I was doing a freelance gig at a studio I hadn't ever worked in. We tracked everything into Pro Tools at 44.1/24-bit, but ran short on time, so we couldn't do rough mixes at the studio. I told my client I'd do a coupla quick bounces for him when I got home.

I bounced the files using Pro Tools' MP3 convertor, at 44.1, and emailed them off. Today, the client called me and told me that the songs were a half-step higher than what we recorded them at!! However, the tempos are exactly the same (they played both tunes to Pro Tools' Click, and I have since then referenced the rough mixes and PT sessions to my own hardware metronome), and there's no apparent loss in sound quality that I would typically associate with vari-speed/pitch shift/etc.

SO WHAT THE HELL HAPPENED!?!? Does anyone have an explanation for this?? I've done countless sessions bouncing between studios, PT rigs, etc, and NOTHING like this has ever happened.

What gives??

[John Suitcase]

I'd say that if the tempo still matches (check the MP3 against your rough mixes against the raw tracks), then it's not clock/sample rate related.

Having said that, I'd be surprised if that's the case. It sounds like a sample rate issue to me.

If all three sound ok to you, that is they all match up, then I'd be inclined to say that the problem may be with the client's machine, or somewhere in between.

[tINY]

What happened, probably, is that the song was recorded a half-step higher than what your client thought.

Is this guitar, wind, vocal, string.... or keyboard music?



-tINY

[bgrotto]

Originally Posted by tINY

What happened, probably, is that the song was recorded a half-step higher than what your client thought.

Is this guitar, wind, vocal, string.... or keyboard music?

It's drums, percussion, bass, guitar, organ, and steel drums.

The organ and pans cannot really be tuned...that's why we are all quite certain the tune was recorded in the right key.

I've been thinking: maybe PT's clock was at 48k, even though the PT session was at 44.1. That'd probably account for the 1/2 step change. I dunno. It's very frustrating.

[Mike Caffrey]

That's similar to what I was going to suggest. I think there was probably an external clock becuase I don't think the mismatch can happen internally.

[bgrotto]

Originally Posted by Mike Caffrey

That's similar to what I was going to suggest. I think there was probably an external clock becuase I don't think the mismatch can happen internally.

Yeah, the confusion for me was that the clients told me the key was shifted UP a half-step...in fact, it was shifted DOWN a half-step.

I contacted the studio, and they do indeed have an external clock, and their standard practice is to run all sessions at 48k. So...mystery solved.

Now, I was wondering if there's a way to convert my 44.1 session so that I don't have to make sure I'm clocking at 48k for the rest of this project!!

Any suggestions?

[tINY]

Sample rate converters.



-tINY

[e-cue]

Originally Posted by bgrotto

I contacted the studio, and they do indeed have an external clock, and their standard practice is to run all sessions at 48k. So...mystery solved.

Not exactly... How did it change pitch, but the tempo remained exactly the same?

[bgrotto]

Originally Posted by tINY

Sample rate converters.

Heh. Thank you for your insiteful answer, but I think you're missing the point. I wanna change the SR for the Pro Tools session; the audio is fine and doesn't need to be resampled.

I was asking if there's a way to trick PT into letting me drop the audio into a new 48k session without it changing the existing samping rate (which is correct, just in a mis-clocked PT session...get it?).

I want to avoid any resampling of the audio so as to best preserve its integrity.

[bgrotto]

Originally Posted by bgrotto

I was asking if there's a way to trick PT into letting me drop the audio into a new 48k session without it changing the existing samping rate (which is correct, just in a mis-clocked PT session...get it?).

I want to avoid any resampling of the audio so as to best preserve its integrity.

Eureka! I got it. For those curious:

1) create a new 48k session
2) Import Session Data from old 44.1 session WITHOUT using SRC
3) Booyah.

That was WAAAAY easier than I thought it'd be. If I wanted to keep the session at 44.1, I could do this:

1) create new 44.1 session.
2) Importa Sessioin Data from old session WITH SRC
3) Booyah.

That, of course, would degrade the audio some. But it'd be at 44.1. Hooray.

[Mike Caffrey]

Why not stay at 48 until the end?

[toolskid]

Originally Posted by e-cue

Not exactly... How did it change pitch, but the tempo remained exactly the same?

I'm curious about that too!

[tINY]

I'm guessing that the tempo change was just less obvious than the pitch change. Though you'd think that 10% would be noticable....




-tINY

[toolskid]

Originally Posted by tINY

I'm guessing that the tempo change was just less obvious than the pitch change. Though you'd think that 10% would be noticable....

-tINY

I'm pretty sure he said that one of them lined up the two differently pitched tracks together in tools!!!!!

still room for operator error tho I suppose!

[bgrotto]

Originally Posted by toolskid

I'm pretty sure he said that one of them lined up the two differently pitched tracks together in tools!!!!!

still room for operator error tho I suppose!

So here's what happened, guys:

The studio's clock (which, embarassingly, I was unaware of) was set to 48k. I created a 44.1 PT session, and recorded the tunes as normal. When I got home, I bounced rough mixes, clocked at 44.1, and emailed them to the client. He told me the mixes were pitch a half-step HIGHER; this is where the confusion set in. In fact, they were a half-step LOWER, which if he'd have told me, I'd have known that the PT system at the studio must've been clocked at 48. I started thinking that maybe there was indeed a clocking error on my part, so I simply set my clock to 48, created a 44.1 PT session, and played a single note on a guitar into the session. Then, I reset my clock to 44.1, and the pitch was indeed lower by just UNDER a half-step.

Sooooo (whew), here's what happened with the tempo: the original 44.1 session had the click set at 66bpm. While we were at the studio (with the clock running at 48), what the screen said was not in fact what the click was playing, as the session was misclocked. That's why when I got to my house, and clocked the session at 44.1, the tempos lined up. Brain-teaser, right? I feel like I'm doing an SAT math problem.

Regardless, everything's cool. I didn't have to do any SRC, and the songs play back at the right tempo and in the right key. Plus now the snare cuts through more clearly!!