Outrigger benefits - anyone got a theory?

[studer58]

sorry, my x y and z's got all scrambled by the posting process so my point is lost...I need to re-compose and re-submit !

[studer58]

sorry, it wont let me re-compose..just slams all my text and ......'s against the left margin and ruins the point I'm trying to make

So....just imagine that y xx and z are equally spaced across the front of stage (top of page) and 'aa' is off to the far left of 'y' and 'a' is about equidistant between 'y' and 'xx' ..that gives you an idea of the locations that my text is referring to ! Hope it makes a bit more sense now ?
Ray

[MichaelPatrick]

studer58,

I understand your illustration and think you've highlighted a real mind-bender for those, like me, who like to calculate delay times. Some old-school recordists don't mess with delays.

[Larry Elliott]

Originally Posted by MichaelPatrick

Some old-school recordists don't mess with delays.

Many of the recordings that we now aspire to be able to replicate were made long before delays were feasible.

Setting delays for direct sound path differences is one thing, but the sound captured by a microphone contains Direct, Early Reflections. Reverberant decay and possibly some Late reflections (echoes). Trying to align all of these is impossible.

I’m a firm believer in the KISS principle, and letting technology serve the art rather than the other way around - things that I’ve learned the hard way during 40 years in the industry.

Larry

[rumleymusic]

So....just imagine that y xx and z are equally spaced across the front of stage (top of page) and 'aa' is off to the far left of 'y' and 'a' is about equidistant between 'y' and 'xx' ..that gives you an idea of the locations that my text is referring to ! Hope it makes a bit more sense now ?
Ray

The way one mics delay affects another is a boggling concept on paper, but not such a big deal in practice from my experience. The sound reaching the y and aa and a mics would be so different with the y mic capturing mostly indirect sound unrelated to the direct sound of a or aa, the sound field wouldn't really get messed up at all, any more than using a and aa in the first place. If you think about it, the whole purpose of spots is not to create the main image, but to lend support to instruments that might get washed out in the mains. The point of delay is of course to make sure the mains are the focus of the recording.

Many of the recordings that we now aspire to be able to replicate were made long before delays were feasible.

Well if you think about it, many of those recordings also didn't have spots OR outriggers to deal with either. It is more of a modern practice to stick everything you can find in front of an orchestra, which causes the problem in the first place. Delay is the small bandage to the gushing wound the engineer created by doing this.

I guess if you wanted to be "old school" you could try sticking up your stereo pair in front of the group and find unique ways to space the musicians so everything sound good in the recording. I could work, (until the cellists start complaining that they are not in their union designated seats and the flutist refuses to play because of the close proximety to the snare drum. Today's orchestra musicians are such babies.)

[pkautzsch]

Originally Posted by rumleymusic

The way one mics delay affects another is a boggling concept on paper, but not such a big deal in practice from my experience. The sound reaching the y and aa and a mics would be so different with the y mic capturing mostly indirect sound unrelated to the direct sound of a or aa, the sound field wouldn't really get messed up at all, any more than using a and aa in the first place.

I think "a" and "aa" aren't mics in his drawing, but sound sources. As the sound from "a" hits the outriggers "y" "z" and main pair "xx" about at the same time, but the "aa" sound hits "y" way earlier than "xx" which it hits still earlier than "z", and both mics are intended to pick up both "a" and "aa", delaying "y" and "z"would mess with the time relationship between "a" and "aa". It would, to be precise, make "aa" sound less far to the side than in the non-delayed outrigger setup.

If you think about it, the whole purpose of spots is not to create the main image, but to lend support to instruments that might get washed out in the mains. The point of delay is of course to make sure the mains are the focus of the recording.

That's why delaying spot mics is perfectly acceptable, and works better the more isolation you get between your spots, and the less reflected sound the spots pick up. Deutsche Grammophon is known to to use quite large numbers of close-ish spots which can be delayed very well.

Outriggers are not spots.
Outriggers are part of a main system. The Center mic in a Decca configuration isn't a spot either. Both are there to MAKE the width of the image: outriggers widen it a bit, and make the strings more spacious, the Decca Center mic helps picking up all strings equally (it's in front of L/R because of the violas!) and avoiding a hole.
Outriggers aren't room mics either.

Well if you think about it, many of those recordings also didn't have spots OR outriggers to deal with either.

Decca actually used outriggers quite early. Instead of using spots they used tons of blankets or wooden panels to optimize the room's acoustics so they got the depth they wanted, and they had the time to re-take if the ensemble's balance was off. That's what you describe as "old school".

I guess if you wanted to be "old school" you could try sticking up your stereo pair in front of the group and find unique ways to space the musicians so everything sound good in the recording. I could work, (until the cellists start complaining that they are not in their union designated seats and the flutist refuses to play because of the close proximety to the snare drum. Today's orchestra musicians are such babies.)

With the exception that you'd probably not put the flute next to the snare, at least in "repertoire" orchestral music.

[rumleymusic]

the Decca Center mic helps picking up all strings equally (it's in front of L/R because of the violas!)

Well the center mic is not in front because of the violas. It is in front because the delay created in relation to L and R creates a greater sense of 3D spaciousness throughout the entire orchestra. This is an example of how lack of delay compensation actually works. The idea was to combine the realistic image of a stereo pair recording with the fast response and wider depth of a muli-miking technique, using delay as a beneficial tool, and avoiding problems too many mics can create.

Yes delaying the outriggers (dare I say treating them as a spots, though I don't know why we all get so worked up over wording) will tighten the image, slightly. The outriggers are still panned L and R. And the amount that they are separated on the channel will still separate the image. To my ears, (and with my mics) this is actually beneficial. You need to listen to the recording to decide whether or not this the right choice. If the main pair does not cut it in terms of spaciousness, then forgoing any outrigger delay is probably the way to go. I don't pan the outriggers hard L/R when I delay them either. I let the main pair create the L/R image, and pan the outriggers according to their placement in the orchestra, delaying them so they match perfectly with the main pair.

With the exception that you'd probably not put the flute next to the snare, at least in "repertoire" orchestral music.

I think you missed my point on that one. I was making a joke about shuffling musicians where they are not supposed to be for the sake of recording.

I think "a" and "aa" aren't mics in his drawing, but sound sources.

That would change my point a little. But still aa would reach Y and Z slightly after xx and the sound from aa would be so close to a shell wall and so far away from the main pair that delay on the outrigger would be all but irrelevant for the placement in the image. There are a million way everything can get smeared. Outrigger delay helps low frequency separation in the front of the orchestra primarily, since the main pair are usually so close together the low wavelengths reach both L and R before waveform completes, and the Basses, being low and usually far to one side, are the most affected. Tuba, not so much. Not to mention that orchestras tend to narrow the further back they go so aa would really be somewhere between y and xx....ahhh this is all getting turned around.

Thanks a lot studer58! You just had to put variables in a subjective equation! LOL. I give up.

[7rojo7]

Turned around? No this is very close... Just to be difficult, music sounds best when you can smell it.
I've developped another sense

[sonare]

Something that has not been discussed is the playback level of outriggers/flankers compared to the main array. I start at -6dB and go down until the expansiveness of the room is still there but the image is not muddled.

RE the Tree-- an engineer who was there in the beginning says the flankers should be hard L & R but the tree L&R should not be. Gains are set to taste.

Rich

[pkautzsch]

Originally Posted by rumleymusic

Well the center mic is not in front because of the violas.

Apparently there are a lot of different "ideas" behind what we call Decca Tree today. The gist of what I have gathered from someone involved with Decca recordings for a very long time is:
- if they recorded an ensemble standing in straight rows, like a choir, they didn't move the center mic forward. It was just three mics in a row.
- if they recorded an ensemble in a half-circle, like most orchestras, they wanted to keep their mics at similar distances from the ensemble. As the violas (or whoever was sitting in the middle) are a few feet back, the center mic had to follow. Remember the tree usually stood right IN the orchestra, with the L/R mics (often) somewhere above 1st stand of violins respectively celli, and pointing outward, therefore, due to the M50's pattern, attenuating the violas quite a bit, in addition to their larger distance from mic. See where the necessity for that center mic and its position comes from? Often they would place the R mic a bit lower to get the celli as present as the high strings.
- Outriggers were usually panned hard L/R (or, rather, assigned, as their early desks didn't feature pan pots...), sometimes the mics could point slightly inward for more "glue", L/R were always hard L/R, and C was always dead center.
- The room was treated heavily with either blankets if too live or wooden panels if not live enough. I recall a photo of a session where you can see dozens of panels laid out all over the seating.
- Everything was decided by carefully listening, and optimizing.
- Lots of experimenting were done.

It is in front because the delay created in relation to L and R creates a greater sense of 3D spaciousness throughout the entire orchestra.

If it was like that, any wavefront would hit the Center mic first, no matter if it was violins, violas, celli, and therefore everything would be perceived in the center.

The idea was to combine the realistic image of a stereo pair recording with the fast response and wider depth of a muli-miking technique, using delay as a beneficial tool, and avoiding problems too many mics can create.

That's precisely what I'm saying. But that's not the Decca Tree idea, nor is it the outrigger idea. It's the idea Deutsche Grammophon has been perfecting.

Yes delaying the outriggers (dare I say treating them as a spots, though I don't know why we all get so worked up over wording) will tighten the image, slightly. *snip* I let the main pair create the L/R image, and pan the outriggers according to their placement in the orchestra, delaying them so they match perfectly with the main pair.

Well, a spot mic, to me, is used to get "more presence", and needs to be panned to where the source appears through the main pair (or...to correct placement of, say, soloists). Outriggers in my terminology are used for "more spaciousness" and not as spots for the outer violin stands.

Originally Posted by sonare

Something that has not been discussed is the playback level of outriggers/flankers compared to the main array. I start at -6dB and go down until the expansiveness of the room is still there but the image is not muddled.

RE the Tree-- an engineer who was there in the beginning says the flankers should be hard L & R but the tree L&R should not be. Gains are set to taste.

Rich

That's again proof that there seem to be lots of different "stories" about the Tree.
How did they achieve panning when they didn't have pan pots (did they really split the mic to two channels, one assigned left, the other right, and one being pulled down, say, 12 dB?). Or did they have pan pots (that would contrast what "my source" told me)?

[rumleymusic]

Outriggers in my terminology are used for "more spaciousness" and not as spots for the outer violin stands.

Fair enough, I shall call my outriggers: "forward string spots" from now on thumbsup

I got all my information on the Decca array from various literature I have read on the subject. I cannot advocate for everything I have read but most sources verify that the center mic is forward facing because of the phase relationship between the more prominent L/R mics. This allows faster attacks to reach the center mic first, which is distributed evenly between the L/R channels, before the more reverberant attack reaches the independent L/R mics that are usually facing somewhat left and right also. It is kind of like a slower attack on a compressor, allowing the transients to appear and sound sharper. The phase is usually not a problem because the delay between the mics only affects higher frequencies, and the directionality of the M50 usually does a good job of separating which high frequency information each mic picks up.

Like all stereo techniques, it is an illusion of depth rather than a proximity issue. The Decca tree's are used in straight line choral recordings quite frequently.

Something that has not been discussed is the playback level of outriggers/flankers compared to the main array. I start at -6dB and go down until the expansiveness of the room is still there but the image is not muddled.

Personally, I start with only the main pair and then add the flankers little by little until I am satisfied. They usually end up at about -6dB, or half the volume of the mains.

[dtf]

Originally Posted by rumleymusic

Like all stereo techniques, it is an illusion of depth rather than a proximity issue. The Decca tree's are used in straight line choral recordings quite frequently.

Personally, I start with only the main pair and then add the flankers little by little until I am satisfied. They usually end up at about -6dB, or half the volume of the mains.

pkautzsch already wrote most of the right things, but I can add that Decca did not have only one fixed way of recording. They worked out the best ways in relation to the location and ensemble - sometimes leaving out the center mic altogether and only using an AB.

The center mic was moved forward mainly for keeping the distance between the orchestral groups the same (therefore reducing the depth). Of course, it was also a result of having L and R closer to the violins and celli, which resulted in the famous "hole in the middle". It really comes from proper understanding of instrumental acoustics.

For choirs, three mics with the center mic forward would not make sense.

Best,
Dirk

[sonare]

Every year or so there is a thread that eventually morphs into a "tree thread" and this is no exception. Perhaps we need a tree sticky?

For the benefit of those who have not seen the earlier responses, here is a previous one:

http://www.gearslutz.com/board/remot...-new-post.html

And for any who might question my earlier comment about folding in the L-R of the tree slightly, here is a partial quote from Simon Eadon who began with Decca in the 60s:

"As for the tree dimensions, they should be 137cm from left to right mic with the Centre sticking out 68.5cm. For orchestral recordings we would use left and right outriggers - anything between 7'-10' from the centre. Obviously the panning of the tree is important. It should not be hard left and right. The centre mic placement should measure one foot back from the front of the orchestra. It is a frequent mistake to move the mics into the orchestra for more strings - moving the mics away is better. The heights typically are 9'6" to 10'09" but I have had tree and outrigers as low as 7' and high as 11' - it depends on the orchestra / venue / repertoire and probably a lot more!"

I am sure there will be doubters among the readers-- but this IS Gearslutz!

Rich

[KINGSWAY]

The outrigger conundrum is a most interesting puzzle. I would make the simple visual comparision:The main array, on it's own would give you an excellent 4:3 picture, with the addition of outriggers you have moved into wide screen territory.
As the inventors of the tree technique, Decca soon realized that the extremes of the
orchestra were being lost, with consequent reduction in perceived depth. Therefore, the standard method soon became a front capture ratio of main tree and outriggers as an organic whole.The actual pans are of paramount importance. The only eq required would be to remove any traffic rumble if necessary.
The subject of delay is fraught with difficulty: The major point is one of dominance,
do not delay your outriggers with respect to your main array if at equal gain.
The inevitable comb filtering will remove any possible advantages. Spot microphones
can be helped using delays, caveat emptor! If in doubt do not use.

[NorseHorse]

There is already comb-filtering before delaying, no?

[panatrope]

Originally Posted by NorseHorse

There is already comb-filtering before delaying, no?

There is always multipath which will result in 'comb-filtering' even with a co-incident pair, unless you record in an anechoic environment (open air, acoustic test chamber). If you use a spaced pair, when you broadcast it in mono, there will always be 'comb-filtering' for off-axis sounds.

However, in the matter of spotters vs main pair and the objective of delaying the spotter to be coherent with the direct path signal to the main pair, you can make it phase -coherent for one instrument in the spotter field, but of course if the performer gets carried away with the performance and moves, the coherence changes and you may hear the shifting comb-filter frequency (the old short-wave radio effect, or as they say in the '70s, 'flanging' - or its precursor the 1959 Toni Fisher recording of "The Big Hurt").

If the approach is to delay the spotter so that the wavefront arrives 10ms after the direct signal to the main pair, two things happen. One, the principal frequency of the phase cancellation moves down to about 50Hz, and secondly, the Haas/precedence/first wavefront theory comes into play - namely that location is determined by the 'direction of arrival' of that first wavefront, and the subsequent signal can be 12-15dB louder with detracting significantly from the localisation of that source.

However, there is plenty of opportunity in this pursuit for much more experimentation, and a whole pile of resulting theories in papers in the AES Journal. And as I remember some luminary being quoted to me in first year Engineering: "There is nothing so practical as a good theory!"

[MichaelPatrick]

I have to say that this thread is both fascinating and revealing. Fascinating because of the many thoughtful posts, and revealing because of all the theories and advice we're getting from practitioners. Thanks!

[rumleymusic]

If the approach is to delay the spotter so that the wavefront arrives 10ms after the direct signal to the main pair, two things happen. One, the principal frequency of the phase cancellation moves down to about 50Hz, and secondly, the Haas/precedence/first wavefront theory comes into play - namely that location is determined by the 'direction of arrival' of that first wavefront, and the subsequent signal can be 12-15dB louder with detracting significantly from the localisation of that source.

That is the approach I agree with taking. Though I usually use about an additional 20% delay rather than a strict 10ms. I don't know if there is any specific theory behind a fixed number rather than a percentage, though perhaps it is to line up the the spots equally in relation to the mains.

The inevitable comb filtering will remove any possible advantages.

I don't think that comb filtering is a worry-some issue with delaying spots or outriggers, at least not in the audible spectrum. It is more of a problem with 2 mics in the exact position picking up the exact same sound (such as two mics close to a guitar cabinet) and as panatrope so eloquently mentioned, the popular practice is to delay spots past the main pair, not try to match phase coherence exactly.

[Nobilmente]

Originally Posted by KINGSWAY

The outrigger conundrum is a most interesting puzzle. I would make the simple visual comparision:The main array, on it's own would give you an excellent 4:3 picture, with the addition of outriggers you have moved into wide screen territory.
As the inventors of the tree technique, Decca soon realized that the extremes of the
orchestra were being lost, with consequent reduction in perceived depth. Therefore, the standard method soon became a front capture ratio of main tree and outriggers as an organic whole.The actual pans are of paramount importance. The only eq required would be to remove any traffic rumble if necessary.
The subject of delay is fraught with difficulty: The major point is one of dominance,
do not delay your outriggers with respect to your main array if at equal gain.
The inevitable comb filtering will remove any possible advantages. Spot microphones
can be helped using delays, caveat emptor! If in doubt do not use.

This has been my experience too.

Delay on outriggers does not make sense for me, they are a simultaneous event with the Tree, whereas spots are fore and aft (nautical, but nice!) and delay might be of help there.

Even given the fore and aft though, I have not used delays for those as they are generally of quite low level if the seating is right. I'm not sure I would want to cope with the theory and the listening all at the same time, I like to move things around until it sounds well, and changing delays whilst doing that would be quite messy. It would be a pity too, if, after getting all the theory right it didn't sound too good.

In any case, if the mics are placed relatively high on the instruments further back, I guess you would need to subtract the "distance from instrument to spot" from the "distance instrument to main pair" to arrive at a correct delay. That could be quite a small distance in some cases.

It's encouraging to see what Mercury achieved with just three spaced omnis, the right room, seating & performance. Aptly named Mercury Living "Presence" I think! No spots, no need for delay.

I used to listen to these earlier LP recordings on a friend's 'super-dooper' hifi when I was at school (60's) (valve amps, large home made monitors with Goodman speakers and a specially adapted room etc) but didn't realise at the time how special everything was - now I know!

Perhaps not using delays has been my loss, and should (ought) be something for me to investigate in future.