"fine/sample delay", mixing down live drums, esp for JustinC!

Question

I'm curious about the "fine/sample delay" and "phase inversion" techniques briefly mentioned by JustinC in an older post regarding recording live drums. I understand using phase inversion on snare mics that are pointing up/down, but I'm not 100% sure about the fine delay adjustment. Is this actually shifting the audio to line up waveforms from mics that are different distances from the same source? I've done this with an upright bass before (F-hole mic and another a few feet away) and it seemed to get a much fatter, clearer sound. I really appreciate any tips you might have on how to pinpoint and solve the problems.

Answer 1

hey over-man,
I was prowling tonight and ran across your post. I don't know your reference specifically but I understand your question.
Sample delay regarding a multimic'ed source:
This is based on phase cancellation and reinforcement. It is initiated by everything in a multimic'ed environment. The times, distances, microphone choice, effects choice-it's everywhere.
The best approach for me (typically) is to get the most appropriate sound earlier on unless you can expect variables ahead of time and factor them in right away. We'll work with an example of simply getting a natural drum sound that will line up with existing tracks (bass, gtr, vox...).
1) You can set up the sound/mics as close as possible to fit the sound you're after (we'll breeze over these points not specific to your question).
2) Before destroying the sound with a handful of plugins adjust delays. Some assumptions can be made:
A: Overheads/Room Mics will typically be the reference for the reason that they are what you have least control over because they represent more sources (Drums in this case).
B: These Mics will often have the largest delay from the start for the reason that they are farthest from the source.
C: These Mics should be adjusted less than others, I would EQ them quite minimally. Solve any problems as early in the signal chain as possible and use only what you know you'll need. Do not apply FX you don't need and position mics to better suit your desired sound.
So your OHs/Rooms are as good as they can sound naturally and sit with the track as well as possible, you can now start adjusting delays/intrtoducing other elements.
The order can vary here but I tend to start with the tracks/mics which will be more prominent. I will use snare in this example.
Some useful tools before we resume: Logic's sample delay, Expert Sleepers' Latency Fixer, because Logic's Sample Delay will not predelay. Found here:
http://www.collective.co.uk/expertsleepers/latencyfixer.html
I would normally just load both of these up, I use full PDC and apply these AUs to each track-remember to turn off Software Monitoring!
Create an instance of Latency Fixer, set it to 2048 samples. You have now added 2048 samples of predelay to your track.
Create an instance of Sample Delay, set it to 2048 samples. You have now adjusted your track to play back at the original record position.
Copy both instances to each adjacent track.
So we now have Latency Fixer followed by Sample Delay on each track for the kit (not necessary for the OHs unless you want to shift everything a little, but I will often do this from the arrange by ms).
At this point you will not need to adjust Latency Fixer's settings and simply use sample delay to offset delay in either direction.
Bring the snare track up to a reasonable level (the closer you are the better). Adjust Sample Delay in either direction and you will notice the sound changes from phase cancellation/reinforcement. Certain frequencies will stand out more at different settings. It normally won't take much, we don't want it to sound like a delay. Someplace within +/- 300 samples is often best for a natural sound. It's a good idea to have a phase inverter (Gain PlugIn) handy for 2 reasons: It may sound better inverted. You can invert the phase to pronounce things you DON'T want to hear and then flip it around the right way after setting the delay to pronounce them. Volume has a huge bearing on interaction here, a good delay setting may sound bad after reducing the level 2dB so Sample Delay setting, volume and phase should all be updated as soon as possible.
By now you have noticed how adjusting the delay affects the sound, inverting the phase is an exaggerated example of this principle. Inverting the phase of two identical tracks cancels them entirely, applying fine delay to one will introduce some of the frequencies and leave some cancelled. You can hear it scan through which frequencies by adjusting the delay. With multiple sources, different mics and different distances these delays already exist. Delaying them so they match (based on distance) is not always the best sounding but try to start somplace close. After setting the delay and Volume as close as possible you should be able to get away with much less surgical types of PluggIng and ultimately have more natural sounding tracks. Will often comprise 75% of the sound (if you want natural tracks).
If there are dynamic effects you know you'll want to apply there are special considerations.
Many of Logic's "lookahead" paramaters delay the signal.
Compression/Gating, etc. (on a per track basis) will alter the volume significantly enough that if you know you want a track compressed... ahead of time you should probably set these as you would then adjust the delay.
I tend to use predelayed tracks as keys for the audible tracks, this allows you to get around the aforementioned lookahead problems and will provide you with more control over the "shape" or envelope.
Concerning phase inversion based on which head you mic: It is a generalized solution to mic setup, though not a hard and fast rule which applies in every case.
Delaying in this manner is aligning waveforms. Based on frequency you are causing some frequencies to stand out more or less since we are working with complex waveforms. Very small deviation from original position will affect higher frequencies most and as you get further away from the original postion it begins to affect mids and eventually low frequencies. Of course this is because the frequency/wavelength. Align and misalign the waveforms for your desired sound. It does require you listen very closely when setting it up but I really think the results are worth the effort. Hopefully you'll find this helpful, it's easier to hear the effect than to describe it, hopefully it is apparent once you hear it.
Finally, I don't have any Platinums. If it doesn't work better for you, don't believe it needs to. Hopefully others will chime in here so you can get some second opinions on the approach and/or general drum mixing techniques.
Cheers, J