chamber ensemble recording...spaced pairs?

[roonsbane]

Daniel,

I think you have hit it right on the head about the sense of mono. I believe it translates to a strange frequency response.

I have a buddy who is such a consistant engineer day in and day out. But one thing I will never agree with him on is panning of piano. He and I were listening to a Brad Mehldau recording and he thinks hearing that movement in panning across the keyboard is completeley weird. I say when he pans in his piano mics in to avoid this I can hear a strange frequency response the minute I walk in the room where as you usually have to be in the center of the speakers to notice that kind of movement. I would rather hear a single microphone or coindident pair over a spaced pair panned way in on piano.

I am not a fan of going over 1 meter on a string quartet, but 1 meter has a nice even spread across the 2 speakers with little or no hole, and with speakers as they are spaced in most control rooms this will give you an image very much as they would be on stage if you were say 20' in front of a group while bordering on spacious.
Cameron

[mpdonahue]

Originally Posted by roonsbane

pkautzsch said:


I am not going to try and explain why it's happening. It's there. Try sticking 4 of the same high quality omni's on a bar; One pair 1 meter apart and another pair at 40 cm and I am positive you will hear it. I can hear it the moment I walk into a room where the pair is too close.

I worked as the staff engineer in one of the very best chamber music halls in the country for 8 years (Jordan Hall, which was designed by the way by the same architect as Boston Symphony Hall at about the same time). This hall attracts many of the very biggest engineers in the business recording chamber music for either release or broadcast. I don't believe that I ever witnessed one single person use anything but a spaced pair of omni's at within 3" of the 1 meter spacing whether with or without additional highlights. Never were they close in like the DPA would recommend and no one used Cardioids and no one use ribbons for that matter in the main array. About the craziest people got were Brad Michaels and occasionally others used Schoeps tubes with the distortion control dialed all the way down. Others tried M50's (Mark Donahue) or M150's, but most of the time you saw Shoeps/DPA's/ Neumann/ or Sennheisers spaced at approximately 1 meter. That goes for Orchestral recording as well. Once you hear a direct comparison, there is a very good reason for this. I always thought there must be something to the DPA recomendations (after all it is DPA) so I have tried it along side the 1 meter setup. Everytime I hear that same buidup with the closer distance.

Cameron

Hey Cam,
With all this talk about spaced omni's and Jordan hall, I'll say that it all depends on the final delivery format to determine the best mic configuration in Jordan or any other hall. For SACD surround I normally use a distorted version of the ITU circle with 5 omnis. Which ones is dependent on the situation and desired result.
For stereo I've settled on a technique that Onno Scholtze showed me more than 20 years ago and it took me almost a decade to really grasp how to work with it quickly and consistently. It is actually based on both of the narrow and wide spaced omni setups listed above. In a line you put a very closely spaced pair (25-30cm) and a widely spaced pair(3m). They are all angled out at 45 degrees. This is a pure stereo pickup that combines the strengths close and widely spaced omnis and avoids the center build up of an M3(3 evenly spaced) setup. I have to admit that I've never been a big fan of any of the setups that sums a single microphone to both channels. The hard Mono component of the signal is tough to balance as it requires constant attention and its character really bothers me.
Finally, I use Omnis for mains 99.9% of the time. Very rarely I will use Mk 21's when the hall is really ugly sounding. The "Smeary" character of directional microphones is almost always a deal breaker for me.
As always, YMMV.

All the best,
-mark

[rumleymusic]

I am not a fan of going over 1 meter on a string quartet, but 1 meter has a nice even spread across the 2 speakers with little or no hole, and with speakers as they are spaced in most control rooms this will give you an image very much as they would be on stage if you were say 20' in front of a group while bordering on spacious.

When is comes to chamber music, I do like to hear a good stereo separation between the instruments. I usually achieve that by moving the mics not 2 feet in front of the ensemble. The separation is more apparent with more direct sound, and it allows a closer spacing without a sense of muddy buildup in the low frequencies. I admit that I tend to space omnis about 3' when recording chamber groups, and really only do the 40-60cm (or less) spacing when I am recording solo instruments or piano.

And like you I do like to hear a little bit of panning on the piano also. Normally I mic piano along the curve with moderately directional omnis (basically any pencil style will do) pointing at the high and low strings. Basically a NOS configuration with omnis.

[didier.brest]

Originally Posted by rumleymusic

And like you I do like to hear a little bit of panning on the piano also. Normally I mic piano along the curve with moderately directional omnis (basically any pencil style will do) pointing at the high and low strings. Basically a NOS configuration with omnis.

At which height ?

[roonsbane]

Hey Mark,

The man himself. Yes I am mr. short term memory. Right! your setup with the premeasured pull lines and the Omni's in an NOS like setup in the middle. I always saw you doing that with larger ensembles as is the Boston Symphony Hall setup. Didn't I see you doing some chamber music with just an A-B on a 1 meter bar live into that Tascam PortaDAT one time? You had that cool talkback system you had made? I remember that I liked your long stereo Mogami cable so much I ended up making one for my self?

I notice at Symphony hall in addition to the rig you just described above you also have a super large deccas tree of MKH800's along with flanks. Somewhere I have a program with a diagram with some 43 mics all laid out.

How do you use the MK800 array within the context of the "other" array? Have a Kelley's roast beef sandwich for Kyle and I.
Thanks Mark!
Cameron

[mpdonahue]

Originally Posted by roonsbane

Hey Mark,

The man himself. Yes I am mr. short term memory. Right! your setup with the premeasured pull lines and the Omni's in an NOS like setup in the middle. I always saw you doing that with larger ensembles as is the Boston Symphony Hall setup. Didn't I see you doing some chamber music with just an A-B on a 1 meter bar live into that Tascam PortaDAT one time? You had that cool talkback system you had made? I remember that I liked your long stereo Mogami cable so much I ended up making one for my self?

I notice at Symphony hall in addition to the rig you just described above you also have a super large Decca tree of MKH800's along with flanks. Somewhere I have a program with a diagram with some 43 mics all laid out.

How do you use the MK800 array within the context of the "other" array? Have a Kelley's roast beef sandwich for Kyle and I.
Thanks Mark!
Cameron

Well as they say, you steal from the best and make up the rest......
In a chamber music setup that is destined for stereo, I will usually use an AB pair on a stereo bar, usually less than a meter. A spot or two for color.
At the BSO we are actually printing several different main systems and ambiance sets for options in post. 99% of the time the main pickup for stereo is the Onno setup. When I mix the stuff, I listen to what will work best for the program at hand and choose accordingly. I'm very lucky to have a great relationship with the producer. She gives me virtually 100% autonomy to work up the sound I want, then we go about making small tweeks to the basic balance to bring out elements of the score. Whether the Maestro likes or not is a different story.... For the most part it is just balance adjustments.
So, you counted 43 mics? That must have been a big program like Les Troyens or some other opera. Normal week is somewhere around 32 tracks.

All the best,
-mark

[roonsbane]

It must have been you colleague Mr. John Newton who had the M50's up in a spaced pair for another session. I remember my buddy Tom Blanco who was the stage manager at Jordan, of course he is a classical music lover and always listens to many Jordan recordings, when he heard the M50's recording in the green room control room he said "Man that sounds awsome. It doesn't sound like Jordan hall, but it sounds awsome".
Cameron

[roonsbane]

Well you would have to have your head up your butt if you are doing several hundred recording each year and don't take any interest in the setups of some very fine engineers. Figuring out what they are doing and deciding what you do and don't like about their stups so that you can strive to improve upon it and hopefully develop your own thing is the interesting part of the job. I have probably 6 composition notebooks full of notes from any setup that I thought was a little unussual along with personal notes about what I did and didn't like about my setup. Eight years in Jordan was the perfect sonic laboratory. I tried a lot of ideas! Of course you have to still go and work in less than ideal halls and hopefully make them sound good as well.

So you are either using the Onno setup or you are using the giant MKH800 Triangle as the main array?

Cameron

[roonsbane]

Just to clarify Mark when you said:

In a chamber music setup that is destined for stereo, I will usually use an AB pair on a stereo bar, usually less than a meter. A spot or two for color.

You're not near 40 cm you are more like a few inches in from a meter if I remember correctly?
Cameron

[Ozpeter]

Kind of related - I've been playing around with samples from the net of recordings made using a Sony PCM-M10 recorder. This (curiously and uniquely) has a pair of built in omni mics very close together, not the usual cardioids, mounted (it would appear) at a 90 degree angle to each other. Separation seemed a bit limited, so I tried playing around with MS widening techniques. I was surprised to hear a complete change in the tonal characteristics when adding 6 to 8dB of extra side channel - a much brighter sound, not just simply a somewhat wider sound.

After some thought, I realised that at low to mid frequencies the original recording is bound to be close to mono, given that the published frequency plots for these omnis show that they have a equal response on and off axis at these frequencies - their spacing is too close for time-of-arrival to be very significant, and the 90 degree angle makes no difference at these frequencies. But at higher frequencies the mics become somewhat more directional, and when adding width via MS matrixing and dematrixing, the difference (side) signal has more HF components than the mid signal, and so the sound will brighten as it is widened.

This phenomenon will be affected by the characteristics of the mic - whether it is really omnidirectional at all frequencies - and with the distance between the capsules and the angle between them. I do wonder however whether, when a "low frequency build up" is experienced with lesser distances between the mics, some MS widening processing would mitigate this to some extent (at risk of some distortion of the stereo image, as LF signals will tend to remain centred, and HF signals will move outwards).

[RobAnderson]

Sorry to join the discussion so late, but in this video you can hear an example of spaced pair omni's on a chamber ensemble:

Sound Insights from Carnegie Hall: Preparing for the dance

The music was recorded with a pair of Schoeps MK2S's 38cm apart. Ignore the close mic's on the stands - these were to capture dialog between the players in-between takes and were not mixed in to what you are hearing in the video.

[mpdonahue]

Originally Posted by roonsbane

Just to clarify Mark when you said: You're not near 40 cm you are more like a few inches in from a meter if I remember correctly?
Cameron

Cam,
Yea, normally between 70cm and 100cm.
All the best,
-mark

[dtf]

Hi,

I mentioned in the very beginning of this thread that small spaced pairs produce good imaging but not the right sound. I have always found it inadequate that stereophonic microphone systems are categorized only by the quality of the image and how this is generated.

It's strange that the other effects of spacing microphone pairs wider are completely overlooked, up to the point that it is claimed they are non-existent.

Best,
Dirk

[rumleymusic]

Hi,

I mentioned in the very beginning of this thread that small spaced pairs produce good imaging but not the right sound. I have always found it inadequate that stereophonic microphone systems are categorized only by the quality of the image and how this is generated.

It's strange that the other effects of spacing microphone pairs wider are completely overlooked, up to the point that it is claimed they are non-existent.

I am just confused about what your distinction is between good "imaging" and good "sound" when it comes to identical mics in an AB pattern. The sound they pick up a 40cm compared to 100cm is not that different, there is just more delay between the mics and a different image on the playback medium. What other effects of spacing AB pairs wider are you referring to that are not related to the imaging?

I recently recorded a concert with a preexisting microphone setup. The main array was 2 Neumann KM183's with APE's spaced about 4 ft apart sharing an angle of about 90 degrees. That was combined with MKH800 outriggers and several DPA 4011 spots on the winds and various other neumann spots. The recording sounded great on headphones, but when I got to my studio to check it. There was a gaping hole in the center. The KM183 were simply too far apart, and with the APE's they were too directional to justify their placement, even with the spots. Not only does that reinforce the old "Do not mix with headphones" theory, but also provided a good "know your equipment" argument. The hole might not have been so bad if a) there were no APE's b) the angle wasn't so wide and/or c) the mics were closer together. It is impossible to say wider or closer is better with so many situational factors to consider.

Another possibility had occurred to me for the problem roonsbane was having, is the mics might not have been picking up enough shared high frequency information, i.e. they could have been angled so that the HF was separated by a directional characteristic in the mic, and the LF was not. I personally have never experienced what he has, but I like my DPA 4061's which are almost perfect omnis.

[roonsbane]

Daniel Said:

The sound they pick up a 40cm compared to 100cm is not that different

Well there sure is a much bigger sonic difference between 40 CM Vs. 100 cm, and the difference between 24 bit vs. 16 bit or 44.1 vs. 88.2. or transformer vs. no transformer or well built tube vs well built solid state.

I have always heard that whole in the middle with a four feet spread. Put them in around a meter and I hear it go away. Stick them around forty cm and that "schlock" appears.

Cameron

[rumleymusic]

Originally Posted by roonsbane

Daniel Said:


Well there sure is a much bigger sonic difference between 40 CM Vs. 100 cm, and the difference between 24 bit vs. 16 bit or 44.1 vs. 88.2. or transformer vs. no transformer or well built tube vs well built solid state.

Cameron

Well you are now talking about completely different things here, but I get your point. I just meant that, all things being equal, the only real difference in a wide and close AB placement is the image, the delay between the 2 mics. There is no mysterious phantom force changing the sound of the microphone.

[David Spearritt]

Originally Posted by rumleymusic

I just meant that, all things being equal, the only real difference in a wide and close AB placement is the image, the delay between the 2 mics. There is no mysterious phantom force changing the sound of the microphone.

But the image deepens (front/back), widens, and gets less accurate on source placement, ie. there are significant changes to the "image" and therefore the whole recorded sound.

[Shureman]

So speaking as a relative novice, isn't looking for the absolutely ideal mic configuration/pick-up pattern for specific musical settings all but fruitless unless there is a universally agreed to, standardized playback environment?

To hold playback variables constant, a "standard" listening setting should have the same dimensions, surface materials, moniters, moniter placement, amps, playback level, furniture, and on and on and on. Without close adherence to these playback variables, reliably determining subtle effects made by variables on the capture end seems impossible.

Please, I'm not inviting a food-fight. I'm just a little skeptical of the usefullness of acoustic determinations made about small changes in mic configurations and pick-up patterns when the experiment's measurements are made in non-standard settings.

Okay, set me straight - please be kind.

Shureman

[rumleymusic]

But the image deepens (front/back), widens, and gets less accurate on source placement, ie. there are significant changes to the "image" and therefore the whole recorded sound.

Exactly, I am NOT making a distinction between image and sound. That is why I thought the comment of good imaging vs good sound a little paradoxical, as if changing the former would not affect the latter.

To hold playback variables constant, a "standard" listening setting should have the same dimensions, surface materials, moniters, moniter placement, amps, playback level, furniture, and on and on and on. Without close adherence to these playback variables, reliably determining subtle effects made by variables on the capture end seems impossible.

You are exactly right. The trick is to find what method is going to sound good on a variety of playback situations, since speaker placement in most homes is not exactly standardized. I find that too much directional information in the microphones, especially in classical recording situations, sounds bad when speakers are too far apart. You kind of have to move around the room to find a spot that sounds good. AB spacings in general have similar information on both channels, just separated by delay that our brain interprets as direction. I think this sounds better than XY or ORTF methods on the majority of speaker setups or headphones.

I guess the question is, how much AB separation is needed for a good image, ie good sound. My answer would be, not a lot, but it depends on every other factor such as frequency separation needed/desired, instrument spacing, directional tendencies of the mic, multi mic reinforcement, etc. Roonsbane says he hears a hole at 4ft but not 1 meter, I think that is about accurate for the max you would want to separate without help. I do think that 40cm is adequate for most situations. But in both cases, again, that depends of the characteristics of the microphone being used and how it will translate when played back.

Please, I'm not inviting a food-fight. I'm just a little skeptical of the usefullness of acoustic determinations made about small changes in mic configurations and pick-up patterns when the experiment's measurements are made in non-standard settings.

Well the separation of the mics only affects the pitch of lowest separation. So a flute does not need a wide AB spacing because the wavelength distances it produces are rather short. The wider you go, the more frequencies you separate. This is standard no matter where you are. If width were the only factor that is.

[Shureman]

Great answer to my question. Makes a lot of sense.

[dtf]

Originally Posted by rumleymusic

I am just confused about what your distinction is between good "imaging" and good "sound" when it comes to identical mics in an AB pattern. The sound they pick up a 40cm compared to 100cm is not that different, there is just more delay between the mics and a different image on the playback medium. What other effects of spacing AB pairs wider are you referring to that are not related to the imaging?

Sorry to reply a little late to this.

I think there is a clear difference between the stereo image and the "sound" of a recording, but perhaps it would have been better to clearify what I meant. Most importantly, if you want to criticize my ineloquancy below, is that these changes happen - "why" is a different question.

First of all, a narrow spaced pair will always generate a more direct recording than a wider pair at the same spot. There is something happening in the low frequencies, which might be described as decorrelation, by which the spaciousness in the lower registers increases.

Another thing is an increase in transparency and naturalness of timbres. A widely spaced pair simply sounds more real, with the instrumental colours getting much more chance to develop.

Just out of interest, you don't hear any difference except in the image when making an AB wider?

[pkautzsch]

Originally Posted by rumleymusic

Well you are now talking about completely different things here, but I get your point. I just meant that, all things being equal, the only real difference in a wide and close AB placement is the image, the delay between the 2 mics. There is no mysterious phantom force changing the sound of the microphone.

There is definitely no change in the sound of "the microphone". One (1) microphone, that is.
If you have a system of two or more microphones, their signals will, upon playback, interact in some way - even if one is assigned to the left speaker, and the other one to the right speaker: We're hearing the left speaker with the right ear, too. That's a special kind of summing, in a way, and we need to cater for this when recording. BTW: that's one of the reasons why binaural miking techniques such as dummy heads, spheres, or Jecklin disks don't work too well with speaker playback.

Originally Posted by dtf

First of all, a narrow spaced pair will always generate a more direct recording than a wider pair at the same spot. There is something happening in the low frequencies, which might be described as decorrelation, by which the spaciousness in the lower registers increases.

This would, put in context with the "special kind of summing" I describe above, be a possible explanation for the LF, or lo-mid buildup mentioned earlier in this thread.
OTOH, I'm not sure if in chamber music you have that much of LF content. A string quartet, by definition, doesn't comprise 6 double basses, timpani, gran cassa, tuba, and bass trombone. In a spaced pair recording, you can't change the image after the fact, since turning your pan knobs inward would result in some degree of comb filtering. But you can change the sound quite easily with a good (hardware or plug-in) EQ.
We're heading towards the old discussion about how "real" a recording can or should be, and how much we need to cater for the difference between a concert hall (with the listener seeing the performers and hearing his neighbours' coughing) and the living room (listener seeing the bookshelf which houses his speakers, and hearing his fridge and the noise from the street).

Originally Posted by Shureman

So speaking as a relative novice, isn't looking for the absolutely ideal mic configuration/pick-up pattern for specific musical settings all but fruitless unless there is a universally agreed to, standardized playback environment?

A calibrated listening environment is standard in film post production. At least in Europe, I think public radio stations have some kind of standard, or tolerance curve, too.
You can't calibrate people's living rooms, right. In the Hi-Fi boom years, though, I think everybody knew about the equilateral triangle. As someone else pointed out in this thread, it's our job to make the recording sound as perfect as possible in the standard setup, but to sound good on non-standard systems too. That's why we have NS-10s and Auratones.

[rumleymusic]

Just out of interest, you don't hear any difference except in the image when making an AB wider?

If you have a system of two or more microphones, their signals will, upon playback, interact in some way

Well I do hear a difference in the sound, but I know that can be directly attributed by the frequency separation of the microphones' pickup and the position of the playback speakers (if there are only 2 mics on separate L/R chanels that is). Which I believe is the very definition of "image" in AB spacings. If you add more mics to the mix, image becomes more of a cumulative thing of course.

Most importantly, if you want to criticize my ineloquancy below, is that these changes happen - "why" is a different question.

Sorry, I don't mean to be critical. Things just come out that way when typing.

[NorseHorse]

Hello Erickx,

How did your chamber recordings in Florence go?

Have you had more time to experiment with omni spacings?

Cheers.

[Erichx]

hey!
They went great...It turns out that all i had were a pair of CAD e70's w/ omni and card capsules, and a Shure VP88 with M-audio fast track ultra 8r interface. apparently florence is a bit behind on the technical side!

But the venues were great, I used mostly the CADs' in close AB about 45cm apart, and got great results. I almost always lost the cello though, and eventually i managed to get hold of an SM57 and put that on it. I have to say that the VP88 was rather useless, because it didn't sound as good as a main pair compared to the CAD's.

still had some problems with stereo image w/ the omni's that close though...the center was always a bit confusing.

some samples of recordings: NYU Final Performance
check them out and let me know what you think! thanks again everyone for the fantastic feedback

[boojum]

Erichx and everybody, thanks for a great thread! It was mentioned earlier in this thread about how it is a good idea to watch what the real gurus do and note it well. I am unable to do that because I am out in the woods so I watch what they are doing here, and note it well.

The 40cm vs 75 - 90cm debate is interesting. Next time I am doing an omni classical pull in a good venue I will set one mic as the base, one at 40cm and one at ~80cm. I can then make the two mixes in post, 40cm and 80cm. I want to hear what the differences are. I have been using the Williams paper on mic spacing as my primer. Time to test it against what is being said here.

Again, all you folks who know by virtue of making a good living at this business, thank you for sharing your thoughts, theories and insights. This remains my master class. I thank you all.

Cheers