R/E/P Community

R/E/P => Klaus Heyne's Mic Lab => Topic started by: klaus on March 12, 2019, 02:10:29 AM

Title: Better Blumlein Technique?
Post by: klaus on March 12, 2019, 02:10:29 AM
I received this as an email, but, while the sender did not wish to post in a forum, he gave me permission to do so:

I have configured a Blumlein array with four KM84 mics.
As you can imagine, a figure 8 can be created with 2 cardioid mics with their axis oriented at 180 degrees to one another, their diaphragms aligned on the same plane, stacked one above the other, and the phase of the rearward facing mic flipped 180 degrees from the forward facing mic. Two such figure 8 pairs of KM84 mics are oriented 90 degrees from each other with the flipped phase facing rearward on both figure 8 pairs.

I would be grateful for your comments, and kindly correct my understanding of this mic arrangement where I may be mistaken:

1. I found that there is no audible effect due to shadowing as a result of stacking the mics participating in each figure 8 arrangement vertically, and oriented 180 degrees to each other, as one might believe would occur. Double blind listening trials of stacked and non stacked mic positions are not distinguishable.

2. The major advantages of this Blumlein array, as compared to a Blumlein array created from two figure 8 mics, is the choice of mics participating in the Blumlein array now includes stock cardioid mics, and control of volume and pan of each of the four lobes of the Blumlein sound field.
The trade-off when gain is unbalanced in the figure 8 members is a reduction of the null in the figure 8 component proportional to the gain delta in each figure 8 component. This trade-off results in a more omni sound field as the affected lobes become more cardioid as the gain delta increases within each figure figure 8 component.

3. If the phase of the rearward and forward facing mics are configured to be 0 degrees from each other (all four mics in phase with each other), then one is left with two pairs of XY coincident arrays, each array facing 180 degrees from one another. The in-phase sound field of each array is wider to capture a wider angle of sound in-phase at the expense of rejecting more sound from the sides of the Blumlein sound field.

The author asked me to add this link to a Blumlein patent: https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=GB&NR=394325A&KC=A&FT=D&ND=1&date=19330614&DB=&locale=en_EP
Title: Re: Better Blumlein Technique?
Post by: afterlifestudios on March 23, 2019, 06:18:41 AM
Perhaps my understanding is not correct, but the few concerns I have about this are:

1.  Would the null at 90 degrees from each of the front/back cardioid pairs be as deep as the null at 90 degrees of a figure eight where the capsules are truly “back to back” with precise spacing. This proposed arrangement has them somewhat vrtically “staggered” so I’m not sure a convincing figure 8 would be achieved.  If this is the case I would expect the pick up patterns to “overlap” more with the other front/back pair and cause phase issues.

2. I find it difficult to believe that there is “no audible effect due to shadowing” when stacking km84’s the way I’m imagining it...  But besides “shadowing”, it would also seem impossible that the metal cylinder housing tubes of the km84 bodies would not create reflections that would combine with the unreflected sound to creat some kind of comb filtering.  But I have not measured it (or tried it), so I cannot difinitively say.

Cool to think about, though...
John


Title: Re: Better Blumlein Technique?
Post by: opacheco on April 01, 2019, 11:23:40 PM
That sound interesting set up!

Do exist a method o way in order to simulate the response or the resultant pattern of  an array like this conforming 4 Cardioid Mics for a Blumlein System with the control for each Cardioid??...I mean use a Geometric Angular Graphing in order to get the Directional Pattern in a cylindrical type plot.

What would be the angular equation for something like that mics arrangement?

Thanks

Title: Re: Better Blumlein Technique?
Post by: David Satz on April 02, 2019, 11:39:42 AM
A forward-facing cardioid is (1 + cos θ)/2, a backward-facing cardioid is (1 - cos θ)/2, and a figure-eight is simply the cosine function by itself.

If you subtract a backward-facing cardioid from a forward-facing one, the two half-"1"s cancel each other out, while the two half-cosines reinforce one another to become a whole cosine--a figure-8. That assumes that the microphones are well matched and that their directional patterns have a common center point.
Title: Re: Better Blumlein Technique?
Post by: afterlifestudios on April 06, 2019, 04:39:30 AM
... their directional patterns have a common center point.

That’s the part I can’t see being the case in the above suggested array... 
Title: Re: Better Blumlein Technique?
Post by: David Satz on April 06, 2019, 05:53:07 AM
Right. But conventional stereo recording takes only the horizontal plane into account. So if the two microphones' diaphragms are aligned one above the other, and precisely equidistant from the sound sources, then (as far as conventional stereo recording is concerned) the microphones are effectively "coincident"--just as with stereo microphones that have one capsule above the other.

Actually I share your doubts to some extent, and wish I knew more about the listening tests that didn't turn up any audible shadowing effects. I believe the sincerity of the report--but one would expect some wavelength-dependent problems in the top octave, depending on exactly how the microphones were arranged. It would be good to quantify those effects even though they might be small.

--best regards

P.S.: It would be interesting to take a stereo microphone that has vertically aligned, single-diaphragm capsules, such as a Schoeps CMTS 301 or 501, and measure its frequency and polar response when its capsules are set to cardioid and combined into a figure-8 as in this experiment. One could also do that for dual-diaphragm capsules as in various stereo microphones from Neumann or AKG.

P.P.S.: Another point of reference for this discussion is the Neumann KM 86 which had back-to-back, separate cardioid capsules that were combined for its "omni" and figure-8 settings. (In fact those two capsules were each the same as a KM 84's capsule--just with shorter pins--though as I recall, the reflectors behind/between the capsules in the KM 86 were shaped more like a convex lens than the hemispherical-shaped reflector of a KM 84.) The dual capsules and the spacing between them helped account for the microphone's strong low-frequency response in the figure-8 mode.

But the down side was irregular polar response at high frequencies. This made the KM 86 sound rather shrill (as I can unfortunately attest) if it was used for Blumlein stereo recording, since the front of a Blumlein pair is at a ±45° angle to each of the capsules--and that is just where the KM 86 had a substantial high-frequency rise followed by a considerable fall (see graph below; I've drawn a circle at 315°, where the microphone's ±45° polar response is shown for the higher frequencies, and attached another graph where I've run a curve through data points for 4 kHz, 8 kHz and 12.5 kHz to infer the microphone's ±45° frequency response from its 1 kHz frequency response as well as I can).

Neumann recommended that the KM 86 be used only at miking distances greater than one meter--but I think that they meant that for the patterns that used both capsules, since there's no great problem with the pattern in the cardioid setting.
Title: Re: Better Blumlein Technique?
Post by: afterlifestudios on April 06, 2019, 07:17:29 AM

P.S.: It would be interesting to take a stereo microphone that has vertically aligned, single-diaphragm capsules, such as a Schoeps CMTS 301 or 501, and measure its frequency and polar response when its capsules are set to cardioid and combined into a figure-8 as in this experiment. One could also do that for dual-diaphragm capsules as in various stereo microphones from Neumann or AKG.


Thanks for the great thoughts and info.  That’s an excellent idea.  I have a sm69 I could try that with.
Title: Re: Better Blumlein Technique?
Post by: klaus on April 06, 2019, 03:38:19 PM
Quote
Another point of reference for this discussion is the Neumann KM 86 which had back-to-back, separate cardioid capsules that were combined for its "omni" and figure-8 settings. (In fact those two capsules were each the same as a KM 84's capsule--just with shorter pins--though as I recall, the reflectors behind/between the capsules in the KM 86 were shaped more like a convex lens than the hemispherical-shaped reflector of a KM 84.) The dual capsules and the spacing between them helped account for the microphone's strong low-frequency response in the figure-8 mode.

Correct about the shapes of the reflectors.
But there is another issue not yet mentioned in the discussion about the KM86's frequency response and suitability for Blumlein: it sports 4 screens in front of the capsule! Three of them are in the outer head basket, and one more in front of the diaphragm platelet.

I speculate that the predominance of lower mids and lows in that mic (or the relatively soft representation of high frequencies and its overall cotton-mouth dynamic behavior) is a direct result of these acoustic impediments, even it that is not clearly reflected in the graphs.
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 06, 2019, 07:56:16 PM
My name is Jeff. I am the person messing around with Mr. Blumlein’s array. Klaus was kind enough to post my Blumlein ideas in this forum. I didn’t think anyone would be interested. I can say that it works for me, and I’m pleased to try to answer any questions if I’m able to.

To hopefully address Dave’s concern, here are the two alternatives I’ve investigated for vertically stacking the 4 cartioid Blumlein array. It is easier to visualize by drawing a schematic:

1st Arrangement:
The two forward facing mics  are immediately one above one other. And the two rearward facing mics are immediately above one another. This arrangement necessitates the  mics participating in each of the two figure 8 components to be separated from one another by one of the microphones, a little less than an inch in the case of KM84 mics. i.e., this arrangement is optimized for coincident phase at the expense of a less symmetrical Blumlein sound field. I would like to put in perspective  Dave’s concern vis-a-vis the Blumlein sound field created from four cartioid mics may be compromised, I would point out that the Blumlein composed of the 4 cartioid mics  I’ve described would be no less symmetrical than a traditional Blumlein array sound field composed from ribbon mics or multi pattern condenser mics that place the ribbons or diaphragms of the two mics a minimum of approximately 2 inches apart. This is often  necessitated by the minimum distance between the  ribbon or diaphragm to the cage of each microphone. This distance multiplied by 2 is the distance between the transducers of both mics in the traditional Blumlein array. Although the ribbon mics do not suffer from the asymmetry of two diagrams to create the figure 8 polar pattern, ribbon mics and side address multi pattern mics often necessitate greater separation between their transducers as compared to small diaphragm condenser cardioid mics of approximately  1 inch diameter.  The small effect it has on the sound field asymmetry and  phase relation consequences are regarded as negligible in these traditional Blumlein arrays. So to is the case with a Blumlein array composed from 4 cartioid mics with similar dimensions as the KM84 mic.. - jeff

2nd Arrangement:
The two mics participating in each figure 8 component are immediately one above the other. This arrangement is optimized for a more symmetrical Blumlein sound field at the expense of a less coincident relationship between the forward facing mics and the coincident relationship of the rearward facing mics respectively.
For what it’s worth, I usually prefer the 1st alternative because the degree of coincidence in the 2nd alternative is audible but small in the top octave or so. A separation of 21 mm diameter mics by about a wavelength of 1 inch will effect frequencies above approximately 13KHz, although diffraction of frequencies above this are to some degree mitigated by diffraction at the diaphragm depending on the diameter of the diaphragm.

Regarding shadowing, although I have never experienced any audible shadowing effect using the KM84s, if anyone does hear shadowing with their mics, one could always use detachable miniature cardioid capsules or try a mic with a smaller foot print.
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 07, 2019, 08:35:56 AM
I just noticed that the last sentence (...) is gibberish.

Did you fix it? Looks coherent to me (aside of mixing up 'effects' with 'affects'...).
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 07, 2019, 11:05:56 AM
An analysis of the Blumlein’s spaciousness is well done at this link:
http://taperssection.com/index.php?topic=110861.0
Title: Re: Better Blumlein Technique?
Post by: klaus on April 08, 2019, 12:33:27 PM
Here is a photo of Jeff's XY arrangement.
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 08, 2019, 01:24:25 PM
Here is the more cogent explanation than the malformed sentence I mistakenly posted previously. I took a photo of an XY array that Klaus was kind enough to attach for me due to my shortcomings. I often use this XY array with a mutual angle if 110 degrees and an additional mic in the center of the sound field Ito fill in the hole that results from the wide mutual angle. The advantage is having this wide sound field captured with an additional on axis mic. Similarly, this is considered by many as an advantage of MS.

I posted it here to visualize 3 of the 4 mics in the 4 cartioid  Blumlein. One can see from the photo that two mics immediately one above the other will provide a much better mono compatibility than if the left and right mics were facing either forward or rearward. The diameter of the KM84 is 21mm, and the diameter of the diaphragm I’m estimating is approximately 19mm (not a published spec that I’m aware of).

Assuming this is a close estimate, a frequency of 17,895Hz has a wavelength of 19mm, equal to the diameter of the KM84 diaphragm. This represents the minimum wavelength that could be accurately captured. Frequencies with wavelengths above 1/2 this wavelength and less than one wavelength are captured less accurately than sound with equal to, or greater wavelengths due to diffraction at the diaphragm.

Accordingly, two mics that include the center mic and either the left mic or the right mic to capture the forward or rearward sound field have excellent mono compatibility in the upper octave of human hearing, as compared to captuing the forward or rearward sound field with the left and right mics separated by the middle mic. Sound arriving at both the left and right mics are not necessarily coincident. This is the reason I prefer alternative 1 that I described earlier for stacking the 4 cartioid Blumlein.
Title: Re: Better Blumlein Technique?
Post by: David Satz on April 08, 2019, 02:06:18 PM
I was trying to figure out how that photo could be a Blumlein pair! Thanks for explaining that it isn't.

While I was struggling to interpret the image, though, it occurred to me that if the two outer cardioids were 90° apart, and if the microphone in the middle were omnidirectional, with frequency response and sensitivity like that of the cardioids, one could matrix the three signals into a Blumlein-ish result: Take the signal from either cardioid, invert the signal from the omni, pad it down ca. 6 dB and sum the two together; do the same for the other cardioid as well, and there you have it. (The narrowing of the omni's pickup pattern in the top two octaves would unfortunately make that an inferior solution.)

All that being said, despite near-ideal behavior in certain respects, the Blumlein approach is highly problematic in practice. It has a rather narrow stereophonic pickup angle, which often requires placing the microphones at considerable distance from the sound sources in order to encompass them in the stereo image. But at the same time, being based on figure-8 microphones, it is as sensitive to rear-incident sound as front-incident sound. At the miking distances that it often requires, you pick up far too much reverberation relative to direct sound; the recording sounds "washed out". I haven't often found Blumlein usable in live concert recording, though for session recording you can sometimes arrange your performers to fit the mike technique, if that's how you want to work.

--best regards
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 08, 2019, 03:55:41 PM
The narrow stereophonic pickup angle of the traditional Blumlein can be altered with the 4 cartioid Blumlein by increasing the gain delta between the front and rear facing microphones, which would otherwise not be alterable with the traditional Blumlein. The sound field becomes more omni-like as the delta increases. Similar otherwise unalterable attributes are adjustable, e.g., one could compensate for the array placement located too close or too far from the reverberation radius, etc. - Jeff
Title: Re: Better Blumlein Technique?
Post by: David Satz on April 08, 2019, 06:40:57 PM
That's true, but when you change that front/back ratio, you are simply setting a different pickup pattern for the two "virtual microphones" whose signals you actually record. It's just like using the remote pattern control on a stereo microphone such as a Neumann SM 69[fet] or any of several past AKG models; their capsules were likewise made from pairs of front- and rear-facing cardioids, and the remote pattern control varied the relative sensitivities of those cardioids.

Since the angle between the main axes of the microphones remains fixed, along with their distance from the sound sources, the amount of reverberance in the recording and the width of the stereo image both depend on that same pattern setting at the same time. It's not a free choice--and in my experience, for any given physical microphone setup, only one setting of the pattern control sounds right at all.

It's the same predicament as in conventional M/S recording--again, a largely equivalent approach achieved by means of a matrix. There's usually just one "sweet(-ish) spot" setting, quickly found by ear; it's certainly not the free, wide-ranging aesthetic choice that might be imagined.

To have any greater freedom you would need Ambisonic ("Soundfield") recording or double M/S. And even with those approaches, if your mikes are too close or too far away or too high up or not high up enough, they're still too close or too far away or too high up or not high up enough.

--best regards
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 08, 2019, 10:01:15 PM
Altering the gain ratio of the front facing and rear facing mics is not tantamount to relocating the array further forward or backward in the recording space. However altering the gain ratio provides a means to either increase or decrease the relative ratio of reverberance as might be the case when the rear facing mics have more reverberant to direct sound than the front facing mics.

The compromises that ensue by altering the gain ratio is immediately apparent as the gain ratio is changed, not unlike mixing any other audio parameter. It is often possible for me to be very pleased with the trade offs, just as it is with any alteration of a recording in post. The 4 cardioid Blumlein provides a means that may be useful that would not be available with a traditional Blumlein array.
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 09, 2019, 09:27:00 AM
Referring to the link at:
http://www.linkwitzlab.com/index.htm
Linkwitz Lab has an excellent analysis of rendering a recording for stereo playback. The SRA, stereo recording angle is discussed, and it is pointed out an SRA of approximately 70 or 75 degrees will render stereo playback for an idealized equal lateral triangular relationship between the listener and the left and right speakers.

Consistent with this, Dave correctly intimates there is a fixed relationship between the placement of the mic array to the sound source and the mutual angle between the mics to obtain the best sounding result that he readily recognizes. Additionally, I would submit that the polar pattern of the mics and their interaction to shape the sound field, i.e., the polar pattern of the array, which also determines  the acceptance angle of the array, and therefore the distance of the array to the performance to achieve the desired SRA.

Unique to the Blumlein’s among coincident and near coincident arrays is the minimization of monophonic reverberation in the sound field.  Dave also correctly points out that the fixed nature of the relationships he mentioned make the Blumlein difficult to obtain a good result unless the sound source is placed as necessary to conform to the Blumlein array  narrow acceptance angle.

And I agree this is not practical or even desirable, as many recording venues with great sounding acoustic characteristics would be excluded. The acceptance angle could be made wider with the 4 cartioid Blumlein by adjusting the gain ratio of the forward facing to rearward facing mics. If it were adjusted so the rearward facing mics were completely muted, then a forward facing XY coincident array with a wider acceptance angle than the Blumlein remains.

To the extent that the acceptance angle can be widened enough to capture the whole performance and allow for some contribution by the rearward facing mics, the less the reverberant aspect of the environment will be rendered monophonically.
Title: Re: Better Blumlein Technique?
Post by: Jeffrey on April 09, 2019, 05:02:04 PM
Please refer to an SRA calculator found at:
http://www.sengpielaudio.com/Visualization-Blumlein-E.htm
Please observe what happens when the calculator is toggled using the radio buttons between bi-directional microphones showing the Blumlein array and the cartioid micrphones which will show an XY array,with both arrays defaulting to a mutual angle of 90 degrees. The resultant SRA will be shown in shadow when you toggle between these 2 microphone polar patterns and arrays. It is easy to see that as if a hybrid of the figure 8 and cartioid mic were available for selection the SRA would increase as the polar pattern shifts from bi-polar to cartioid. This is tantamount to modulating the relative gain of the forward and rearward facing microphones in a 4 cartioid Blumlein array. - Jeff