Value of phase correcting multiple mics?

[melonmachine]

I'm about to start an acoustic recording, most songs have many different instruments, which will be dubbed individually.

I'm thinking of using multiple mics on most things, so that I can have a choice of different tones to combine / choose from when I mix, without resorting to a lot of EQing.
Also I have two pairs of very different sounding pres - Altec 1567a and early 70s QuadEight, so again different sounds to chose from.

I'm recording with PT and was thinking that I would put all the mics into the same phase by aligning their timing.

Is this a good idea? My guess is that with everything in phase it should be easy to combine mics, and with no spillage from multiple mics it shouldn't throw up problems.

Have I missed anything though? Is this actually a good idea or a naive mistake? I've never made one of those!

Giles

[Wayne]

Originally Posted by melonmachine

...I'm thinking of using multiple mics on most things, so that I can have a choice of different tones to combine / choose from when I mix, without resorting to a lot of EQing.

On the plus side; Lots of options for experimentation on mix time. The down side; 'Lots of options for experimentation on...'

I do this when the opportunity is an option just to get time on different mics in various applications -but it can also turn into a royal pain if you haven't yet paired it down to the 'keeper(s)' come overdubs, "let's try a few more and keep those" or worse -recall time..!

Anywho.. better to start with them distanced' to sound at least useful in combination at tracking. This could be trying for equal source-mic distance for best shot at 'in phase', or 'out of phase but complementary', rather than dealing with track sliding on top of picking mics at mix.

... and with no spillage from multiple mics it shouldn't throw up problems

I don't get that part?

[bgrotto]

My approach is to use Digi's Time Adjuster plugin to experiment with lining mics up. You can easily bypass it, so you can actually A/B whether or not it's doing anything good.

Generally speaking, I don't have great luck with lining stuff up; the sound I got was the one I want, phase-coherent or not! Often, it's the comb-filtering and such that gives a sound its 3-dimensionality.

That said, I'm in the middle of tracking a record for roots/americana musician Tony Furtado. If you're unfamiliar with Tony, he's a really amazing multi-instrumentalist; he'll rip on just about anything with strings and a fretboard. I set up a bunch of mics in a sort of "acoustic instrument station", and in addition to experimenting with different combinations for his many different instruments (and styles!), I fiddled (no pun intended) a bit with re-adjusting the timing on some of the mics. For some sources, it really worked wonders. It brought his Dobro to life, and saved the day on an acoustic guitar track.

So, yeah, it's definitely worth experimenting. I'd just advise against actually moving the regions themselves; using the sample delay plug gives you waaay more flexibility, and it's undoable.

[drumkideric]

Personally, I prefer getting the mic placement right, to avoid most phase issues. Just listen and adjust the mics as much as possible. I find aligning it digitally takes away that depth and space, but that is just me.

[Mike Caffrey]

Originally Posted by melonmachine

I'm about to start an acoustic recording, most songs have many different instruments, which will be dubbed individually.

I'm thinking of using multiple mics on most things, so that I can have a choice of different tones to combine / choose from when I mix, without resorting to a lot of EQing.
Also I have two pairs of very different sounding pres - Altec 1567a and early 70s QuadEight, so again different sounds to chose from.

I'm recording with PT and was thinking that I would put all the mics into the same phase by aligning their timing.

Is this a good idea? My guess is that with everything in phase it should be easy to combine mics, and with no spillage from multiple mics it shouldn't throw up problems.

Have I missed anything though? Is this actually a good idea or a naive mistake? I've never made one of those!

Giles

I think it's a bad idea because it does not solve the problem.

Arguably you could excessively EQ one of the mic and use that phase shift to solve the problem, but that doesn't either.

First identify the problem - when two mics are out of phase, the sound bad, not because they're out of phase, but because of the combfiltering (which happens because they're out of phase.)


The reason people think time alignment fixes phase issue is because phase is taught by
showing two sine waves moving one (in time).

IF you're recording sine waves in an anechoic chamber with identical mics, time alignment will work, but you're not.


If you record a source with a 57 up close and an 87 far away, you're going to have two different sounds. So if you align the wave forms based on time, you just creating different phase issues, not solving them - and if they happen to be 180 degrees out with the time align, you've made it worse.

IF you use two identical mics, they'll still have different sounds form they're location in the room (there are phase issue caused by walls - think about the summing effect of reflection bounding off the ways and summing with the source).

So, if you want to use to mics that can be successfully time aligned to improve phase, you pretty much have to set them up in an XY pattern with will make them time aligned in the first place.


There are three ways to properly fix combfiltering due to phase issues - changing mic placement, flipping polarity/phase and using the Little Labs IBP box.

Time alignment is just a method to change the combfiltering.

In then end, combfiltering is the problem you want to address, so on occasions, changing the alignement or regions can improve the sound, but there are other better methods that dont' kill the purpose of the distant mic - part of the sound of a distant mic is the time delay.


So keep the delay, and fix your combfiltering with a different method.

[steveschizoid]

Originally Posted by Mike Caffrey

If you record a source with a 57 up close and an 87 far away, you're going to have two different sounds. So if you align the wave forms based on time, you just creating different phase issues, not solving them - and if they happen to be 180 degrees out with the time align, you've made it worse.

IF you use two identical mics, they'll still have different sounds form they're location in the room (there are phase issue caused by walls - think about the summing effect of reflection bounding off the ways and summing with the source).

There are three ways to properly fix combfiltering due to phase issues - changing mic placement, flipping polarity/phase and using the Little Labs IBP box.

I have to take the devils advocate. By the above reasoning, flipping the phase will also only create different comb filtering. I think there are some situations where phase flipping works best (mic'ing a source - like a speaker cone or a drum head - from both sides at the same distance) and others where time aligning works better (for example where two differently placed mics in front of a guitar cabinet create an issue). Of course, in either situation, you'll get the best result if you are careful with mic placement. In the former case, the more the two sounds phase cancel before phase flipping, the better the result. In the latter case it will sound best if there is no issue in the first place, but, since the two will sound different irregardless of time alignment, I don't see the harm in lining them up - especially if phase flipping doesn't help (which it won't unless they are somewhere near 180 degrees out of phase, right?). If it sounds better, it is better.

The IBP box looks interesting. I wonder if it has a sound of its own though.

[Full Clip Audio]

A lot of people may disagree but I personally agree with Bgrotto and think that the phase "issues" and comb filtering is what makes some things like acoustic guitar sound incredible in many cases! I of course always try to find a balance between that huge, deep spaciousness and a mono compatible, focused sound. One of my favorite things to do is to set up the mics with just enough difference in distance that when you flip the phase of one, it folds down into a very centered and focused sound and when left alone it has a very exaggerated wide sound with a scooped feeling middle. Then when there are two guitar parts I will flip the phase on the mic on one of them (usually the lead part) and it has its own perfect place to sit in the middle of this ultra wide image. I use this sort of thing to make nice pockets for things to sit and am still able to keep things sounding good when folded down into mono.

[rockstar_josh]

Yeah, I think it goes to the idea of tracking the way you want to hear it... rather than getting the hottest, most coherent signal and then adjusting it all during mixdown you can use mic placement to set yourself up for depth and ambience, saving yourself some time later. Sure, you don't want cancellation or obvious comb flitering, but the different delay times from source to the various mics in place are what adds to the perception of depth and roominess.

[Mike Caffrey]

Originally Posted by steveschizoid

I have to take the devils advocate. By the above reasoning, flipping the phase will also only create different comb filtering. I think there are some situations where phase flipping works best (mic'ing a source - like a speaker cone or a drum head - from both sides at the same distance) and others where time aligning works better (for example where two differently placed mics in front of a guitar cabinet create an issue). Of course, in either situation, you'll get the best result if you are careful with mic placement. In the former case, the more the two sounds phase cancel before phase flipping, the better the result. In the latter case it will sound best if there is no issue in the first place, but, since the two will sound different irregardless of time alignment, I don't see the harm in lining them up - especially if phase flipping doesn't help (which it won't unless they are somewhere near 180 degrees out of phase, right?). If it sounds better, it is better.

The IBP box looks interesting. I wonder if it has a sound of its own though.

Yes, changing polarity only changes the combfiltering.


Aligning close mics is the only time where I think it's likely to be a good idea, however it would be better t place them in a way in the first place that didn't require aligning.

Besides, if they are two different mics with different paths, there's no reference to align the waves. The waves that you see are a representation of the sound and as mic diaphrahms of different materials react differently, aligning to identical looking parts of the wave may not be proper time alignment.

So I'm still against the idea of aligning waves/regions. I'm not against moving them randomly until they sound good. In the end it's all just changing the combfiltering.

I think that an engineer who makes the arbitrary choice to set a limit of this one particular option - in other words, decide to never time align tracks - in the long run that engineer will change other behaviors and end up happier with their sounds in the long run.

[Paul Hammond]

I asked the exact same question as the original poster - many years back - to a well known engineer with multiple gold and platinum records. He told me to "get a life". I knew exactly what he meant. The stories of Roy Thomas Baker recording Queen's drums made me forget all about phase anomalies in multiple mic kit setups. Cheers, PH

[Mike Caffrey]

In an era where record companies maintained barges in NYC's East River that had their own zip codes so that they could ship two million units the week an album is released, it's pretty hard to not have multiple gold and platinum albums.



Have you ever heard any of the raw Queen drum tracks? Mixed the sound great, raw, pretty ratty, but there are certainly no phase problems.

[Sinewave]

There are three ways to properly fix combfiltering due to phase issues - changing mic placement, flipping polarity/phase and using the Little Labs IBP box.

I believe the Neve Portico series has a phase alignment box also.

[melonmachine]

thanks for all the in depth answers guys. Plenty to think about!
Why is it that whatever I ask about recording I get the same two answers:
1 it's not that simple.
2. listen for what's right.
oh yeah, because......

i think I'll try it on some test tracks to see what I can get out of it. It would help when I've got someone coming for a short session. I could set up a lot quicker.

BTW, Wayne, what I meant about not having spillage is that if there's 2 different sources on the same mic I'd have thought it'd be impossible to time align them to be in phase

cheers, Giles

[msmith]

Good luck with that...

Moving a pre recorded track in "time" is not the same as repositioning the mic in the "space" of that room...

[beebay007]

once I got a nice thin telephoney vocal sound by using 2 different mics close together, flipped out of phase and added to mono..

[Mike Caffrey]

Originally Posted by msmith

Good luck with that...

Moving a pre recorded track in "time" is not the same as repositioning the mic in the "space" of that room...

You win the prize for the shortest, most accurate comment in the thread.

[EvilE]

Originally Posted by bgrotto

My approach is to use Digi's Time Adjuster plugin to experiment with lining mics up. You can easily bypass it, so you can actually A/B whether or not it's doing anything good.

Generally speaking, I don't have great luck with lining stuff up; the sound I got was the one I want, phase-coherent or not! Often, it's the comb-filtering and such that gives a sound its 3-dimensionality.

That said, I'm in the middle of tracking a record for roots/americana musician Tony Furtado. If you're unfamiliar with Tony, he's a really amazing multi-instrumentalist; he'll rip on just about anything with strings and a fretboard. I set up a bunch of mics in a sort of "acoustic instrument station", and in addition to experimenting with different combinations for his many different instruments (and styles!), I fiddled (no pun intended) a bit with re-adjusting the timing on some of the mics. For some sources, it really worked wonders. It brought his Dobro to life, and saved the day on an acoustic guitar track.

So, yeah, it's definitely worth experimenting. I'd just advise against actually moving the regions themselves; using the sample delay plug gives you waaay more flexibility, and it's undoable.

some how i knew benny would post on this thread. hey dude, good advice. anyway, i like that idea a lot. otherwise, seems like you should try to move the mics until they all line up....or sound good together! do a couple little 10 second tests recorded in and check the phase in the edit window. once they work together, you can have like a dry or wet option, dark or bright option that work together or apart. think about the mix before you start tracking! also there's always the 3:1 ratio rule....that is usually pretty fool proof on stuff that isn't bass heavy.