Drum Editing With Mic Bleeding

[joe_04_04]

I just got my m-audio fast track ultra 8r, and so i wanted to jump right in and use my 8 piece mic set on 3 new songs. I tried to mic it the best way possible to avoid bleeding, but this is always a problem. I was wondering what the best techniques are that allow you to either use quantization (elastic time) or beat detective without getting severe phasing problems. Obviously EQing and cutting out where the drums bleed are big factors..just wondering if there are a whole slew of tricks out there..since im new to micing this many drums. By the way, I have 2 Overheads in an X-Y coincident, 3 tom mics, a bass drum mic, snare, and a pencil condenser on the hi hat.

[KingFish]

you have to group, and edit ALL THE TRACKS in the kit, together - "As 1 Instrument" to avoid phasing. you can't move a "Out of place Snare" on the snare track, rather move all 8 tracks from the snare transient, to the grid line.

Here's a Beat Detective tutorial

which will NOT apply to the ProTools Essential version, only LE with a Toolkit, or HD but you'll see the process of drum quantization in the clip.

[joe_04_04]

I've always mostly understood this, but it still is a bit unclear. Because moving the "out of place snare drum" may cause other instruments to be out of time, if the drummer was that off (which isn't uncommon these days). Also, maybe someone can elaborate on the pros and cons with correcting the tempo within a multitracked drum session with elastic time vs beat detective.

P.S. - hows come the gentlemen in this video isn't having any phasing issues, im probably mistaken, but when he deletes the incorrect quantization, and moves the drum individually, it would seem like that would pose a problem. He is still working in grouped edit mode so im guessing all of the tracks are moving, but doens't this cause the other tracks to be out of time? And he says he will discuss phasing problems, but never does (lol).

http://www.youtube.com/watch?v=eFTSiFt01uU

[DigiTechSupt]

Let's see if I can't clear up some of the confusion.

When you record multi-track drums, you have several mics that are picking up bleed from drums other than the drum the mic is intended for. For instance - the snare mic will probably have a bit of the toms, hi-hat and cymbals in it.

If you were to edit/quantize just the snare track, the bleed from the other drums will now 'smear' or phase compared to their primary mic, which is why you want to edit/quantize ALL tracks, so that those relationships are maintained and phasing/smearing is kept to a minimum.

The reason you're not hearing phasing in that video is that, while he's moving an individual track and it's moving all other tracks, it's only moving the audio around the edit point, which will contain bleed from the primary drum track that was just moved.

Now keep in mind that, while these tools are powerful, there are limitations to what they can do. If the drummer was really bad, timing-wise, there may not be anything you can do to fix it and have it sound 'great'.

With EA, what you're trying to identify on EACH track are the primary transients that the mic is picking up. If you have those set correctly the quantization should work properly as long as you're not moving something so far from it's original, recorded position that the algorithm causes artifacts.

The difference between EA and Beat Detective is that EA maintains a continuous file, while Beat Detective actually chops the audio up into regions and then moves the regions. With Beat Detective, if you move things too far you can get gaps in the audio. Also with BD if there is a lot of bleed on a mic, you can get phasing/echo problems if a region has bleed and it's moved in relation to the primary drum track.

Example - let's say I've got a snare track with hits on 2 and 4 (or should have been if the drummers timing was good, but it's not *quite* there). It's got bleed from the hi-hat that's relatively loud on the snare track and the hi-hat is played on 1-2-3-4. If I use BD to quantize the snare and move it so it falls spot on 2 and 4, that bleed from the hi-hat has also moved and you'll get phasing/slap/echo depending on how far you had to move the snare to get it in time.

With EA, you have markers on the snare track that also move the hi-hat track and vice versa. The snare track markers would be on the snare hits and the hi-hat track markers would be on the hi-hat hits. If you now quantize the snare track, it moves the audio on the hi-hat track around the marker points that the snare track had. Same goes for the hi-hat - if you quantize it, it also moves the audio on the snare track in the same proportion. This keeps the relationships between the primary track and any bleed you may have in other mics intact.

A vast majority of phase issues I hear complaints about are due to 3 factors:

1. Make sure you have markers set at the beginning (within the region, but before audio) and end (after the last audio, but before the end of the region). This 'locks' the area you're going to quantize properly. If you don't have them, it moves audio from the beginning or end of the actual region and can cause weird anomalies.

2. The amount of quantization is too much - even the best algorithm sometimes will have problems if you're moving things too far. This usually occurs more often when a region is stretched, rather than compressed.

3. Incorrect markers - they're not being placed properly or there are 'invalid' markers - ones that are not on an actual transient.

It's important to have an understanding of what a tool is doing to comprehend what, exactly, may be going wrong. User error is a big factor, limitations of the tool itself are next (trying to make it do something it wasn't intended to).

Hope this helps.