SOUND ON SOUND

[blinddrew]

I had a lucky accident the other night when I was recording some backing vocals. I'd forgotten to mute the speakers, fortunately I generally monitor at a low level and I was deliberately going for a close-mic'd, heavy proximity effect so the signal to noise ratio was such that I got away with my amateur mistake.
But that got me thinking.

Say you have a standard pair of stereo monitors set up in front of you in a typical near-field position. You then put a mic in front of you to sing into such that it's pretty much forming the third point of the equilateral triangle.
Now, assume a mono signal is sent to those speakers, and let's also assume a nicely treated room with a stable phantom image in the centre of the speakers.

Would a cardiod mic give better rejection than a hypercardiod?

I.e. Does the microphone respond to the phantom image or the actual sound sources?

Most of my brain is saying the phantom image is made by the actual interaction of soundwaves in the air and therefore that is what the microphone will respond to.
But a niggling part of it is thinking that this seems a little oversimplified.

[Tim Gillett]

Any mic can only respond as per its polar pattern. So in this case the hypercardioid rejects better assuming its deepest null points align with the speaker positions. The cardioid is better with a single speaker directly behind it.

[Kwackman]

I remember an article or interview about this sort of thing years (maybe decades!) ago.
Where they used the speakers as you did with a mono signal, then swapped the wires on one of the speakers so it's polarity was reversed. Theory being there'd be a lot of cancellation at the mic.
I'm not sure if it worked, or even if I read it properly!
Dave Stewart & Annie Lennox maybe?
Anyone else with a better memory?

[Aled Hughes]

Any mic can only respond as per its polar pattern. So in this case the hypercardioid rejects better assuming its deepest null points align with the speaker positions. The cardioid is better with a single speaker directly behind it.

I don’t think you’ve understood his question. He is asking whether or not a microphone ‘hears’ a phantom central image.

My gut feeling is that a phantom centre image relies on two ears and a brain in between, so the single mic wouldn’t be aware of it. But I have nomidea really!

[blinddrew]

Any mic can only respond as per its polar pattern. So in this case the hypercardioid rejects better assuming its deepest null points align with the speaker positions. The cardioid is better with a single speaker directly behind it.

I don’t think you’ve understood his question. He is asking whether or not a microphone ‘hears’ a phantom central image.

My gut feeling is that a phantom centre image relies on two ears and a brain in between, so the single mic wouldn’t be aware of it. But I have nomidea really!

Yep, that's pretty much it. Is the phantom image a property of the ears and brain or the interference pattern of two independently produced waves?

[CS70]

I.e. Does the microphone respond to the phantom image or the actual sound sources?

Not sure if I fully follow you, but insofar I know we perceive position due differences in level and timing of the sound that gets to the two ears, the sonic "shadow" of the head and the angle with which the sound wave front hits our ears (which is why ears aren't flat).. stereo out of two fixed speakers deals mostly with the first two (and the third, well, it's always there and matters more for low freqs).

So no, a microphone doesn't hear a stereo image, but the mono summation of what arrives at its position.

An hypercardioid has a back lobe, hasn't it? In this case, yes I'd say it would pick up more sound from the speakers (or as u say, a cardioid would have better rejection)

[Hugh Robjohns]

Would a cardiod mic give better rejection than a hypercardiod?

No, the hypercardioid would give much better rejection than the cardioid -- assuming ycather align the nulls towards the speakers. (Actually, a supercardioid would be even better as its nulls are closer to the required angles for your scenario). This is a standard setup for floor wedges on stage under a hypercardioid vocal mic, after all!

I.e. Does the microphone respond to the phantom image or the actual sound sources?

The actual sound sources!

The phantom image is entirely a construct of our human sense of hearing being fooled by sounds arriving at both ears at the same time and amplitude. In real life, that can only happen if the sound is being emitted by a source on the median plane, so our brain invents a phantom image.

The sound waves from the two speakers will interact, of course, and produce a horizontal interference pattern, but there will be multiple peaks and troughs occuring at different angles for different frequencies -- not a single central lobe!

[CS70]

I remember an article or interview about this sort of thing years (maybe decades!) ago.
Where they used the speakers as you did with a mono signal, then swapped the wires on one of the speakers so it's polarity was reversed. Theory being there'd be a lot of cancellation at the mic.
I'm not sure if it worked, or even if I read it properly!
Dave Stewart & Annie Lennox maybe?
Anyone else with a better memory?

I did try - probably after reading the same article as you :) - and I still have the XLR phase inverter in the drawer somewhere.

Even if I spent lots of time carefully positioning the speakers and the mic, I wasn't at all impressed by the cancellation... but admittedly I tried in a poor room with no absorbers so it could be that in a better, less reflective room it works better.

Another variation is the "Beatles trick" where you have a figure 8 microphone and two people singing against each other, with the base on speakers at 90degrees.

That I've tried as well, and it works much better and it's easier... no complete "cancellation" as you can still hear the base a little but definitely usable.

[CS70]

assuming you can align the nulls with the speaker axes.

Right, good point!

[Hugh Robjohns]

Where they used the speakers as you did with a mono signal, then swapped the wires on one of the speakers so it's polarity was reversed. Theory being there'd be a lot of cancellation at the mic.
I'm not sure if it worked, or even if I read it properly!

Yes it works, and he's you remembered it correctly. It's a workable technique for recording vocals in the control room with live monitoring over the speakers... If the artist doesn't mind outof phase speakers... Makes me feel I'll!

[blinddrew]

The phantom image is entirely a construct of our human sense of hearing being fooled by sounds arriving at both ears at the same time and amplitude. In real life, that can only happen if the sound is being emitted by a source on the median plane, so our brain invents a phantom image.

Thank you, I wasn't sure if that phantom image was a human construct or a wave interference pattern. :thumbup:

In that case, going back to my original lucky break, I got even luckier than I thought. :)

[James Perrett]

Where they used the speakers as you did with a mono signal, then swapped the wires on one of the speakers so it's polarity was reversed. Theory being there'd be a lot of cancellation at the mic.
I'm not sure if it worked, or even if I read it properly!

An alternative to this would be to record two tracks - one with the vocalist singing as normal and the other with the vocalist in the same position but quiet. If you invert the polarity of the second track you should end up with just the difference between the two.

[Tim Gillett]

Where they used the speakers as you did with a mono signal, then swapped the wires on one of the speakers so it's polarity was reversed. Theory being there'd be a lot of cancellation at the mic.
I'm not sure if it worked, or even if I read it properly!

An alternative to this would be to record two tracks - one with the vocalist singing as normal and the other with the vocalist in the same position but quiet. If you invert the polarity of the second track you should end up with just the difference between the two.

The difference being the vocal itself? What would be the benefit then over the vocal track?

[The Elf]

Where they used the speakers as you did with a mono signal, then swapped the wires on one of the speakers so it's polarity was reversed. Theory being there'd be a lot of cancellation at the mic.
I'm not sure if it worked, or even if I read it properly!

An alternative to this would be to record two tracks - one with the vocalist singing as normal and the other with the vocalist in the same position but quiet. If you invert the polarity of the second track you should end up with just the difference between the two.

A trick I use is to do both of the above - reverse the polarity of one speaker (keep them equidistant from the mic position) AND record a no-vocal take and flip it. Between these techniques you can get something pretty clean.

[Tim Gillett]

Where they used the speakers as you did with a mono signal, then swapped the wires on one of the speakers so it's polarity was reversed. Theory being there'd be a lot of cancellation at the mic.
I'm not sure if it worked, or even if I read it properly!

An alternative to this would be to record two tracks - one with the vocalist singing as normal and the other with the vocalist in the same position but quiet. If you invert the polarity of the second track you should end up with just the difference between the two.

A trick I use is to do both of the above - reverse the polarity of one speaker (keep them equidistant from the mic position) AND record a no-vocal take and flip it. Between these techniques you can get something pretty clean.

I get it now. Use a hypercardioid mic and that would make three stages of attenuation.