Taperssection.com
Gear / Technical Help => Microphones & Setup => Topic started by: boltman on April 25, 2017, 08:54:05 PM

For the purposes of this discussion, I'm being totally reductive. Just want to get a basic sense here.
In midside, with an omni mid, one could expect the equivalent of crossed cards, with a 50:50 mix, I believe.
With a cardioid, it would approximate crossed hypers.
What would the general equivalent be with a supercardioid mid?

For the purposes of this discussion, I'm being totally reductive. Just want to get a basic sense here.
In midside, with an omni mid, one could expect the equivalent of crossed cards, with a 50:50 mix, I believe.
With a cardioid, it would approximate crossed hypers.
What would the general equivalent be with a supercardioid mid?
Picture something between hyper and figure 8

Sounds logical.

Keep in mind that forward directionality is tied in with the choice of Mid pattern. That is, the pattern of the Mid mic determines the maximum achievable forward directionality of the stereo configuration. But because that maximum forward directionality is achieved at a ratio of 100% Mid / 0% Side, it's equivalent to just the Mid in mono and not useful for stereo. As the proportion of Side is increased the forward directionality of the Mid mic is traded off for increased stereo width, until at a ratio of 0% Mid / 100% Side there is no remaining forward directionality influence by the Mid mic at all. You just have Side with opposite polarity to each speaker.
So using an omni Mid means no forward directionality and the sum of the combined patterns will always omnidirectional. With an omni Mid the virtual patterns will vary across the entire first order pattern range from omni through bidirectional, but the angle between the resulting virtual patterns is always 180 degrees apart regardless of what Mid:Side ratio is used. So when using an omni Mid and a 50:50 Mid/Side mix, the result is crossed cardioids, but with an angle between them of 180 degrees rather than a more typical stereo X/Y angle.
The inverse situation happen when using a bidirectional Mid. In that case the virtual patterns always remain fig8 at all Mid:Side ratios, and only the angle between the virtual 8's changes.
Using other Mid patterns lands somewhere between those extremes, with both virtual patterns and virtual angles varying with mix ratio, up to the maximum possible forward directivity as determined by the Mid pattern.
So, one way to determine what Mid pattern to use is to consider how much forward directionality you want given your recording position. If you need to minimize pickup from behind, use the most directional Mid you have available. If you want to minimize pickup from the sides, use a bidirectional mid along with a minimal amount of side in the mix ratio.

If you have an ios device, this is very very useful: https://itunes.apple.com/us/app/midsidemicvisualizer/id486145346?mt=8

Thanks guyshelps to clarify it for me. Thanks for the link, Noah.

For the purposes of this discussion, I'm being totally reductive. Just want to get a basic sense here.
In midside, with an omni mid, one could expect the equivalent of crossed cards, with a 50:50 mix, I believe.
With a cardioid, it would approximate crossed hypers.
What would the general equivalent be with a supercardioid mid?
With an omni mid you get backtoback cardioids pointing away from each other.
Read the Dooley and Streicher MS paper (https://soundlink.co.uk/docs/MS%20stereo%20%20dooley%20&%20streicher.pdf)  you will see all the diagrams in this. :coolguy:

Just keep in mind that similar to idealized polarpatterns plots and response graphs, the patterns shown in that paper are ideal/mathmatical/theoretical rather than realworld/actual/asachieved ones.
The two are closely related actually. As the section heading in that paper indicates, those pattern diagrams represent "MS Stereo: Concept and Theory". Also read the section titled "Practical Application Notes" at the end. Most everything talked about in that section is describing things which directly relate to how realworldachieved patterns do not behave quite like the simple and clean looking conceptualmathematical pattern diagrams suggest.
However, it gets the basic concept across nicely.

About the problem of talking about "idealized" patterns when the actual performance of so many microphones doesn't live up to them: That's not just empty words. Many microphones have a polar pattern that varies significantly at different frequencies. That makes for real problems with coincident (X/Y) recording, and to some extent with M/S as well although less so (as long as the figure8 is a real figure8 across the spectrum).
The worst offender without a doubt is largediaphragm, dualdiaphragm cardioids, which is about 90% of the condenser microphones used in recording studios. They tend to be cardioid only in the midrange, becoming wide cardioids at low frequencies and narrower (~supercardioidish) at high frequencies. Whenever people talk about "cardioids" without saying what kind of cardioids, in my opinion the talk is likely to steer a lot of people wrong as far as practical usage is concerned. A cardioid pattern that remains a real cardioid throughout the main part of the audio spectrum can only be obtained from a smalldiaphragm, singlediaphragm capsule. And even the best of those run into limitations at high frequencies.