Stereo Mic Techniques - Phase or Time Differences vs Volume

[James B]

I haven't quite been able to get my head around this.

If I record something (say a solo acoustic guitar) with a spaced pair is the stereo image formed by the difference in sound arrival time or differences in phase or the relative volumes at the two microphones?

What I really want to know is - in what cases should I be adjusting the channels to an equal amount of gain if I suspect the differences in volume were caused by a lack of attention when setting the levels and when should I leave it because that difference in volume is key to the stereo image? M/S? Blumlein? XY?

I don't think I've done a great job of explaining this...

[dsw]

Does the image move to the center as you adjust the levels?

[MoreSpaceEcho]

in what cases should I be adjusting the channels to an equal amount of gain if I suspect the differences in volume were caused by a lack of attention when setting the levels and when should I leave it because that difference in volume is key to the stereo image?

not to be a jerk, but...just listen to it.

[Dakota]

If I record something (say a solo acoustic guitar) with a spaced pair is the stereo image formed by the difference in sound arrival time or differences in phase or the relative volumes at the two microphones?

All 3 of those factors simultaneously contribute to the perception of stereo image. Which of the three are dominant factors and which are lesser factors in making an image sound stereo varies widely depending on the setup and application.

What I really want to know is - in what cases should I be adjusting the channels to an equal amount of gain if I suspect the differences in volume were caused by...

Gain staging each mic while recording isn't about that - it's about each mic pre gain staged to the good input level zone of your recorder, input levels not too hot or too weak. Assuming both mics are recorded to separate tracks, get the input levels similar and healthy for each mic. This will often require somewhat different gain settings on the two mic pres.

Then later, in the mix - that's when you'd make decisions about the proportionate level of each track.

[Dakota]

in what cases should I be adjusting the channels to an equal amount of gain if I suspect the differences in volume were caused by a lack of attention when setting the levels and when should I leave it because that difference in volume is key to the stereo image?

not to be a jerk, but...just listen to it.

+1

[James B]

I hear the "just listen to it" remark, but I'm pretty curious about how this this works and some of the different techniques sound too similar to tell to my ears. Or they sound different but I can't quite define how.

Thanks Dakota, I think you've cleared some things up for me there.

[vvv]

If you are ITB, I believe most DAW progs have a stereo/phase analysis - I use one in Cool Edit; I'm pretty sure the Roger Nichols site and mebbe Voxengo have freebies.

I mean, yeah, use yer ears, but the visual feedback can show you what yer hearing, and that can be useful.

Me, I'm a very visually-oriented person prob'ly, as I've said before, from years of viewing porn.

Gear porn, of course.

[Dakota]

You're very welcome, James B -

About educating one's listening, perhaps try this:

Take any mono track in your DAW, duplicate the same file to another track as well (with them both time aligned the same). Pan one hard right and one hard left. Set both at the same volume.

Listen. Sounds totally mono. Drop the level of the left one 3db. Listen. Sounds panned stereo a bit. Repeat at 6db drop, 12db drop, 24 db drop. Then return that track to the original volume and do the same routine with the right track. That isolates hearing how level contributes to perceived stereo image.

Then with them both back to the same volume level, slip the left one later in time by .1 millisecond. Then .2ms, .5ms, 1ms, 5ms, 10ms, 25ms, 50ms, 100ms, 250 ms. Then put it back and do the same routine with the right. That isolates hearing how arrival time contributes to perceived stereo image.

Then get one of those plug-ins that rotates phase. Keep the right track phase unchanged, and rotate the left track from 0 degrees in increments, stopping to listen at many points, particularly listening closely at 180 degrees (full polarity flip), and continuing on to 360 degrees from there. Return that to neutral and do the same deal with the right track. That isolates hearing how phase relationships contribute to perceived stereo image.

And mess with any combination of the three. Your ears will better and better hear how each factor contributes in a complex real-world recipe like stereo mic'ing an acoustic guitar.

And +1 vvv: also boot up some stereo and phase analysis plugs while mucking about with this kind of thing.

[dfuruta]

Most of the time, phase and time differences are the same, no? Unless you've got one mic behind the instrument and the other in front or unless the mic capsules behave weirdly with off-axis stuff, at least.

If you're recording with a pair of directional mics, the effect of the pickup pattern is, I believe, the main thing. With many stereo techniques (the ones you mentioned, certainly), the mics are fairly close together and so time & phase differences aren't going to be that big?remember that 1 foot apart is only 1ms difference. With XY, Blumlein or M/S, your mics will be much closer than that.

If you've got two coincident figure 8 mics pointing 90 degrees apart (Blumlein, say), each will pick up sound the strongest in the same place as the null of the other mic. So, when you pan it to L&R, you get roughly the same soundstage back. The same is true to a lesser extent with the other patterns except omni, since the nulls aren't as deep or as large. Remember, also, that all mics are directional at high enough frequencies.

With a spaced pair, which factor matters most depends on how close you are to the sound source, where the mics are pointed, and so on.

If you need to match the gain after the fact, can you take something that was exactly in the middle of the soundstage and make that equally loud in both speakers?

[dfuruta]

Then get one of those plug-ins that rotates phase. Keep the right track phase unchanged, and rotate the left track from 0 degrees in increments, stopping to listen at many points, particularly listening closely at 180 degrees (full polarity flip), and continuing on to 360 degrees from there. Return that to neutral and do the same deal with the right track. That isolates hearing how phase relationships contribute to perceived stereo image.

Keep in mind that these devices rotate the phase around some particular frequency, and so it's not necessarily straight-forward to understand what's happening. It's certainly very different from moving a mic...

[Dakota]

Then get one of those plug-ins that rotates phase. Keep the right track phase unchanged, and rotate the left track from 0 degrees in increments, stopping to listen at many points, particularly listening closely at 180 degrees (full polarity flip), and continuing on to 360 degrees from there. Return that to neutral and do the same deal with the right track. That isolates hearing how phase relationships contribute to perceived stereo image.

Keep in mind that these devices rotate the phase around some particular frequency, and so it's not necessarily straight-forward to understand what's happening. It's certainly very different from moving a mic...

Totally! I'd just proposed that so the OP could at least start directing his ears to raw phaseyness, somewhat teased apart from time&phase locked together. The nuances of the actual accurate physics of all this is a lot for a beginner to get a handle on all at once. - Or a pro as well... good ears and lots of practice, no substitute for that.

[RoyMatthews]

Not that everything everyone else posted isn't dead on but I think what James was talking about in the OP is how does one 'calibrate' a stereo pair of mics. He mentioned both spaced, AB, pairs of mics that depend on the time difference between capsules and coincident pairs such as XY, Blumlein, and M/S which base the image on just volume differences between sides.

It's fair to wonder if you have 2 microphones set up do you set the gains on the preamps to be the same and place the instrument(s) in that space or set the gain independently so the image is centered.

I do notice this most with acoustic guitar as James mentioned. The mic facing the hole of the guitar is going to be louder in a stereo array where the preamps are set the same.

I think the quick answer is use your ears and set the gains so the guitar is centered. Or place the guitar in relation to the mics so that it's centered.

I feel like setting gains differently so that the guitar is centered works fine except with M/S. I usually move the guitar then so it's where I want it. Sometimes I'll goose the L-R levels after I decoded the M-S.

If I have a spaced pair I'll consider placing the mics pretty much wherever as long as they are the same difference from where the pick hits the strings just like measuring the mics from the snare in a Glyns John drum setup.

Does this make any sense or did I just jump to conclusions? Was this what the first post was even about? I may be a little drunk.

[MoreSpaceEcho]

you did well.

i think for acoustic guitar, in XY the gains are gonna be about the same. spaced pair, if you have one back by the bridge and one circa where the neck meets the body (as is common), i'd think the bridge one is gonna need to "look" louder, as the neck one is gonna be brighter (and therefore sound louder at equal gain). blumlein i've never tried but i'd think it'd be similar to XY.

[Nick Sevilla]

is the stereo image formed by the difference in sound arrival time or differences in phase or the relative volumes at the two microphones?

Hi,

You are referencing three different things, which are inextricably interrelated.

The illusion that two microphones recording the same sound source create, is principally created by the time difference in the arrival of the same sound to both microphones. This is the most important CUE that our brain uses to know the location of a given sound source. The reason two microphones are the most efficient way to represent a sound source in a three dimensional space, is that they match the number of auditory organs(ears) that we possess. Tests have been done with more mics, and with less, and two is the most economical, efficient way of making this illusion work for most humans.

Phase is the Term used to describe a MOMENTARY difference within, in our case, two signals which have been combined together into one output. It is the Amplitude difference between two signals, in our case acoustic ones (not Absolute Phase, But Relative Phase). If you know how acoustic signals work, then you can understand that they add and subtract when added together inside the same space. This phasing occurs all the time, and varies according to the signal and the space it occupies, both of which work on each other. Phase is normally measured Instant by Instant, with no regard for time. Which is why it is only relevant to measure when you have an issue which is present ALL the time,in your recording.

As to Relative Volumes at the mic positions, again, these differences help our brain cue into the positional information of the sound source. Without any difference in volume between the two microphones, there cannot be any positional information.

Try this for a few minutes in your recording studio:

Put your mics in your preferred Stereo micing position, and place a static sound source in or near the position where your guitar will be. Calibrate the mic preamps to be within 0.1 dB of each other. You can use a sine wave tone through a small speaker for example. Make sure the sound source is the same distance from both mics, using a tape measure.

Then record your guitar as normal, without making any changes to your setup.

The result will be the correct Stereo recording of your guitar.

Cheers

[vvv]

FWIW, I was doing my bastardized psuedo M/S thing the other day with a M179 in cardoid pointing at the pan of a banjo, and the figure-8 M179 behind the cardoid mic.

It sounded like poo, but I didn't wanna re-do the banjo performance.

So I simply dropped the inverted figure-8 track and panned the original hard right with the cardoid panned about 10:00 left.

Did I get a accurate stereo representation - obviously not.

But in the mix of the kinda alt.cuntry-rawk tune, it was cool "stereo-enuff".

I guess my point is, eh, ...


... on top of my head.