i have been to a many concerts by large, well known orchestras (as an audience member) . Of course, I always observe the microphone arrangements (always seem that such concerts are recorded, presumably at least for archival purposes).
Nearly always the concert is multimiced. the usual arrangement has a main pair (I assume that is what it is) that are somewhere around 2 meters apart, and placed at a location which would yield an orchestra angle of at least 150 degrees.
When I plug these numbers numbers into the Sengpiel audio calculator, the resulting graphic looks abysmal.

What gives?

The main pair is being spaced to sound ‘wide’ for headphone and earbuds listening, with the spots used to populate the sparse wasteland in the centre with ‘details’ ?

If you plug in two omnis at 2m apart you get a very small 29 deg Stereo Recording Angle.

From the explanation page located at:
https://sengpielaudio.com/Fragen08.htm

The Stereo Recording Angle SRA:
The SRA is not the angle between the microphone axes, but an invisible angle, that gives you information about your stereo image. It tells you how "wide" the recording will sound. Every acoustic sources that are outside of the SRA will be located full at "L" or "R". So a smaller SRA will give you a wider stereo image. Everything inside of the SRA will find a place between the left and the right speaker.

yes, of course. here the issue is a wide Microphone spacing combined with a relatively wide Orchestra angle leading to a seemingly very poor SRA.
Orchestral angle as here:https://sengpielaudio.com/Ausdehnungsbereich01.gif

If, from the page linked below, you compare the three AB variants listed: the 30cm AB, the 60cm AB and the 120cm AB Sengpiel SRA optimal blue-shading areas, it's visually obvious that you're likely going to to be happiest in the 60cm (+/- 15cms) Goldilocks-realm

Thus...remarkably little wiggle room...yes ?
In fact, the oft-cited 67cms seems to nicely hit the target ?

https://sengpielaudio.com/Fragen08.htm

So why an AB pair at a 1.5 meters?
Something is going on here I do not understand. I mean, major orchestras are doing this- why?
Meanwhile, Micheal Williams' paper would suggest that at an orchestra angle of 150 degrees, one should go no more than about 38cm. weird.

Photos or it didn't happen. I haven't seen such arrangement much... Big AB plus closed mics, yes, or one of those 4 mic set-ups, but a single spaced pair above conductor spread that much - I have not seen. Maybe find a video or photo example to check?

One possible explanation is that the 1.5 m spaced mics aren't being panned hard L and R (which is the only thing the Sengpiel will illustrate).

Remember that the L and R mics (spaced 4.5' apart) of the classic Decca Tree we not panned hard L and R. And the main pair of it's 4-mic variation (spaced 3') was panned-in even more, since there was no center mic.

Another explanation is that there seem to be many engineers that care little about accurate stereo imaging, and only seek a big lush sound from the mains; I've heard many a Classical release that has a massive hole in the middle (they are especially onerous on headphones, where they create a physically uncomfortable phasiness).

i sometimes wonder how colleagues approach their tasks, but without having worked in the same room with a similar approach (apart from the main microphone system), with a similar ensemble and for a similar type of release/broadcast and without having compared their results with mine, i would be reluctant to criticize them; habits and tastes are too different...

that said, ime there are indeed a few setups that i would characterize as being a receipe for desaster, independent of the task!

from what I think I saw, it was a big AB plus outriggers and about a dozen spot mics.
could very well be that I got it all wrong, but that is what it looked like to me.
my next date with such a symphony is May 25. I will try to get a decent photo.

OK, here's an old chart I dug up.
and here's what was written about those front mics: "... the overall pickup design is based on four spaced, breasted and angled, Main omnidirectional microphones. ... the distance between these hanging microphones is approximately 7.5'; their height above the deck is around 9'. This spread is not typical for arrangements without a center microphone. For example, the downstage microphones in a Decca tree are likely to be spaced less that 7.5'. However, after many years of experimentation, we feel our current Main array produces a useful combination of flexibility, timbral accuracy, and musically appropriate balance for most standard symphonic concert literature."
And that is what I am seeing. I am not saying they are wrong; I am just saying I do not understand such setups, and I can not reconcile them with sengpielaudio.com.

Again, the chart nor the text say anything about how those four mics were panned (especially the inner pair).

Panning a widely spaced pair inwards just moves the two concentrated mono messes inwards, it doesn't improve the stereo image in any way, and may contribute to combing.

Actually I think closer to 54cm is the sweet spot for SRA and distribution of sources between speakers as evenly as possible for orchestral mains and where the mics typically end up in relation to the front row of players.

I’ve had cause to use a 3 mic Decca Tree on a small chamber ensemble (where a 5 mic Tree was unnecessary/too wide) …in that case, panning the outer mics hard L/R gave a good result. The 5 mic Tree is more of a ‘system’, and hence the recommended selective panning-in is effective and necessary

That's rather an assumption on your part; no where in the Dunkerley book is there mention of panning the tree any differently when used without outriggers.

Those at Decca (Dunkerley and Wilkinson included) who used the four-mic version of the Decca Tree obviously achieved very fine results with it; the 4-mic Tree pans the 3'-spaced main L/R inward a full 50%, in conjunction with a higher mix level of the outriggers than the classic 5-mic Tree. In theory, this should produce a comb-filtered mess, but in practice (if done correctly) it can sound fabulous.

But when a 3' spaced pair is used without outriggers, the in-panning required to effectively fill the middle is only about 80-85%.

This kind of of wide AB doesn't try to produce the kind pin-point imaging one gets from ORTF or Blumlein; the traditional 5-mic Decca Tree doesn't either - it has significant flaws in localization, but many, many of those who record full orchestra prioritize spacial richness and a sense of envelopment (especially at low freq) over a literalness of imaging.

Those who have actually done 3' AB, with in-panning compensation, will tell you it sounds nothing like "two concentrated mono messes" (neither do any of the many fine Dunkerley recordings done with the 4-mic Decca Tree). It has been my experience that anything beyond 3' cannot be effectively corrected by in-panning, and offers no advantages anyway.

It's highly recommended that people actually try something before rejecting on grounds of theory or 'numbers-on-paper'.

‘No where’ ? It’s hiding in plain sight ….the Reaper +\- 90 is equivalent to 100% L/R…**

The other DAWs listed are self explanatory. Most importantly, my ears tell me it’s right….irrespective of tables or dogma.

The -3dB for the centre mic is spot on….and I agree that the 4 and 5 mic Trees are a different kettle of fish entirely, re panning and level requirements, as ‘systems’

** The pan law settings for Reaper (any version from 5 upwards) as below need to be made in Preferences, so that the table settings (for all Tree variants) correspond. Pyramix, ProTools and RME TotalMix have similar pan law settings requirements, outlined in the book.

Hi Wildplum - The Sengpie Audiol charts, terms, framework are not connected to human perception and are not useful for determining good stereo arrays. For instance, compare:

NOS (30cm) = 81.01 SRA
XY (0cm) = 195.9 SRA

According to Sengpiel, XY should be "wider". But XY is demonstrably narrower than NOS BY FAR because it lacks time-of-arrival difference. And every cm reduction from 30cm makes the image narrower, but the SRA increases.

Trust your ears, not the charts.

The 5 mic Tree panning recommendations…as opposed to the 3

Thanks for the reply. And maybe this is part of my confusion about these charts.
If you look at the localization of the source as presented by its position between the speakers, my interpretation would be that NOS yields a wider perceived image by the listener than would X/Y (given the assumed location of the speakers).
In other words, a smaller SRA will yield a wider stereo image. At least that is the way I interpret the charts.
I think Micheal Williams approach ("The Stereophonic Zoom") might be better because (I think) I read that he based his calculations on actual listening experiments. But the two approaches sometimes (often?) yield different results. Hence my curiosity.
Yes, listening and trusting your ears is better, but in my end of the world, there often is no time for doing any serious moving around of things, once they are first setup. Then, experience and intuition must take over. I am hoping that if I can gain an understand these two approaches (and their respective charts), it might help me in those situation where my experience and intuition are, shall we say, not what they should be.

A helpful free app that you can download to your cellphone (and probably computer also) is Neumann’s ‘Recording Tools’ …it will expand on the Williams and Sengpiel’s tables (likely more helpful than both !) and you can adjust mic spacings, angles and SRA on your phone with your finger at a gig, to give you immediate information !

Here’s a description of how it works:

https://www.soundonsound.com/reviews...ecording-tools

I am not sure of this, but are Neumann definition of SRA and that of Sengpiel mean the same thing?.
Neumann: "If the SRA term is unfamiliar, it defines the relationship between how far to the left or right of the mic array a sound source can be placed before it reaches the left or right loudspeaker when auditioned in a normal stereo monitoring setup. "
Sengpiel: "The SRA is not the angle between the microphone axes, but an invisible angle, that gives you information about your stereo image. It tells you how "wide" the recording will sound. Every acoustic sources that are outside of the SRA will be located full at "L" or "R". So a smaller SRA will give you a wider stereo image. Everything inside of the SRA will find a place between the left and the right speaker."
(Note that Williams "Stereophonic Recording Angle" is not the same: SRA= sound source sector + "sideroom". Ain't this stuff fun!?)

Neumann and Sengpiel are broadly in agreement that it’s a measure of orchestral image width between the speakers (throw away your headphones !)

Williams is correct in his precision of definition… but Sengpiel’s interactive app lets you toy with the notion of images bunching up in the centre, being evenly spread across the L/R spectrum, or clustered around each speaker leaving a hole in the middle.

Try the 3 approaches and commit to the one that gives you the best results…

I too have suspicions about the rigour of Sengpiel’s testing methodology…so why not put that free Neumann app on your cellphone immediately, then start playing around with the variables !

I am a seasoned practitioner.

Right you are; I stand corrected.

What's your point?

I was referring to the four mic tree; not the three or the five.

I find that the book has several instances of ‘reiteration’….where you can think a particular item has been extensively covered, only to have it revisited in a later chapter, with additions or qualifiers. I put it down to the Decca book having multiple authors ! Makes simple skim or overview reading impossible…because it’s easy to miss things.

Simply that each variant of the Tree has it’s own panning table devoted to it in the book, not a one size fits all.

The curves given by Williams are derived from localization curves published in a master's thesis by Gert Simonsen (1984). These curves are significantly different than those measured by other authors and, in my experience, are not to be believed. You can find a comparison (by Sengpiel, of course!) here.

The Neumann and Schoeps apps hew closer to consensus localization curves and (again, in my experience) produce more accurate predictions.

David