assorted questions

[tek0010]

Capsule Distance?

What is the effect of the distance between capsules when setting up the Output Formsts in MIR? With real microphones in the physical world, adding distance between capsules causes the capsules to hear things at slightly different times from one another, according to the speed of sound moving through the air. Does MIR simulate this aspect of distance between the capsules? Or if capsule distance does something else, what exactly does it do?

Bass Response?

I have the following basic understanding about real mics in the physical world: As you bring a directional mic closer to the sound source, the bass response gets amplified ("proximity effect"). Does MIR simulate this "proximity effect" for directional mics?

I have also read that real-world directional mics are worse at capturing the bass, compared with omni mics. I'm not sure if this is the same thing as the proximity effect -- as you move a directional mic further from the sound source, the bass gets attenuated -- or whether the bass-problem with directional mics is something separate from the proximity effect. In any case, I notice in MIR's description of the M-S Omni configuration:

"Due to the use of the omni, this is the first choice for arrangements that need lots of energy in the low frequency range."

Basically, why do you recommend this for arrangements that need lots of energy in the low frequency range? Other choices would result in bass-loss? Would the bass-loss depend on the mic distance from the Instrument, as in the proximity effect, or is it from some separate phenomenon? Please elaborate?

Off-Axis Coloration?

I read that directional mics color the incoming sound depending on the angle between the incoming sound and the mic axis. Does MIR simulate this?

Thank you

[Dietz]

Welcome tek0010.

thanks for these interesting questions.

It is crucial to understand that all capsules MIR Pro displays for its Main and Secondary Microphone arrays are purely virtual. They are all derived from the original impulse response recordings which use a meta-format called Ambisonics. (I tried to give a quick overview of the technical background in an addendum to MIR Pro's manual called “Think MIR!”. It is available from your VSL User Area: -> http://eu.vsl.co.at/downloader.aspx?FileID=7629 )

Ambisonics recordings are gathered by means of a coincident microphone array. These recodings can be decoded virtually to pretty much every single- or multi-microphone setup (as long as it is a coincident setup, too). This means that we can freely define the number of capsules, their polar pattern, their volume, the directions they are facing into and the angles they are using in all three dimensions.

While there actually is no “distance” between coincident capsules by definition, MIR Pro uses some clever de-correlation algorithms to allow for more spatial enveloping by the reverb tail. (… please google the pros and cons of coincident microphones in comparison to spaced microphone arrays – this would be way beyond the scope of a little forum reply. 8-) …). The amount of this de-correlation applied to one of the virtual capsules in MIR Pro is shown as “distance” to make the process easier to understand.

Re: Proximity Effect: There is no proximity effect other than the one that happened naturally in the Venue when capturing the impulse responses (which is highly desirable from an audio engineer's point-of-view!).

 

Re: Off-axis colouration: Luckily there is none in a virtual microphone capsule. The polar patterns decoded from the Ambisonics signal can be considered “perfect” mathematically.

 

Re: Omni patterns: It is a well-known phenomenon that lower frequencies have less directivity than higher ones, thus the bass-components of a signal will be audible from more angles in a hall then the treble. An omni capsule, which picks up sound from all three dimension, will have a more distinct bass response than capsules with narrow polar patterns, therefore.

 

@Another User said:

Basically, why do you recommend this for arrangements that need lots of energy in the low frequency range? Other choices would result in bass-loss? Would the bass-loss depend on the mic distance from the Instrument, as in the proximity effect, or is it from some separate phenomenon? Please elaborate?

 

There are many factors influencing the recorded frequency range of a signal in a hall, and the same is true for MIR Pro. Like in the case of “real” recordings I would aim for a well-balanced overall frequency spectrum first, and boost the bass range of a few selected instruments only, otherwise your mix will get “muddy” and “boomy” quickly (… that's why spot-microphones are so important). An M/S-setup with an Omni capsule as M-channel is a good starting point in any case, but other variables like the instruments' positions, the chosen Main Microphone positions and all the musical aspects like arrangement (!) and performance are equally important.

 

HTH,

[tek0010]

Thanks for your explanations. 

@Dietz said:

While there actually is no “distance” between coincident capsules by definition, MIR Pro uses some clever de-correlation algorithms to allow for more spatial enveloping by the reverb tail. (… please google the pros and cons of coincident microphones in comparison to spaced microphone arrays – this would be way beyond the scope of a little forum reply. 8-) …). The amount of this de-correlation applied to one of the virtual capsules in MIR Pro is shown as “distance” to make the process easier to understand.

So increasing MIR's virtual-capsule distance is like increasing the ratio of side-channel to mid-channel? Is that a good way for me to imagine it (since I'm already familiary with mid/side processing)?

[Dietz]

No, M/S processing is something different. The "Distance" parameter increases the artifcial de-correlation of the signal derived from the virtual capsule. The actual volume relation between M and S doesn't change. That's why it works for stereo as well as for surround-setups or a single virtual microphone capsule.

BTW: The de-correlation doesn't affect the first few milliseconds of the signal, to keep the all-important positioning cues intact. And of course the dry signal components of an instrument (a.k.a. "spot microphone") aren't affected either, although they undergo a complete Ambisonics encoding/decoding process otherwise, to match the relative position of the original multi impulse responses exactly.

HTH,

[tek0010]

Most of your answers above make sense to me, but I'm confused by your description of the Distance parameter (for a mic capsule).

@Another User said:

please google the pros and cons of coincident microphones in comparison to spaced microphone arrays – this would be way beyond the scope of a little forum reply. 8-)

I actually have researched this already, and feel confident in my understanding of real-world coincident mics v. real-world spaced-arrays. MIR’s "spaced" arrays are something very different from real-world spaced-arrays, though, and it’s MIR’s "spaced" arrays which I’m working to understand here.

I understand well what 2 spaced omnis would do in the real world (cf. Format-B above). But in MIR — Format-B yields mysterious audio results, despite the simplicity of its design. Before I try to design my own Output Formats, I hope to at least understand what something so simple as Format-B does?

[tek0010]

@Dietz said:

Ambisonics.... This means that we can freely define the number of capsules, their polar pattern, their volume, the directions they are facing into and the angles they are using in all three dimensions.

By "volume" here, you mean their size in 3-dimensional space (as opposed to decibels 'volume')? How does the spatial volume of a capsule affect the audio it produces? This is counterintuitive to me.

[Dietz]

It's really volume - as in "louder" or "quieter". The graphics are just there to give you a quick idea of what to expect from a chosen setup.