Great...but just to frug with your mind, here's an alternate Onno spacing, one he came to later in his life, and likely one you'll admire even less: see post #18

RIP Onno Scholtze

As a bonus, you get one last TF video...I thought I'd lost the link for this one...but just re-located it again >>> go to around 17 mins in for some 66.5 cm exposure YouTube

With this sort of determination on site, pre-concert or pre-recording session, you're going to be making your spacing decisions on headphones or portable-enough monitor speakers in a (very likely) untreated general purpose room or office or green room...unless you're lucky enough to be in a facility which accommodates regular recordings, and has thus created a dedicated control room ?

So you're making decisions about relative phase integrity, stereo image width, bass extension etc on a set of compromised transducers, in a relatively unknown or hostile room. Maybe you even have a goniometer running on your DAW, or patched into a spare mixer channel ?

You can see where this is going...you're going to be relying on your knowledge of the venue, the size and scale of the ensemble (and the material to be played) and probably also upon: 1) your little notebook which you religiously carry with you to all such events...which contains details of mic type and spacing, and measurements/comments pertaining to prior successful (or otherwise) recording events there; 2) safety measures such as 4 mics on a single bar; 3) 'golden ratios' or 67 cms or other fallbacks which have had more global and general application for yourself and others in the past (ie they've been used on commercial CD releases, for example !)

Whatever data translates successfully to generating a satisfying stereo image back at base, and informs your mic placement practice at the venue, is valuable information...in what's often a compromised monitoring situation at the source of it all.

Yes, that is an interesting quote:

I have not come across it and I have never tried two omni's in the middle, but it is an interesting thought. I have used 3 meter wide AB, but always in combination with one omni or two non-omni arrays in the middle.

I will try this out in the future. Thanks for pointing this out.

BTW, the use of two omni's in a kind of Ortf-ish way as such is not new to me. It was one of the earliest surprises bestowed on me by the late Dick van Schuppen, and it made me aware that in the world of recording everything outside the common and accepted practice is worth to be checked out by oneself...

As regard to the 67 cm spacing. I am aware that there are some good folks recommending it, but in my experience one’s conception of what is best, will vary over one’s lifetime. There used to be years when I happily used 60 to 100 cm, but when I hear those recordings back, I am more and more concerned about the inherent lack of cohesion in the spatial image. I now find it too varying and to my ears it brings about too much disorder, unless it gets tamed by a second array in the middle. This is entirely dependent on taste, and taste seems to drift in one’s lifetime. Each engineer has to judge these things by himself. No use to follow others without trying out yourself.

Fortunately you have numerous historical antecedents to fall back upon , or ignore, as well as others you can invent for yourself...plus the oncoming possibilities of the more affordable Ambisonic-like offerings from Rode and Sennheiser, which can be used in stereo mode also.

You might be favouring, from your comments about 'disorder in the middle'..or is it a hole in the middle ...something like a Decca Tree, or a RCA/Mercury curtain of 3, or a Telarc centre ORTF or Blumlein flanked by outriggers...there's a plethora of precedents to embrace or reject, whatever you fancy.

If you study the Faulkner videos cited above, there's a consistency of methodology and reason/purpose behind his approaches which contradicts your suggestion that he simply adheres slavishly to previously tried and true (or solely empirical) methods...perhaps his physics degree helped him there also ?

His 4 mic array has both pragmatic value and theoretical/demonstrable flaws...simultaneously !! I'm sure he'd be the first to concede that, but it works for him and others across many contexts, hence the popularity.

Tony is not someone to slavishly follow dogma. On the contrary. The 67 cm spacing will be to his own liking, just as it was to my liking in the past. These listening preferences can not only vary from person to person, but also from ones period in life. Personally I have not such a high regard for so called golden standards. I do not believe in one ideal figure like 67 for AB, or 18 cm for Ortf-ish setups. I let my ears be the judge, not those of others.

To clarify my view of 67 cm or shorter spacing; it is not the hole in the middle that bothers me at these spacings. It is more the feeling that with a smaller spacing there seems to be a real continuous space around you (particularly on headphones). With larger spacing or angling out, that continuity seems to fall apart to some extent. I also avoid spacings under 40 cm because that brings other unpleasantness to the sound. In reverberant churches I sometimes put a Jecklin disc in between, but I would never space the mics closer than 40 cm, which was the original idea of Jecklin. These are all very personal choices, and there's no universal good or wrong.

I am glad that you understood my point that mixed arrays should not be seen as the Holy Grail of recording. They are useful to overcome difficult situations, and to mend problems in post. If you have time to move your main array around in a good sounding hall or church, then always strive to just use one of the two arrays for 85% or more of your sound.

I find 2 pairs on a bar really comes into its own, and TF confirms this, when the alternative is either filling the stage with spot mics for the further back instruments (ie just zoom in with your centre directional pair for more focus) OR when the performers move around on stage and it's impossible to adjust the stand to accomodate during the show.

Particularly if say choral singers rearrange themselves for more depth/ fewer rows of width for a couple of pieces...again use more cardioid in the blend to reach deeper into the ensemble, if needed to retain focus. If done too frequently within a performance you can end up with constantly-shifting 'focal length'....but that may be preferable to those same pieces being too washy if the mic choice was set for AB Omni as set and forget....

Maybe that's why ORTF and later added reverb is usually touted as the safest 'one size fits all' solution, as it's hardly ever out of focus (nor universally great either) !

Don't forget that the typically recommended AB width for a Decca tail pair on a concert grand is about 8"/20cm.....it's contended that wider than this the image loses coherence and starts sounding like '2 mics working separately', which Decca carefully avoided.

Be sure to study the Williams 'stereophonic zoom' PDF linked above....you may have developed a preference for a particular SRA and his calculations can help you shoot for consistency there (also Eberhard Sengpiel's calculation widget...similar approach), not forgetting the free mobile phone apps also: Neumann's "Recording Tools" and another one called "Stereo Mic Tools" (can't recall the developer) which are great aids on location, as they use the phone's camera to provide you with real-time angle/width measurements which removes guesswork !

what puzzles me regarding the discussion of 4-mic-arrays is the repeated claim that they would have more 'reach': this simply ain't true, no mic system has more 'reach' than any other, not even shotgun mics. some mics do have higher sensitivity than others and more directional mics have more damping towards the rear (and/or additional directions) but that's it - 'reach' is bad terminology or a simplification which (obviously) can lead to wrong understanding...

spacing mainly determines how far down in the fr some phase cancellation shall occur (besides the risk of the 'hole in the middle' - which can easily get filled up by using any kind of l/c/r system rather than a widely spaced a/b) which imo is as much a matter of taste as the basic choice of a mic system - i say taste 'cause i don't think all variables can get merged into a formula.


[and while we're at it: i mostly don't like any spaced technique much as they do yield a less accurate stereo image - somewhat different story if used for multichannel/surround reproduction though (where my beloved coinicident/directional mics sound rather boring, especially the rear channels. also, they cannot picture any movement in the soundfield convincingly).

i mention this to illustrate that i'm not here do downplay aesthetic qualities of any spaced technique or to hype functional aspects of coincident techniques; whatever one is going to use, there is alway a price to pay...

...but if using both at the same time, imo one does NOT get best of both worlds!]

You can indeed get a bit more "reach" by using cardioids instead of omni's, but I wouldn't say that it defies the need for some spots on woodwind and/or percussion instruments. That balance most often be problematic, unless you would move the main array further into the orchestra, but that causes other problems...

i cannot agree on the term 'reach', not even when using (highly) directional mics which i mostly do anyhow...

i do agree that if the mains cannot picture what one wants to get, there is no way around spots (well, i haven't tried a dozen of arrayed shotguns and applied sophisticated gain shading amongst the array?!).

i don't consider spot mics to be problematic to get integrated (unless one is mixing straight to 2track) but they certainly do need more gear/tools/time/work.

Are you on a sponsorship retainer from SRG SSR to mention directional mic pairs at every possible opportunity...I hope it is a worthwhile payment/loyalty scheme for you ?

The term 'reach' only has validity when 2 microphones (or 2 pairs of mics) of different directional characteristics are used at the same time on the same musical subject.
Let's say the omni pair (non-optimally located) is being used for the pickup of an orchestra or wind ensemble or a large choir ...etc
Let's say the omni pair is largely ok for the front to middle of the orchestral/choral depth...but that the rear rows of singers or the woodwinds or the wind orchestra percussion is sounding too cloudy and indistinct.

Obviously, bringing the ORTF or near concident pair (on the same bar) up in the overall omni mix a little has the net effect of 'apparently' increasing the focus and volume and perceptibility of those distant elements, to match the omni pair which defines most of the sound.

So 'reach' is used in a very qualified and contextual way...as the purposeful introduction of a different mic type to achieve a specific aim (and maybe only only on occasions, when those more distant players or singers are performing quietly, or otherwise 'lost in the mix') It only works as a descriptive term when 2 mic types are being used simultaneously...and one pair is failing to 'reach into' the ensemble as much as the other is able to...

As mentioned several times previously...it's a rescue remedy, not a substitute for informed, careful mic placement (ideally of a single pair).....but sometimes, that's not possible, due to logistical or administrative constraints (or the video guy telling you where you can't put your mic stand, 20 mins before the show begins !)

i'm neither affiliated with srf (which is the current name of the swiss national radio and tv broadcaster) nor do they have a bias on directional mics: i assume the majority of their mics being used by the most of their engineers for recording of classical music are omnis.

i occasionally get to work for/with them and either use what they are proving or double mic if i feel i cannot live with another engineer's choice - guess what: i then might put my directional/coincident main pair on the same bar as their spaced omnis - wouldn't call it an array though and would never use both pairs at the same time, not even to a very small degree (well, maybe to blur the stereo image and feed the omnis into a efx device).

___

imo the term 'reach' is just not very well chosen, accurate or appropriate, regardless of setting/setup...

studer,

you're bringing up a(nother) interesting topic: as mentioned previously, i'm not getting sponsored by any company or 'endorse' any product and i'm a freelancer/independant/self employed.

but to get to the point: i'm sometimes getting seriously embarrassed by the fact the national broadcasters do offer their service for free and provide opportunities (offering 'reach' :-) i cannot... - while i can compete on the technical level (using studer etc.), i cannot go onair and don't want to offer my work for free (only to do a job)!

dunno about other countries/areas: this does happen in the german speaking area a lot and on rare occasions, i've even seen multiple broadcaster parking their trucks outside festivals/concert halls with engineers fighting to get access to the venue; things usually get settled elsewhere though (and most any engineers i get accross seem to be both very professional and easy to work with).

also, i wish that some posters here would declare their position regarding institutions, distributors, manufacturers, sponsorships etc. in their profile...

I definitely don’t work for the Russians. No Collusion!

well i do (as much as i do for the americans) sometimes....

...not much in their favour though: i get to travel to the eastern ukraine to survey the minsk agreement on behalf of the osce and findings are often quite clear (and contradict what gets written elsewhere...)

a rather serious topic but i fear we're getting off track [and mods for good reason will not like us going any further] - so:

sorry to the op, i appreciate the discussion!

"what puzzles me regarding the discussion of 4-mic-arrays is the repeated claim that they would have more 'reach': this simply ain't true, no mic system has more 'reach' than any other, not even shotgun mics. some mics do have higher sensitivity than others and more directional mics have more damping towards the rear (and/or additional directions) but that's it - 'reach' is bad terminology or a simplification which (obviously) can lead to wrong understanding..."

We probably need some better alternative terminology if you don't like the word "reach". After all, the dictionary definitions of the word "reach" do not include what I think we are talking about. Stereo or surround microphone phased arrays of most if not all varieties do, in my opinion, deliver more "reach" both subjectively and scientifically. What I understand to be happening is that our ears are working hard to differentiate direct coherent wanted sounds from more randomised indirect incoherent sounds like room honks and excessive reverberation. The more clues and cues we can give our ears, the easier job our brains have pulling musical sounds out of reverberation and reflections. With stereo, we are making the job easier if our microphone system includes useful inter-channel timing information. If you ever have to deal with radio antennae or radar, there is a similar quest to pull coherent wanted information out of incoherent unwanted clutter and noise using timing and phase cues.

I enjoyed the kind thoughts about my late highly valued friend and colleague Onno Scholtze. Onno was a great guy and a great engineer, committed to making honest recordings with integrity. We knew each other very well and I worked with him on projects including using his 4-way 4006 array. Onno was constructive in the design of the Schoeps 2H capsule which is a fine mic..

"If you ever have to deal with radio antennae or radar, there is a similar quest to pull coherent wanted information out of incoherent unwanted clutter and noise using timing and phase cues"

i'm not saying that the mic arrays as discussed here cannot show some (minor) effect and i'm in no way trying to challenge your authority, it's just that for my taste, i cannot hear much (if any) benefit.

[i do have to deal with antennas a bit: working for the osce/blue helmets, i'm fixing them in war zones - in my experience, it takes more than just a four antennas (plus phase shifters) to get a significant steering of the beam.
can't say whether/to which degree this holds true in our case (mics to obtain 'reach') and if so, if it wouldn't be more feasible to use more mics, of the same design and delays - maybe you did try different arrays? would be interested in hearing about any experiments, even if they were not or not much of a success!]

Beamforming is a big topic and definitely in the same zone as phased arrays of microphones. I have tried all sorts of arrays here. Vertical configurations can give you forward gain as well. We recorded some Bach cantatas once using eight KM84's mounted in a way which looked like an old anti-aircraft gun. It looked a bit daft and did not really work wholly convincingly. I think it would be possible now to use DSP to process arrays looking for coherence signatures.

There is an interesting clever Polish surround microphone called Zylia which probably does this sort of DSP thing. I bought one of the first release but have not got on too well with it. The Windows drivers have taken a while to be fine-tuned, and for me working on location I cannot really get on with a mic which is connected via a USB cable. The cable-run in the Royal Albert Hall is typically 85 metres and there is no easy way to sort this out for a USB mic. Dante network connection would possibly have been a better solution in retrospect.

wow! would love to see a pic and hear results (but of course can fully understand if you cannot or do not want to show any details).

No pictures nor audio of the KM84 folly, sorry. We recorded it as an experimental 'B' setup on a backup deck, but I have no idea where the material would be any more, it was a long time ago. We seldom took any session stills at that time.

Vertical arrays are interesting because the direct sound will not vary a huge amount with similar delays dependent on path-length, but the ambient field will show useful deviation. I have always wanted to try putting out two identical arrays one physically a foot or so directly in front of the other. We could delay the front mics to be in time-alignment with the ones behind on direct sound from a designated forward focus-point, and the combination would have forward gain that way as well.

In the end, the results which convince me most sonically and musically are the simple ones. As soon as you have microphones all over the place, timing goes all over the place and creates a confusing mush. For me timing is an important issue. When you think about how your hearing works, timing has to be important when our personal sound perceptors are our ears which are close together spatially and the brain is going to be scouring for differential timing and phase cues to interpret what is going on.

Hi Tony, thanks for joining in again. It's always nice to have someone thinking out of the box.

What is your own idea about angling the outer Omni mics vs keeping them in parallel? Did you find the 67cm by calculations or by listening tests?

Have you ever tried Blumlein at the centre of such an array?

Sorry for all these questions, but instead of speculation it is better to hear it straight from you...

Hi Lucas_G,

Thanks for your kind note.

I angle the outer omni outwards usually. Most real-world omnis tend to be directional at hf so you might as well take advantage of it. Mics like M50s are omni at lf but intentionally increasingly directional at higher frequencies.

I have tried using Blumlein fig8's in the middle and it worked fine though it did not focus in on the rear of the soundstage so well. We have a Soundfield ST450ii and I might try it in the middle sometime.

I did not like Blumlein as a center pair for my array as it gave a slight “hole in the middle” effect. XY cardioids worked okay, but ORTF or NOS cardioids are my preferred center pair. Inre the AB omni pair, I often point them outwards or upwards, depending on the situation, and I find myself using closer spacing now than I used to.

Friday I will record orchestra with a Soundfield and 65cm AB.
I will add a hall pair and some spots. Hope it'll work.
I once tried 100 cm AB and Soundfield in the middle and that was actually quite ok.

Do you ever use Michael Williams Stereophonic Zoom to determine the spacing of stereo pairs? 65mm may be an 'average' omni spacing but, if you believe Williams, the spacing should be adjusted according to the SRA of the pair on a case-by-case basis.

according to the Sengpiel SRA calculator, I should use a 53 cm AB, thanks for reminding me.

Lucas, how are you processing the Soundfield mic in the middle of this array? I have a Josephson c700s that I've been using for ambisonics with 360 video, and I'm interested in trying it in an array w/ my pair of c617set in AB. Have you experimented with Harpex or the Blue Ripple Harpex Upsampler to move into 3rd order, then derive ORTF or other patterns from it after that? There are so many good options for ambisonic processing right now after the emergence of ambisonics for 360 video.

Hi Ryan,

I try to let one set be predominant, and I always prefer the omni's (DPA 4006a or MKH20) to be the dominant set. In such a case I usually end up with Soundfield being at least 6 dB less loud than the AB. I experiment in Harpex with Blumlein or NOS. I most often use Blumlein but with the polar pattern a little bit less figure eight and the angle around 100 degrees.

When the acoustics or placing prohibits the AB being the dominant set, I usually end up with the Soundfield in Blumlein (angle usually a bit wider than 90 degrees) and I add the omni's bass by using a HP filter to get more and better bottom end.

I also use 40 to cm 50 cm for AB width in this scenario. I stopped using 67 cm some time ago. For really big orchestra I use 150 to 200 cm widt, but then the function of the central array becomes a different one.

BTW, I compared VVMic , Surround Zone, and Harpex with the new Rode plugin. I would rate them as follows:

Harpex 10
Rode Plugin 8
Surroundzone 6
VVMic 6

I will try to post some sound bites soon...

Thanks Lucas! I agree - Harpex, and Blue Ripple Sound by Richard Furse (which uses Harpex algorithms in some of the plug-ins) have the best overall sound of any of the ambisonic processors I've tried so far. Angelo Farina is doing some great work processing HOA signals from an Eigenmike as well, getting far superior results to the original Eigenmike decoders. It would be great to hear an Eigenmike w/ his processing as the center of an array too! IEM and aalto SPARTA are making some cool things now as well.

I'd *love* to hear some samples from your recordings using the Soundfield plus AB pair. I'm moving to Los Angeles next month and am looking forward to getting back into music recording more, as I've drifted into Ambisonics and 360 video for the past few years. I think it'll be very interesting to see the ways people combine Ambisonics with traditional recording techniques, especially with all of the excellent tools available today, many for free.