Taperssection.com
Gear / Technical Help => Post-Processing, Computer / Streaming / Internet Devices & Related Activity => Topic started by: poorlyconditioned on June 19, 2006, 11:51:58 PM
-
I've tried a few matrices where I have different clocks. For example, a UA5 for mics and a standalone CD recorder from the soundboard.
I noticed that the vocals tend to "waft" in and out. There is a definite improvement using the soundboard in the mix, but it is not as nice as if I run using the same clock. It even sounds better if I do an on-the-fly mix with a two-track, provided the PA delay is less than say 10ms.
I guessing these are phase issues, since flipping the sign/phase on one track sometimes helps, but only in places. But it is not stable over the recording, or even the song for that matter.
I'm thinking the vocals in each source are partially cancelling each other out. Say the vocals are at 2kHz. That is a wavelength of 0.5ms or 20k samples. Is it possible to line up to less than 1ms even? Would it even be stable (eg., for song-by-song alignment)? Is there a plug in that can vary the phase at a particular frequency? I'm using Wavelab 4.0.
Comments?
Richard
-
Different recorders def record at different speeds. The phasing will more or less be inevitable
if you line it up perfectly at the beginning, it will no doubt go out of sync later in the recording, so you need to check it in multiple places and corrent as necessary
splitting up the tracks and syncing track my track may be more accurate
-
Different recorders def record at different speeds. The phasing will more or less be inevitable
if you line it up perfectly at the beginning, it will no doubt go out of sync later in the recording, so you need to check it in multiple places and corrent as necessary
splitting up the tracks and syncing track my track may be more accurate
Yeah, I split into tracks, but there is still phasing. I think this is *really* subtle. Like on the order of less than 1ms. I can line up to below 10ms by eye, and maybe better by ear, but I don't know if the time will even stay accurate to within 1ms over a single track even.
Requires more experimentation I think.
Richard
-
it might be, depends on how different the recoding speeds are
or maybe line up the track to the middle on the other track , minimizing sync problems at the end of the track
-
Comments?
Yep...two:
#1) Get an R4! ;D
#2) You should be able to minimize any phase issues in Wavelab. First, split your SBD (or AUD...doesn't matter) copy in to tracks. Open a new montage in Wavelab and add your two stereo tracks from the UNSPLIT mic source. Begin laying in your SBD tracks one at a time. It's easier to find and align a reference point on one end of a small track than it is to find it on one end of a LONG track. Once you have your starting point synced up, you can stretch or shrink the physical wave length to match time. Once you get the track in line, drop in the second track and repeat. Complete this until you're done lining up your tracks. I've fixed even the funkiest of delays by following this procedure. I can't see how a particular frequency would be delayed any more than another, so it's likely the entire PA wave you need to shift.
#3) See #1.
Mike
-
you could rerecord one of the tracks(or both) into the JB3 line in so that they are clocked the same.