Topic: Strange matrix-in-post issue

[tim in jersey]

This one really has me scratching my head.

Another taper and I recently recorded a Greensky Bluegrass show @ The Historic Blairstown Theater.

I had access to the SBD and grabbed a patch. 1/4">1/8">M-10. I also ran my AUD rig FOB mk41>kc5>cmc6>SD 722. The venue is a tiny theater and the show was sold out which made it very inconvenient to stop/start both decks so I just let the M-10 run for both sets knowing I had ample battery power and card space.

I was able to successfully do a matrix-in-post of my source (mk41>kc5>cmc6>SD 722). Results are here: http://archive.org/details/gsbg2013-08-02.mk41.mtx.flac16. Not my first time doing a matrix using 2 different sources from 2 different decks that are not jammed to the same clock. I've gotten pretty good at it and have been pretty satisfied with the results. Enough so that I can't yet justify purchasing a 4 track for the ~15-20% of shows where I have SBD access (even though my inner gear-slut tells me otherwise).

I use Sony Vegas and my matrix-in-post methodology is pretty simple and logical, IMHO. I knock down peak values of each recording to -6 to -9dB to avoid clipping once the two are mixed down. I import both as stereo files in Vegas. I pan the AUD hard left and the SBD hard right and use headphones to shift the events and "rough in" the delay at the beginning of the recording. I've found snare hits and/or stage banter are great for this, BTW.  I then pan each to center and listen for "the echo/delay" and shift them by  2 or 3 or 4 thousandths of a second to get them properly aligned and eliminate the "reverb". Then I snap to the end of the recording and use the same methodology using  the "shrink" recording function in Vegas. Has alway served me well....


So anyhow, I offered to do a matrix-in-post of my buddy's Neumann source (I was clamped to his stand). He sent me his raw, unedited source which is KM-184>V3>SD 722. I used above-mentioned method, and set 1 turned out great (same sample rate/bit depth etc.). Set 2 is a complete disaster. When I try to align the 2nd set opener I get a 1 minute discrepancy within the first 10 minutes, which is crazy big. Both are in pitch as near as I can tell...

I'm about to abandon this project.....

[page]

Just for reference:
your 722 = deck 1
your m10 = deck 2
his 722 = deck 3 (assuming it wasn't clocked to deck 1)

So his set 1 works through the entire set, and set 2 has problems even if you resync it from scratch at the beginning of the set (as though it's from a separate/fourth deck)?

[yousef]

My first thought is that if you're getting such a large discrepancy within such a short amount of time without any obvious pitch issue, there must be a glitch in one of the sources and some missing audio.

I had a maddening experience mixing an aud source with a multitrack tape which kept dramatically losing sync during the final 10 minutes. The band's producer eventually sussed it out: the multitrack had dropped an entire bar during one song... Somehow this had happened exactly on the beat and at the very end of a guitar solo so to the unfamiliar ear there was no audible glitch whatsoever.  Of course, this caused a sudden loss of sync rather than the more typical gradual drift - could one of your sources have multiple tiny dropouts?

Have you tried using a sync point later in the recording?

One unrelated thing occurs to me - it sounds like you're pre-processing the files before bringing them into Vegas to reduce the volume. I do this using the volume sliders in Vegas and so save a separate render on each file. I'm sure the difference is negligible but I always to have an absolute minimum of re-rendering.

[OOK]

did he pause his deck at any time? that might be the screw up... there is were your missing minute is...

[tim in jersey]

did he pause his deck at any time? that might be the screw up... there is were your missing minute is...

I thought of that too. Yes he did pause, but only between set 2 and encores near as I can tell. Other than that not as near as I can tell.

I usually sync the entire set in one pop, but I even went so far as to break each song in to individual tracks and try to sync seperately. I just found it really weird that both set 1 AUD sources synced easily to the SBD source  (only had to shrink the SBD by 2-3 thousandths) and the first song of set 2, which has no apparent drop-outs/pauses/pitch problems would be off by ~1minute.

I'm gonna futz with this some more and if I find more clues I'll post here and maybe we can all learn something. I'm beginning to think maybe set 2 was corrupted during the upload process. And yes, I downloaded the Neumann source several times and ran a checksum to be sure they were the same file...