Become a Site Supporter and Never see Ads again!

Author Topic: 4-mic Phased Array Recording  (Read 10731 times)

0 Members and 1 Guest are viewing this topic.

Offline voltronic

  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 2559
4-mic Phased Array Recording
« on: April 16, 2015, 08:37:53 PM »
After picking up the DR-70D, I wanted to try out a 4-mic array for the first time.  I decided to go with a variation on Tony Faulkner's 4-mic Phased Array which uses a pair of subcards flanked by omnis, all angled outwards at +/- 45 degrees.  Tony's preferred spacings are 47cm for the subcards and 67cm for the omnis.  Until my wide mic bar from followinbob arrives I am a bit limited for wide mounting, so for this particular recording the subcards were 36cm and the omnis 46cm.  Next concert I record I will try the 47/67 spacings.  (Note that if you search for "Faulkner Array", you will come up with the "other" array he is known for, which is 2 straight-ahead figure-8s spaced 20cm.)

Below is a dropbox folder with two tracks, and three different mixes of each.  I would appreciate feedback on which mixes are preferred.  I think all of them sound pretty good but in different ways.
1. Subcards 6dB lower than omnis.
2. Omnis 6dB lower than subcards.
3. Both equal (or as close as I could get setting levels by eye on the 70D.
4. Separate tracks of each pair of mics.

This is a high school small ensembles concert (some groups are student-led), so expect flaws in the performance.  I'm simply posting for critiques on my recording, and I chose the two most contrasting pieces.
The recording at this stage hasn't been altered in any way, not even a level adjustment.  I'm working on another set that is about 10-11dB higher and run through RX to kill the HVAC drone and will post those when done.
EDIT: The mp3 files are unaltered, but now there is also a folder of FLACs which have been run through Izotope RX - declick was used to reduce applause, I raised the gain 3dB, and denoised to remove the HVAC noise.

https://www.dropbox.com/sh/8t5hzhwlzhehjkt/AAAMfW5gdFpJGAP_08wZ8z2ba?dl=0

SETUP:

Line Audio CM3 (subcard): 36cm, angled +/- 45 deg out
Naiant X-Q (omni): 46cm, angled +/- 45 out

12ft up, stand in orchestra pit about 4 feet from stage lip.
Mic array about 4 feet above and 8-10 feet back from conductor position.

Mics > Tascam DR-70D (no external preamps)

Tascam DR-70D Settings:
- 24/48 BWAV
- 48V Phantom
- Limiter OFF
- HIGH gain - subcards at 10:00; omnis at 8:00.
- levels of all 4 channels set roughly equal by eye

Relevant links on this technique:
https://www.youtube.com/watch?v=8uCcFIyJJ-w&feature=youtu.be (start around 30 min in - the inner mics are a modified ORTF which he no longer uses)
http://www.recordproduction.com/tony-faulkner.htm (start about 17min - here he describes the Phased Array)
https://www.gearslutz.com/board/remote-possibilities-acoustic-music-location-recording/930912-three-mixes-boojum-jnorman-case-study.html (Tony F contributes to this thread)
https://www.gearslutz.com/board/remote-possibilities-acoustic-music-location-recording/730017-faulkner-arrays.html
https://www.gearslutz.com/board/remote-possibilities-acoustic-music-location-recording/492654-faulkner-array-insta-snake-action.html
http://taperssection.com/index.php?topic=152926.30

EDIT: One thing I forgot to mention earlier is that neither of the spaced stereo pairs in this array might seem to be a good choice on their own for relatively close recording if you are looking at SRA graphs, either in Tony Faulkner's preferred spacings, the modified "Boojum/Jnorman" versions, or the further modified one I used.  The subcards would seem to be spaced a bit too wide and the omnis a bit too narrow.  Clearly, there is more than meets the eye going on with the interaction of the two pairs.  From Tony's interviews and written statements, it seems that having the omnis 10cm out from each subcard is an important aspect of getting this right.  Since the widest I could set my CM3s was 36cm, that's why I went with 47cm for the omnis.
« Last Edit: April 18, 2015, 08:44:08 PM by voltronic »
DPA 4061 | Line Audio CM3 | Naiant X-Q
Naiant PFAs | Shure FP24
Tascam DR-70D JWMod | Sony PCM-M10

Tascam DR-70D FAQ
Team Line Audio
Quote
I am hitting my head against the walls, but the walls are giving way.    ///    If a composer could say what he had to say in words he would not bother trying to say it in music.
- Gustav Mahler

Offline 2manyrocks

  • Trade Count: (12)
  • Taperssection All-Star
  • ****
  • Posts: 1647
Re: 4-mic Phased Array Recording
« Reply #1 on: April 16, 2015, 10:54:32 PM »
How about posting the omnis and CM3s as separate files for us to mix to our own taste?  I think the instrumental piece is an easier selection to focus on.  Of the samples posted, I thought the omni -6 was a bit more clear on the instrumental. 

About how big is the stage and how many players?  Any observations on the acoustics of the room that you think we ought to know?

Offline voltronic

  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 2559
Re: 4-mic Phased Array Recording
« Reply #2 on: April 17, 2015, 06:00:15 AM »
How about posting the omnis and CM3s as separate files for us to mix to our own taste?  I think the instrumental piece is an easier selection to focus on.  Of the samples posted, I thought the omni -6 was a bit more clear on the instrumental. 

About how big is the stage and how many players?  Any observations on the acoustics of the room that you think we ought to know?
I can post the separate files.  the CM3 track on its own doesn't sound that great, but the omnis on their own are pretty good.

Stage is the typical size for a high school auditorium - just big enough for a full choir and orchestra.  The hall itself is pretty dry for the most part and has some serious acoustic issues due to parallel untreated sidewalls.

Vocal example was 19 students in two parallel rows, spread 2/3 of the width and all the way downstage.  The soloist was right of center as he appears on the recording.

String example was 23 players in the traditional chamber strings arrangement of (L-R) violins-violas-cellos with the two basses behind the cellos.  They were a bit more upstage to allow room for the conductor.
DPA 4061 | Line Audio CM3 | Naiant X-Q
Naiant PFAs | Shure FP24
Tascam DR-70D JWMod | Sony PCM-M10

Tascam DR-70D FAQ
Team Line Audio
Quote
I am hitting my head against the walls, but the walls are giving way.    ///    If a composer could say what he had to say in words he would not bother trying to say it in music.
- Gustav Mahler

Offline voltronic

  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 2559
Re: 4-mic Phased Array Recording
« Reply #3 on: April 17, 2015, 06:41:18 AM »
Dropbox folder now has individual CM3 and X-Q tracks added.
DPA 4061 | Line Audio CM3 | Naiant X-Q
Naiant PFAs | Shure FP24
Tascam DR-70D JWMod | Sony PCM-M10

Tascam DR-70D FAQ
Team Line Audio
Quote
I am hitting my head against the walls, but the walls are giving way.    ///    If a composer could say what he had to say in words he would not bother trying to say it in music.
- Gustav Mahler

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12607
  • Gender: Male
Re: 4-mic Phased Array Recording
« Reply #4 on: April 17, 2015, 09:47:43 AM »
Thanks for posting this. 

As you know I'm very interested in these types of arrays.  In real-world taper-section recording terms, exact spacing may not matter as long as the resulting recording sounds good, but in terms of what Faulkner is doing, the spacing isn't arbitrary.  I think he's arriving at the spacings he's using not by calculation but empirically, through careful monitoring while setting up.   We don't really have that luxury.  But that he calls it a Phased Array is a hint that dialing in the appropriate spacing is important due to the complex phase interaction between the four points along the line (the four mics on the bar).

Your observation about the three general mix level relationships, and how each can be appropriate in different ways, parallels my experience using a three mics in a row (spaced omni pair with a third mic in the center, either a third omnis or a directional), and five mics in a row (wide spaced omnis, plus a 3-channel OCT/OCT-2 setup in the center), mixed to stereo.

Looking forward to listening to this but may not have a chance to until I get back from a trip in about a week and a half.
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Bruce Watson

  • Trade Count: (0)
  • Taperssection Regular
  • **
  • Posts: 86
  • Gender: Male
Re: 4-mic Phased Array Recording
« Reply #5 on: April 18, 2015, 02:35:53 PM »
...But that he calls it a Phased Array is a hint that dialing in the appropriate spacing is important due to the complex phase interaction between the four points along the line (the four mics on the bar).

Not exactly. It's called (I seem to recall Mr. Faulkner denied creating the name) a phased array because he lines up the mic capsules to insure the four caps are in phase. The reason for this is the wave lengths at the higher frequencies. I'm sure this crowd gets the implications.

That said, what I've read on this indicates that he gets the exact spacings by listening, and that the 67cm / 47cm spacings are a good place to start. He doesn't encourage anyone to take those spacings as inscribed in stone.

Offline 2manyrocks

  • Trade Count: (12)
  • Taperssection All-Star
  • ****
  • Posts: 1647
Re: 4-mic Phased Array Recording
« Reply #6 on: April 18, 2015, 05:19:24 PM »
I find myself perplexed at the CM3 portion of the recording.  Expected more from them and the 70d.  An EQ notch clears things up a bit, but I'm not sure what has happened here. 

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12607
  • Gender: Male
Re: 4-mic Phased Array Recording
« Reply #7 on: April 18, 2015, 06:21:46 PM »
...But that he calls it a Phased Array is a hint that dialing in the appropriate spacing is important due to the complex phase interaction between the four points along the line (the four mics on the bar).

Not exactly. It's called (I seem to recall Mr. Faulkner denied creating the name) a phased array because he lines up the mic capsules to insure the four caps are in phase. The reason for this is the wave lengths at the higher frequencies. I'm sure this crowd gets the implications.

Not exactly.  Its not as simple as bringing the signals into phase.  Allow me to explain what is going on phase-wise with this geometry.

The signals from the four microphones will be in-phase for any plane-wave with an origin perpendicular to the axis of the array, regardless of the microphone spacing.  In less technical terms, that means sounds arriving from directly in front (or behind, above or below) with their source a long distance away will be in phase at all frequencies regardless of the spacing between microphones. This is the 'forward-gain' aspect of this array.  And this forward-facing-gain aspect is compounded by the number of elements in the array arranged in a line perpendicular to gain axis.

However, for any sounds arriving from off center, there will be a complex phase relationship between the signals of the four microphones.  Pick any to microphones in the array, and the phase-relationship of their signals will change based on 3 variables: the angle of arrival; the spacing between those two microphones; and the frequency in question.  Change any of those variables and the phase relationship changes.  Sounds from off axis don't get the same 'directional-gain' across all frequencies.

Here are the basic implications:
>The signals will be more in-phase at the lowest frequencies, and will have increasing phase difference at higher frequencies.
>The out of phase component of the signals will be closer in-phase for sounds originating near the median plane, and will have increasing phase difference at wider angles of arrival.
>The out of phase component of the signals will be more in-phase at closer microphone spacings, and will have increasing phase difference at larger microphone spacings.

That's the case for any two microphones spaced apart from each other. 

When there are four microphones instead of just two, those complex phase relationships are multiplied by six.  That's because there are then six pair relationships between the four microphones, rather than one pair relationship between two.  So the phase relationships get incredibly complex away from the median plane.

Quote
That said, what I've read on this indicates that he gets the exact spacings by listening, and that the 67cm / 47cm spacings are a good place to start. He doesn't encourage anyone to take those spacings as inscribed in stone.
^^
This is the practical take away, and what I was primarily attempting to convey in my previous post.  He's listening while adjusting the spacing, and that's the only practical way to optimally arrange things.

It's very informative to listen while actively changing the spacing of a single pair of omnis.   I encourage everyone to try this themselves, listening to the performance with headphones while varying the spacing of the two omnis (most here won't have an assistant who can slowly vary the spacing while we are listening on a speakers in some isolated room).  I think many recordists think of microphone spacing as just affecting left-right imaging and other SRA aspects, but having done this myself a number of times, I find the tonal and 'textural' aspects are often more significant.  Any one here who tries it is likely to hear this immediately, far more clearly and obviously than the SRA changes, especially when listening while making the spacing change. Comparing short segments recorded with a few different spacings is also helpful and often the more practical way to do this, but the feedback-loop and mental association with this relationship is far less direct.

One can hear these complex phase relationship shift up and down in frequency.  There will be frequency specific reinforcement at phase angle differences near zero and multiples of 360 degrees, and attenuation near 180 degree phase angles and multiples of 180 degrees.  At frequencies where the first cancellation and reinforcement happens, the attenuation and increase is quite audible. It isn't heard so much as a level differences at higher phase angle differences where phase rotation happens at ever-increasingly narrow frequency bands, but the 'texture' and 'diffusivity' changes if you'll allow me those subjective descriptors.

This is likely to explain what you are hearing and EQ adjusting for in the CM3 portion, 2manyocks.  If the spacing was being actively changed, you'd be likely to hear that particular aspect you are compensating for with the EQ notch shift upwards and downwards in frequency along with the change of spacing.   When one is listening while setting up, a big part of 'tuning' the spacing is 'tuning' those phase/frequency relationships.
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12607
  • Gender: Male
Re: 4-mic Phased Array Recording
« Reply #8 on: April 18, 2015, 06:36:53 PM »
I seem to recall Mr. Faulkner denied creating the name

For the reasons described above, any non-coincident microphone array can technically be considered a "phased array".  Adding additional non-coincident microphones makes an array more significantly phased.  If Orwell was a recording engineer he may have phrased it this way-  "All spaced microphone arrays are phased, but some are more phased than others."

Mr. Falkner of course is more astutely aware of this than most, so it is not surprising that he would not give this particular setup the name "phased array".  It's not incorrect in a technical descriptive sense, but it is overly general.  And I suspect he's most interested in it in a practical sense than a technical sense anyway.
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline voltronic

  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 2559
Re: 4-mic Phased Array Recording
« Reply #9 on: April 18, 2015, 06:43:22 PM »
Thanks for that very detailed explanation.  Quick question: in the future I may find myself in the situation where my entire array is going to be closer to and higher above the ensemble.  In that case I have angled slightly down when using one stereo pair, but now I want to be careful to preserve consistent arrival time at the 4 capsules.  I would be using a posi-lok clutch to move the entire bar, but my 4 mounts aren't the same height.  Should I adjust so that the capsules are still aligned in a straight vertical plane, or more like a plane angled toward the choir below?  I'm visualizing the old PZM technique on a big plexiglass sheet flown above and angled toward performers on stage.
DPA 4061 | Line Audio CM3 | Naiant X-Q
Naiant PFAs | Shure FP24
Tascam DR-70D JWMod | Sony PCM-M10

Tascam DR-70D FAQ
Team Line Audio
Quote
I am hitting my head against the walls, but the walls are giving way.    ///    If a composer could say what he had to say in words he would not bother trying to say it in music.
- Gustav Mahler

Offline voltronic

  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 2559
Re: 4-mic Phased Array Recording
« Reply #10 on: April 18, 2015, 06:51:25 PM »
Also, would it be helpful if I posted FLACs instead?  I have read that MP3 encoding may do some summing of common information between channels, and that may affect what we are hearing. 

Regarding the CM3 track on its own, I'm not sure what EQ adjustment 2manyrocks made, but taken on its own it's the "worst" recording I've made with them.  But look at the SRA for that spacing and it's extremely narrow for how close my stand was.  I do think they really help complete the image of the Omni's though.
DPA 4061 | Line Audio CM3 | Naiant X-Q
Naiant PFAs | Shure FP24
Tascam DR-70D JWMod | Sony PCM-M10

Tascam DR-70D FAQ
Team Line Audio
Quote
I am hitting my head against the walls, but the walls are giving way.    ///    If a composer could say what he had to say in words he would not bother trying to say it in music.
- Gustav Mahler

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12607
  • Gender: Male
Re: 4-mic Phased Array Recording
« Reply #11 on: April 18, 2015, 07:08:08 PM »
I want to be careful to preserve consistent arrival time at the 4 capsules.

The arrival time is going to be all over the place for all the different positions in the ensemble.

I think you just need to try things and listen for what works.  The 'forward-gain' aspect of this assumes equal or near equal path lengths from the sound source in question to all microphones.  Falkner talks about using this for that 'forward gain' aspect to help when not very close.  But that aspect of it may not be important for you if you are using it close to the ensemble.  In that case it's probably more about the timbrel and imaging aspects of the four in combination.
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12607
  • Gender: Male
Re: 4-mic Phased Array Recording
« Reply #12 on: April 18, 2015, 07:09:50 PM »
mp3 is probably fine.  Can you hear the same aspects in the mp3 compressed version?  Downloading them now.
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline voltronic

  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 2559
Re: 4-mic Phased Array Recording
« Reply #13 on: April 18, 2015, 07:14:42 PM »
mp3 is probably fine.  Can you hear the same aspects in the mp3 compressed version?  Downloading them now.
For the most part, yes but it might be useful to have the full frequency range.  I think I'll post FLACs of my lightly processed version which is a few dB higher and with the HVAC reduced.  The noise removal in RX was very transparent.
DPA 4061 | Line Audio CM3 | Naiant X-Q
Naiant PFAs | Shure FP24
Tascam DR-70D JWMod | Sony PCM-M10

Tascam DR-70D FAQ
Team Line Audio
Quote
I am hitting my head against the walls, but the walls are giving way.    ///    If a composer could say what he had to say in words he would not bother trying to say it in music.
- Gustav Mahler

Offline voltronic

  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 2559
Re: 4-mic Phased Array Recording
« Reply #14 on: April 18, 2015, 07:22:05 PM »
I want to be careful to preserve consistent arrival time at the 4 capsules.

The arrival time is going to be all over the place for all the different positions in the ensemble.

I think you just need to try things and listen for what works.  The 'forward-gain' aspect of this assumes equal or near equal path lengths from the sound source in question to all microphones.  Falkner talks about using this for that 'forward gain' aspect to help when not very close.  But that aspect of it may not be important for you if you are using it close to the ensemble.  In that case it's probably more about the timbrel and imaging aspects of the four in combination.
OK, that makes sense.  And regarding the spacings, I fully intend to adjust by listening.  As TF says they are starting points although he mentions the 67 cm omni spacing quite often.  Clearly he has found that consistently reliable for what he does, but we don't have his equipment nor are we recording at Abbey Road...
DPA 4061 | Line Audio CM3 | Naiant X-Q
Naiant PFAs | Shure FP24
Tascam DR-70D JWMod | Sony PCM-M10

Tascam DR-70D FAQ
Team Line Audio
Quote
I am hitting my head against the walls, but the walls are giving way.    ///    If a composer could say what he had to say in words he would not bother trying to say it in music.
- Gustav Mahler

 

RSS | Mobile
Page created in 0.279 seconds with 42 queries.
© 2002-2018 Taperssection.com
Powered by SMF