Become a Site Supporter and Never see Ads again!

Author Topic: The World's Finest Blumlein Array  (Read 2238 times)

0 Members and 1 Guest are viewing this topic.

Offline Len Moskowitz (Core Sound)

  • Site Supporter
  • Trade Count: (3)
  • Taperssection Member
  • *
  • Posts: 336
    • Core Sound
Re: The World's Finest Blumlein Array
« Reply #45 on: August 06, 2020, 10:50:19 PM »
what would be a recommended calibration interval for octomic, and at what cost?

We recommend re-calibrating every three years. Cost is $250.
Len Moskowitz
Core Sound
www.core-sound.com

Offline kuba e

  • Site Supporter
  • Trade Count: (1)
  • Taperssection Member
  • *
  • Posts: 342
  • Gender: Male
Re: The World's Finest Blumlein Array
« Reply #46 on: August 07, 2020, 07:42:19 AM »
It literally has nothing to do with the pattern of the microphone array's capsules. The capsules could just as well be omnis as the directional capsules we use in OctoMic.

Len, I'm sorry we don't have the same view. I'm out of the internet now. But I'm sure everything will be explained. If OctoMic can be assembled with omni capsules, then you use a different principle than I was thinking. I am not an expert on this, I hope this will help someone clarify.
« Last Edit: August 07, 2020, 07:50:58 AM by kuba e »

Offline EmRR

  • Trade Count: (3)
  • Taperssection Member
  • ***
  • Posts: 552
    • ElectroMagnetic Radiation Recorders
Re: The World's Finest Blumlein Array
« Reply #47 on: August 07, 2020, 12:03:58 PM »

In the overall combined pattern sense mentioned in the posts above, Blumlein presents a horizontal toroid (donut/fat-inertube) shape, and crossed hypers a toroid which is ballooned out on one side more than the other, sensitive across the full horizontal plane but with increased bias toward the front.  We've discussed in other threads how blumlein has reduced sensitivity to sound arriving from both directly above and below, particularly how it might produce a somewhat more even-sounding audience portrayal by reducing sensitivity somewhat for nearby audience noise immediately surrounding the recording position when the mics are on a tall stand.

Quote
[..] They reasoned an omni combined with a hypercardioid looking up and another looking down, both hypers polarity reversed, would create that pattern.  The OctoMic is ideally suited for that job.  You could likewise make it point to a lesser part of the horizontal spectrum. 

I don't think a monopolar toroid is possible to construct that way.  Thinking through it..  Ignoring the omni microphone for the moment, if done perfectly, two summed hypers of opposite polarity pointing in opposite directions would produces a figure 8 response.  The omni component of the hypers is subtractive and cancels out, while the bi-directional component is additive, producing just a figure 8 oriented in the same direction as the positive polarity hyper.  Two summed hypers of the same polarity pointing in opposite directions would produce an omnidirectional response.  The omni components then sum, and the figure 8 components cancel out.   That remaining omni component from both hypers would have inverted polarity in comparison to the omni microphone, so the result of the full combination of all three microphone signals is theoretically a complete null.

Being 2nd order, OctoMic should actually be able to provide a toroidal monopole response, possibly by using the advanced plugins Len mentions.

Sure, Blumlein makes a bumpy toroid of sorts, but in stereo.

The hypers in my example are in polarity relative to each other, opposite the omni.  At the right level and placement, the hypers should not make omni, they should leave a toroidal hole at 90/270º when mixed with an omni.   Tweaking between hyper and super is probably the ticket.   Second order patterns from an OctoMic should do a better job of it. 

Experiments to do!
Mics: DPA 4060 w/MPS 6030 PSU/DAD6001/DAD4099, Neumann KM 131, KMR 81i, Oktava MK 012, Sennheiser MKH 105, MKH 20, MKH 30, MKH 40, MKH 800 TWIN, lots of other studio appropriate choices
Recorders: Zoom F8n, Sony MZ-R50, portable MOTU based multitrack DAW for client work

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (15)
  • Needs to get out more...
  • *****
  • Posts: 13938
  • Gender: Male
  • Gunther Theile nailed it!
Re: The World's Finest Blumlein Array
« Reply #48 on: August 07, 2020, 01:41:58 PM »
The hypers in my example are in polarity relative to each other, opposite the omni.  At the right level and placement, the hypers should not make omni, they should leave a toroidal hole at 90/270º when mixed with an omni.

Unfortunately no, I had to think about it yesterday to get it myself (it's the second example in my post quoted above).

Two coincident microphones of any pattern with the same polarity but opposite orientation will sum to omni (except two figure-8's which sum to null).

Regardless of the pattern used, the omni compnents are of the same polarity and add together, while the bi-directional components are of opposite polarity and subtract from each other, cancelling out.  The symmetry of identical patterns facing opposite directions means all the bidirectional component cancels out regardless of the proportion of bidirectional in the nominal patterns of the mics used.

[edited to add- using two hypercardioids rather than two cardioids just means you get a less sensitive omni out of the result]

Summing the resulting omni with a polarity inverted omni 3rd microphone results in null (after adjusting levels to match).



It may help to imagine taking it on step further by varying the pattern of one of the two opposed directional microphones.  What then results is some 1st order directional pattern other than omni, yet it always remains a familiar 1st order pattern.  Flipping polarity on one of the mics flips the orientation of the resulting 1st order pattern.
« Last Edit: August 07, 2020, 01:48:53 PM by Gutbucket »
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (15)
  • Needs to get out more...
  • *****
  • Posts: 13938
  • Gender: Male
  • Gunther Theile nailed it!
Re: The World's Finest Blumlein Array
« Reply #49 on: August 07, 2020, 01:53:25 PM »
Here's another way of thinking about it-  Imagine setting up a "native" ambisonic b-format array made using three fig-8s and one omni.  You can use that to derive a 1st order pattern pointing in any direction, but you can't derive a higher order pattern like a toroid shape from it.   The three mic example of two hypers and and one omni are just a subset of that.
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Len Moskowitz (Core Sound)

  • Site Supporter
  • Trade Count: (3)
  • Taperssection Member
  • *
  • Posts: 336
    • Core Sound
Re: The World's Finest Blumlein Array
« Reply #50 on: August 07, 2020, 02:00:08 PM »
Here's another way of thinking about it-  Imagine setting up a "native" ambisonic b-format array made using three fig-8s and one omni.  You can use that to derive a 1st order pattern pointing in any direction, but you can't derive a higher order pattern like a toroid shape from it.   The three mic example of two hypers and and one omni are just a subset of that.

And because of the separation in space of the mono mics, the results will be quite poor.

If we then go to a true ambisonic microphone (instead of a native B-format made with mono mics), the first-order patterns you'll get from a first-order ambisonic microphones are much, much less stable and precise than the ones you'll get from a well-calibrated second-order ambisonic microphones like OctoMic.
« Last Edit: August 09, 2020, 08:34:34 AM by Len Moskowitz (Core Sound) »
Len Moskowitz
Core Sound
www.core-sound.com

Offline voltronic

  • Trade Count: (34)
  • Needs to get out more...
  • *****
  • Posts: 3153
Re: The World's Finest Blumlein Array
« Reply #51 on: August 07, 2020, 02:05:08 PM »
Two coincident microphones of any pattern with the same polarity but opposite orientation will sum to omni (except two figure-8's which sum to null).

Regardless of the pattern used, the omni compnents are of the same polarity and add together, while the bi-directional components are of opposite polarity and subtract from each other, cancelling out.  The symmetry of identical patterns facing opposite directions means all the bidirectional component cancels out regardless of the proportion of bidirectional in the nominal patterns of the mics used.

[edited to add- using two hypercardioids rather than two cardioids just means you get a less sensitive omni out of the result]

I would think hypercardioids would have other problems besides being less sensitive in this arrangement.  Hypers have that pronounced rear lobe, which will overlap the front lobe of the other coincident mic oriented 180 deg opposite and vice versa.  A sound source directly on-axis to the array would be picked up strongly by the front lobe of the forward-facing mic (duh), but also somewhat well from the rear lobe of the rear-facing mic.  Would that overlap not cause the bass to be overly bloated for sources on-axis to the front of the array, since that is where that rear lobe is most sensitive?
DPA 4061 | Line Audio CM3 | Naiant X-Q | AT 853  //  Naiant PFAs | Shure FP24  //  Zoom F6 | Sony PCM-M10
MOTU M2 // KRK Rokit RP5 // Sennheiser HD 650

Team Line Audio

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (15)
  • Needs to get out more...
  • *****
  • Posts: 13938
  • Gender: Male
  • Gunther Theile nailed it!
Re: The World's Finest Blumlein Array
« Reply #52 on: August 07, 2020, 03:25:51 PM »
What I outline above is the basic phenomena, without any of the messy complications of actual implementation.  It helps to ignore real world implementation complexities when seeking to understand the basis of how things work at a fundamental level. Things get very complected quick when moving on to actual implementation.

It's important to recognize that difference, especially in this type of discussion where both aspects are being discussed in the same thread.

Real world implementations can't work at all unless the basic underlying phenomena is correct, but a bad real world implementation can screw up a fundamentally correct underlying basis.


Len, what you say is not incorrect, but you are describing real-world implementation complications.  Voltronic I think you are mixing up the two.  The omni resulting from summed hypercards in my theoretically perfect example has less sensitivity because hypercardioids don't have much omni component in them to begin with.  They are mostly bi-directional component which is cancelled out.  Not sure what you are getting at with the bloated bass stuff, but since the proposed scheme won't work fundamentally, its pointless to speculate on the real-world problems of an attempted implementation.

musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (15)
  • Needs to get out more...
  • *****
  • Posts: 13938
  • Gender: Male
  • Gunther Theile nailed it!
Re: The World's Finest Blumlein Array
« Reply #53 on: August 07, 2020, 04:11:31 PM »
Going back to this-
(Thanks to this thread, I understood now how ambisonic works - it is the same as Mid/Side manipulation of coincident pair of mics, but this is in 3d space and with the possibility of more combinations)

In terms of the fundamentals that's actually a pretty good basis for understanding basic 1st order ambisonics. At it's core ambisonics can be thought of as a further extension of basic Mid/Side.. or more specifically, of sum/difference processing of the basic omni and bidirectional components which in varying combinations form the basis of all 1st order sensitivity patterns.

I don't mean to differ with Len in his reply to that post, and I don't think we actually disagree on that point as I believe was replying with the specifics of implementation in mind rather than the fundamentals.

To clarify, not simply the manipulation of a pair of "microphones" but of three (or four) at minimum.  Specifically two crossed figure-8s and one omni for a 2-dimensional single plane 1st order ambisonics (horizontal only), and an additional figure-8 required to extend that to the 3rd dimension (height).  Those "microphones" correspond to the 4 channels of ambisonic B-format (W = omni, X = front facing 8, Y = left facing 8, Z= up facing 8 ). 

I put "microphones" in quotes above because I'm describing the fundamentals more than actual implementation.  Although it can be done using actual microphones of those patterns (doing so is referred to as a "native" B-format recording array), actual ambisonic microphones such those from Core Sound's do not actually use three figure-8s and an omni.  Instead they use a different, more advantageous arrangement of microphones which significantly helps improve a number of implementation issues, in combination with some special sauce processing.  The actual output of the microphone elements of such an array prior to the application of the "special sauce" is referred to as A-format, and is specific to each microphone design.  In typical use, that A format output is what gets recorded to the machine, yet is unusable until the "special sauce" is applied resulting in "universal" B-Format (actual output is either B-format or direct output of some form of mono, stereo or multichannel output converted internally from B-format).  Only at that point does it cross from the real-world-implementation-specific realm into the realm of fundamental ambisonic B-format signals which can be be manipulated with sum/difference techniques in the same way as Mid/Side, only with far greater flexibility.  More advanced tools and plugins can do stuff beyond simple sum/difference processing, but sum/difference manipulation of the fundamental omni and bi-directional components is the underlying basis of ambisonics.
« Last Edit: August 07, 2020, 06:15:11 PM by Gutbucket »
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Len Moskowitz (Core Sound)

  • Site Supporter
  • Trade Count: (3)
  • Taperssection Member
  • *
  • Posts: 336
    • Core Sound
Re: The World's Finest Blumlein Array
« Reply #54 on: August 07, 2020, 05:46:01 PM »
I want to again emphasize that the process of going from A-format (the raw capsule audio) to B-format (what is called "encoding") has almost nothing to do with the directivity of the microphone array capsules.

There's no way that you can combine - for example - cardioid capsule directivities in the simple way Gutbucket has been discussing to get the directivity of an ambisonic second-order channels. Or omni capsules. Or any other mono mic patterns.

Yet, those capsules can be used in ambisonic arrays.

---

That said, you can combine virtual mics decoded from B-format in the way Gutbucket describes. The first-order virtual mics you'll get from OctoMic (the continuum from omni to sub-cardioid to cardioid to super-cardioid to hyper-cardioid to figure-8 dipole) are among the best (if not the best) in the world. (See the polar patterns on https://www.core-sound.com/OctoMic/2.php .) So the combination of those patterns are similarly excellent.

And of course, with a good second-order ambisonic microphone you get many more useful patterns than first-order mics (ambisonic or mono) can ever offer.

For Blumlein though, to get the crossed dipoles, we don't need to decode B-format to two dipoles, because the first-order B-format X and Y channels are already azimuth-plane dipoles. And being B-format, they don't have proximity effects.
« Last Edit: August 07, 2020, 05:58:23 PM by Len Moskowitz (Core Sound) »
Len Moskowitz
Core Sound
www.core-sound.com

Offline EmRR

  • Trade Count: (3)
  • Taperssection Member
  • ***
  • Posts: 552
    • ElectroMagnetic Radiation Recorders
Re: The World's Finest Blumlein Array
« Reply #55 on: August 07, 2020, 07:27:17 PM »
The hypers in my example are in polarity relative to each other, opposite the omni.  At the right level and placement, the hypers should not make omni, they should leave a toroidal hole at 90/270º when mixed with an omni.

Unfortunately no,


Well, I just tried it with an MKH 800 Twin and a 4060, it does work, crudely.    I put a metronome on the end of 18" of string and ran a circle slowly around the mic pair to maintain distance and level as best as possible. 

Best set supercardioid in my test.  When you are at 90/270º relative to the hypers, you get maximum output from the mono array of 3. It may not be a toroid, but it's at least a greatly flattened sphere, less output up and down, more side to side.    It's not purely with respect to frequency, but that has value in practice as well.  The opposite facing hypers or supers by themselves definitely have less output at 90/270º, it does not make an omni in the same way that the two capsules combine to omni: that has no drop off in level.

As visual example, go to

http://www.sengpielaudio.com/Visualization-AB60-E.htm

set it XY, supercardioid, point them 90/270º.   Sure looks like it's on the way to the void of a toroid. 


Should be able to do a refined version with second order patterns. 
« Last Edit: August 07, 2020, 07:32:33 PM by EmRR »
Mics: DPA 4060 w/MPS 6030 PSU/DAD6001/DAD4099, Neumann KM 131, KMR 81i, Oktava MK 012, Sennheiser MKH 105, MKH 20, MKH 30, MKH 40, MKH 800 TWIN, lots of other studio appropriate choices
Recorders: Zoom F8n, Sony MZ-R50, portable MOTU based multitrack DAW for client work

Offline Len Moskowitz (Core Sound)

  • Site Supporter
  • Trade Count: (3)
  • Taperssection Member
  • *
  • Posts: 336
    • Core Sound
Re: The World's Finest Blumlein Array
« Reply #56 on: August 08, 2020, 10:15:57 PM »
If you're looking for a toroid in the azimuth plane, combine the V and U second-order B-format channels.
« Last Edit: August 08, 2020, 10:30:10 PM by Len Moskowitz (Core Sound) »
Len Moskowitz
Core Sound
www.core-sound.com

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (15)
  • Needs to get out more...
  • *****
  • Posts: 13938
  • Gender: Male
  • Gunther Theile nailed it!
Re: The World's Finest Blumlein Array
« Reply #57 on: August 10, 2020, 09:54:41 AM »
As visual example, go to

http://www.sengpielaudio.com/Visualization-AB60-E.htm

set it XY, supercardioid, point them 90/270º.   Sure looks like it's on the way to the void of a toroid.

Keep in mind that Sengpiel visualizises the stereo interaction, not the electrically summed behavior of the two channels.

So you had the MKH800 twin set to output two opposed supercardioids of the same polarity, then summed those resulting outputs together?
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Len Moskowitz (Core Sound)

  • Site Supporter
  • Trade Count: (3)
  • Taperssection Member
  • *
  • Posts: 336
    • Core Sound
Re: The World's Finest Blumlein Array
« Reply #58 on: August 10, 2020, 10:19:45 AM »
Why would you bother with summing first-order patterns when you'll get much, much better results in both the spatial and frequency domains with the simple sum of the V and U second-order B-format channels?

The summing is done incredibly simply: route the two  channels to a single recording output.

If you want to hear it on one of our B-format recordings, pick one and I'll decode & record it for you.
« Last Edit: August 10, 2020, 10:41:48 AM by Len Moskowitz (Core Sound) »
Len Moskowitz
Core Sound
www.core-sound.com

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (15)
  • Needs to get out more...
  • *****
  • Posts: 13938
  • Gender: Male
  • Gunther Theile nailed it!
Re: The World's Finest Blumlein Array
« Reply #59 on: August 10, 2020, 11:43:25 AM »
Because he is using the microphones he has available, and the primary intent at this point was to test the Gearslutz hypothesis about the ability of producing such a pattern, more than testing presumed usefulness of it.

Thanks for your offer to make such a decode of the B-format recording.  I do intend to play around with those recordings at some point with all this in mind, just need to make some time to do so.  Are you sure a toroid shaped pattern is the result of summing channels V and U?  Given the interaction of their inverted polarity lobes in the sum, I can't see how that works if the 2nd order channel assignments the same as those shown here:



Which instead seems to imply a sum of (R and -W)
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

 

RSS | Mobile
Page created in 0.056 seconds with 56 queries.
© 2002-2020 Taperssection.com
Powered by SMF