Double miking acoustic; placement and mix techniques

[Wyllys]

Bob...

Greeat info.

At first glance the "reversed ORTF" looks to me to be an "expanded x/y" setup.
I was going to say expanded coincident pair but that makes no sense.

Now I have to look up the Jellyfish. Looks like a neat device...right up there with my flux capacitors.

Cheers.

[ukejon]

Phenomenal answer, Bob. I have an ORTF bar so reverse setup will be easy. Back to you soon with a track.

[rick-slo]

Originally Posted by Wyllys

Derek, et al...
Perfect or absolute phase correlation at all frequencies is an impossibility. The only scenario where phase coherency can be absolutely perfect is a single sine wave. The best you can do is to time align the zero point of all tracks. All your mics will then be at a virtually equal distance from the source.

All complex wave forms can be represented as a sum of various sine waves (Fourier analysis). For perfect correlation phase correlation of the right and left tracks in a stereo recording all the sine waves would have to be in phase. The right and left tracks being identical (except perhaps in amplitude (volume)) you would have mono. Thus my comment earlier.

Now if you mean, for example, that when you record two different signals (waveforms), as in your typical stereo recording, you can't phase align all the frequencies post recording, then that is very true. My point was that you would not want to do that anyway (even if it were possible). If one is adjusting time delay on one track for "better" phase alignment with another track it would pertain to specific frequencies (done for improvement to one's ears hopefully). Interestingly for your typical mike setup for recording guitar in stereo where the mikes are fairly close to equidistant from the guitar low frequencies are for all practical purposes nearly in phase (for example the wavelength of 100 hertz being about 10 feet) (discounting also possible deleterious effects from room reflections (walls, ceiling)). Track time delay adjustment amounts I generally use when playing around with phase correlations are audibly applying to quite a bit higher frequencies .

[ukejon]

By the way, Bob, I had to look it up but now you've convinced me to strive for microphonic euphony!

[rick-slo]

Regarding the vertical ORTF I did some similar which came out well. I put the mikes about eight inches apart and about fourteen inches out from the guitar soundhole. Both mikes faced the guitar but one mike I angled up about sixty degrees and the other mike down about sixty degrees.


Usually I do not care for mikes this close together and this close to the soundhole usually has too much proximity effect but it worked well with this specific mike orientation.

Here is a recordings. The only post recording addition is a small amount of reverb.

http://dcoombsguitar.com/Guitar%20Music/ASimpleThingAM.mp3

'
'
'

[Wyllys]

Originally Posted by rick-slo

All complex wave forms can be represented as a sum of various sine waves (Fourier analysis). For perfect correlation phase correlation of the right and left tracks in a stereo recording all the sine waves would have to be in phase. The right and left tracks being identical (except perhaps in amplitude (volume)) you would have mono. Thus my comment earlier.

Now if you mean, for example, that when you record two different signals (waveforms), as in your typical stereo recording, you can't phase align all the frequencies post recording, then that is very true. My point was that you would not want to do that anyway (even if it were possible). If one is adjusting time delay on one track for "better" phase alignment with another track it would pertain to specific frequencies (done for improvement to one's ears hopefully). Interestingly for your typical mike setup for recording guitar in stereo where the mikes are fairly close to equidistant from the guitar low frequencies are for all practical purposes nearly in phase (for example the wavelength of 100 hertz being about 10 feet) (discounting also possible deleterious effects from room reflections (walls, ceiling)). Track time delay adjustment amounts I generally use when playing around with phase correlations are audibly applying to quite a bit higher frequencies .

I think perhaps stereo and mono carry connotations which might prejudice things a bit at worst or cloud the issues at hand a bit at best. It's just sound in the end.

The most valuable things I've taken from study and practice are understandings of polarityhase and mono-compatability. Understand these two issues and everything else falls nicely into place. The fascinating thing is what transducers to use and how to array and position them. A mm here or there, a degree here or there in the angle, maximizing desired frequency response when combining elements in an array...there's enough there for several lifetimes of work.

Fortunately, good engineers all over the world have been working this out pretty much since the advent of radio broadcast if not before. What I find when I finally grok the import of someones work is the elegant simplicity behind the complexities.

And if it sounds good we win.

[upsidedown]

This thread has officially gone over my head; in terms of understanding both its content, and the gear involved! I'm scratching my head with the two sticks I use for making fire.

But it's enjoyable.

[rick-slo]

Originally Posted by upsidedown

This thread has officially gone over my head; in terms of understanding both its content, and the gear involved! I'm scratching my head with the two sticks I use for making fire.

But it's enjoyable.

Fire makes light.

[sdelsolray]

I have found that the off axis response of a mic correlates with how well that mic will record a source when placed in non-direct positions, such as XY, ORTF, over the shoulder, Blumlein and M/S (for the side mic). In these positions, the front of the mic is not pointing at the guitar but is instead is pointing somewhere away from it. If the off axis response of the mic is cheesy/inconsistent/sporadic then the recording will suffer. Excellent off axis response is usually only found in certain high end mics. Average or poor off axis response is common among most other mics.

This is perhaps why so many folks prefer direct aiming of mics at the source, which most of the spaced pair and AB placements do (as do a J-Disk and Faulkner array).

I'm am leaving aside the desire of many to create an enhanced stereo recording (i.e., exaggerating the L/R spread of what is essentially, at a normal listener's distance when listening live, a single source), which is a different topic altogether, although it can be influenced by the off axis issue.

I am also leaving aside room issues, which can be a dangerous omission.

[upsidedown]

Originally Posted by rick-slo

Fire makes light.

Exactly right!

[ukejon]

Originally Posted by sdelsolray

I have found that the off axis response of a mic correlates with how well that mic will record a source when placed in non-direct positions, such as XY, ORTF, over the shoulder, Blumlein and M/S (for the side mic). In these positions, the front of the mic is not pointing at the guitar but is instead is pointing somewhere away from it. If the off axis response of the mic is cheesy/inconsistent/sporadic then the recording will suffer. Excellent off axis response is usually only found in certain high end mics. Average or poor off axis response is common among most other mics.

This is perhaps why so many folks prefer direct aiming of mics at the source, which most of the spaced pair and AB placements do (as do a J-Disk and Faulkner array).

I'm am leaving aside the desire of many to create an enhanced stereo recording (i.e., exaggerating the L/R spread of what is essentially, at a normal listener's distance when listening live, a single source), which is a different topic altogether, although it can be influenced by the off axis issue.

I am also leaving aside room issues, which can be a dangerous omission.

This issue has puzzled me as well. The Massenberg setup is moderately direct if close to the guitar but not if set back a bit. How well do mics (in my case cardioid mics) hear from an angle? Polarity pattern images suggest they do hear somewhat well from reasonable angles, right?

[Bob Womack]

Originally Posted by ukejon

This issue has puzzled me as well. The Massenberg setup is moderately direct if close to the guitar but not if set back a bit. How well do mics (in my case cardioid mics) hear from an angle? Polarity pattern images suggest they do hear somewhat well from reasonable angles, right?

This is entirely based upon the design and quality of the microphone. Here is where you find the real differences between good mics and not-so-good mics. Some exhibit reasonably flat response as you move off-axis and others don't. All those little lines on a good response chart really tell you something. For instance, here is the chart on an AKG C451B:



Note the two frequency response lines on the left and the split, two-spectra polar response curves on the right. On the left we see the two response curves, on-axis and 180' off-axis. From this we can see that at 180' from absolutely on-axis, there is a smooth -28db reduction in response between approximately 600 and 3000hz.. There is a gentle curve up from there towards a treble peak at 16khz. with a small hump at 5khz. On the bass side we see a more humpy curve up with the hump at 300hz., a trough, and then a tilt up towards 125hz. (if the roll-off isn't engaged). On the right we see the two polar plots to show the bass and treble spectra, bass being the left half. Notice that the polar plot for all frequencies is extremely consistent right out to about 100'. Overall gain of the mic is down by about 5db as you reach 90' and by 10db at about 115'.

Using the split in the polar chart to look at low and high frequencies, we begin to see the relative lift in bass (left side of polar chart) and treble (right side) smoothly develop from about 115' on to 180' while the overall gain drops. If we were to set a pair of these in a 90' coincident pattern at say 14" covering a three-foot-long guitar, the angles of incidence that are of merit will range from 0' to about 90'. If you interpolate the above slope between the on-axis and off-axis response you will get the impression that the frequency response through the various angles of incidence to 90' will be consistent and smooth because, in this case, the AKG C451B has a single-line plot of lobes of response all the way around to that point. In fact, the same can be said of its use in the inverted ORTF because the roughly 120' angles of incidence introduced in a 14" placement fall within the area where the mic exhibits reasonably flat response. Thus, if you like a mic with a presence peak, the AKG C451B proves to be a good candidate for use in patterns that feature off-axis pickup of this type.

However, you can get all kind of crazy response patterns off special design or lower-quality mics. The funny thing is that there are uses for all of these patterns. For instance, a mic that rolled off the high frequencies as you moved off-axis might be better to poke into a drum kit to pick up a snare on-axis while rejecting the adjacent hi-hat. When shopping, always try to look for a comprehensive a/b gain vs. frequency line curve and multi-frequency polar chart and work out the response you can expect from the mic's off-axis lobes.

You can verify some of this with a mic by speaking into it and rotating it while monitoring on headphones. I remember the mic preview room at the now-defunct Mars Music chain. That place was heaven for a recording engineer because they had all their good mics set up with headphones so you could try them out for yourself. To quote inspector Clouseau, "Not anymore..."



Bob

[rick-slo]

Like Bob said in his last post, there may be uses for frequency dependent variation in off axis response. It gives a broader choice of a mike tonal palette.
Depending on how directional a mike is you may get some out of phase response between two mikes when angling them around, but probably not much
recording an acoustic guitar and if noticeable at all then only when panning towards center.

[ukejon]

Bob, in theory would the Shure KSM 137s be a good choice for the reverse ORTF?

[Bob Womack]

Originally Posted by ukejon

Bob, in theory would the Shure KSM 137s be a good choice for the reverse ORTF?

If you like the sound of the mic on-axis, its polar plots (HERE) show a little on-axis peak at 10khz that seems to go away as you work around the sides of the pattern, starting right at 20'. The upper mids (around 6400hz.) drop at little as well.

Bob