Gearslutz

Gearslutz (https://www.gearslutz.com/board/)
-   All Things Technical (https://www.gearslutz.com/board/all-things-technical/)
-   -   imaging improvements to narrow omni spacing. Live recording. (https://www.gearslutz.com/board/all-things-technical/1284547-imaging-improvements-narrow-omni-spacing-live-recording.html)

Geoff Poulton 27th October 2019 03:29 PM

imaging improvements to narrow omni spacings
 
What techniques are you employing for recordings made with narrower omni spacings to make the image sound wider please?

0VU 27th October 2019 03:38 PM

Quote:

Originally Posted by Geoff Poulton (Post 14288349)
The 17mm spacing

You do mean 17cm don't you? kfhkh

Geoff Poulton 27th October 2019 03:43 PM

Quote:

Originally Posted by 0VU (Post 14288369)
You do mean 17cm don't you? kfhkh

Oh Yes. 17cm!

I've changed it now.

studer58 27th October 2019 04:35 PM

That you were able to get stereo imaging from a pair of narrowly spaced (and especially angled as well !) omnis is actually not surprising. Most omnis do become more directional at higher frequencies, which their polar plots will show you...especially above 5kHz

Admittedly you won't be getting ORTF-like stereo separation, and you'd expect the width to be somewhat curtailed....but as you've found, closely spaced omni mics certainly don't produce a mono image !

I maintain that a Rode NT-4 XY mic will give you a small but plausible stereo spot miking width, if the cardioid capsules are replaced by the Rode NT 45-O omni.....which conventional wisdom and theory would say is impossible, as they're effectively coincident ! That might be so for the bass, but not for the upper HF region.

Geoff Poulton 27th October 2019 04:58 PM

Quote:

Originally Posted by studer58 (Post 14288451)
That you were able to get stereo imaging from a pair of narrowly spaced (and especially angled as well !) omnis is actually not surprising. Most omnis do become more directional at higher frequencies, which their polar plots will show you...especially above 5kHz

Admittedly you won't be getting ORTF-like stereo separation, and you'd expect the width to be somewhat curtailed....but as you've found, closely spaced omni mics certainly don't produce a mono image !

I maintain that a Rode NT-4 XY mic will give you a small but plausible stereo spot miking width, if the cardioid capsules are replaced by the Rode NT 45-O omni.....which conventional wisdom and theory would say is impossible, as they're effectively coincident ! That might be so for the bass, but not for the upper HF region.

So I'm probably on the right track to improvement if I use a shuffler do you think? That way the bass would widen too? The set up was not something that I would have considered doing, but I was surprised how much of a stereo spread there was from it, and the MK5 does have a 2db rising high end from around 2.5kHz if I remember correctly. It really differed little, if at all, from ORTF as far as width is concerned.

I guess you might expect to lose image definition employing omnis in any case, which would be a trade off for the other advantages of spaced omnis. I don't use spaced omnis at all very much as a main pair, unless they are in the Faulkner phased array, which I do use often. From ORTF experience I don't think the distant instruments would have had very much more definition than they had with this closely spaced and angled omni set up. The width has probably been saved by having the mic quite close to the ensemble, which will have also worked in the omnis favour.

There was nowhere to monitor on speakers and it sounded well on heaphones. I am very pleasantly encouraged by what I'm hearing now - not anticipated.

It will teach me to use MS more in these situation perhaps!

heva 27th October 2019 07:18 PM

Try Blumlein shuffling, invented for narrow spaced omnis (though with a baffle).

Geoff Poulton 28th October 2019 01:49 AM

Quote:

Originally Posted by heva (Post 14288640)
Try Blumlein shuffling, invented for narrow spaced omnis (though with a baffle).


Thank you for the thought.

I have a DAV SIPP which I will use for the shuffle tomorrow. Crossed fig8s may benefit from shuffling in some instances too, but it is not always needed with that setup. I feel sure the narrower spaced omnis will benefit though, we shall see. Sounds with a wide frequency content, can split off and appear to be where they are not, or appear larger, convering a wider area then they really do. I think it's just a case of not overdoing it and taking care to listen.

I was very surprised at the width of the recording though, I half expected everything to be bunched up in the centre a bit like 90 degree coincident cardiods, but it is pretty well evenly spread from speaker to speaker, the 110 degree angle, high frequency directionality and closeness of the set up are responsible for that I think, the shuffling should enhance it further by increasing the bass spread more evenly too.

studer58 28th October 2019 02:33 AM

Try plugging your mic type, angles and distances into this calculator and see what sort of SRA it gives you: http://www.sengpielaudio.com/Visualization-AB30-E.htm

If you're unfamiliar with the SRA concept, give yourself an hour or two to chew on Williams' paper: https://microphone-data.com/media/fi...%20zoom-10.pdf

Stradivariusz 28th October 2019 05:23 PM

Can anyone suggest something to read on this subject?

On this moment I'm reading this: https://www.audiosignal.co.uk/Resour...uffling_A4.pdf

Excited to read more!

deedeeyeah 28th October 2019 05:39 PM

i'd try a 1st order allpass filter on one track.

GYMusic 28th October 2019 05:48 PM

Hope this is on topic. Learned about ORTF in the late 60's. Been a fan ever since for drum overhead. Sometimes using just the pair and a kick mic.

Also used the Crown SASS system years ago with great success.

http://www.coutant.org/sass/126982.pdf

Geoff Poulton 28th October 2019 05:58 PM

Quote:

Originally Posted by Stradivariusz (Post 14290204)
Can anyone suggest something to read on this subject?

On this moment I'm reading this: https://www.audiosignal.co.uk/Resour...uffling_A4.pdf

Excited to read more!

You will find this interesting to read.

http://www.davelectronics.com/sipp.htm

The DAV SIPP is used for more than just shuffling, there are several processes it deals with. I think it's pretty wonderful!

There are quite a number of papers around on Google relating to shuffling, but if you look around (youtube?) there are example videos of its effect that demonstrate it being switched in and out of a circuit as an illustration. You can choose to use it or not, depending on whether or not you feel it is needed.

Geoff Poulton 28th October 2019 06:03 PM

Quote:

Originally Posted by deedeeyeah (Post 14290221)
i'd try a 1st order allpass filter on one track.

What frequency would you aim for with that deedeeyeah? Or would it depend on.........?

Geoff Poulton 28th October 2019 07:42 PM

Quote:

Originally Posted by GYMusic (Post 14290242)
Hope this is on topic. Learned about ORTF in the late 60's. Been a fan ever since for drum overhead. Sometimes using just the pair and a kick mic.

Also used the Crown SASS system years ago with great success.

http://www.coutant.org/sass/126982.pdf

ORTF, Faulkner phased array and Blumlein 90deg crossed fig 8s are my three most used set ups.

Room permitting, Blumlein for smaller ensembles, I love it!

deedeeyeah 28th October 2019 08:12 PM

Quote:

Originally Posted by Geoff Poulton (Post 14290261)
What frequency would you aim for with that deedeeyeah? Or would it depend on.........?

yes, it depends...

short answer: the lf range as you experience the largest difference in level with the phase relationship off, the mids as there is simply the most 'traffic'/our ears are the most sensitive and the hf range which can make the stereofield more wide.

___


long answer: i mostly measure the room before i set up spaced mics and then decide on their spacing (although i hardly ever end up using spaced mics as mains but anyway):

let's assume you experience a 'room tone' at 100hz: you could almost cancel out this frequency by using a spacing of 83,5cm which would get you (up to) 90° degrees phase offset at 100hz while all other frequencies would have a less (or a less noticeable) phase offset.
then you set identical apf's on both tracks but invert phase on just one track until you get +90° one one side and -90° on the other side at the frequency which you want to cut down.
with multple apf's you can affect the phase relationship in multiple frequency areas - but also completly mess up things, so handle with care! :-)

i'm using a measurement mic, an fft to analyze and a speaker processors for precise alignment (and i record clave clicks right in front of the main mics). without fft, a goniometer and even a simple spl meter are also helpful (although way less precise).

___


p.s. i learned s few things from measuring and aligning large speaker systems and especially from designing and measuring subwoofer arrays: same principles on the way in as on the way out...

GYMusic 28th October 2019 08:40 PM

Quote:

Originally Posted by Geoff Poulton (Post 14290482)
ORTF, Faulkner phased array and Blumlein 90deg crossed fig 8s are my three most used set ups.

Room permitting, Blumlein for smaller ensembles, I love it!

Never tried the Faulkner phased array. TNX for the tip.

chris661 28th October 2019 11:41 PM

Any chance of processing it as M/S and pushing the -side channel up?

Probably the simplest way to do it, but I'd be interested to try the other methods.

Chris

Geoff Poulton 29th October 2019 02:46 AM

Quote:

Originally Posted by deedeeyeah (Post 14290545)
yes, it depends...

short answer: the lf range as you experience the largest difference in level with the phase relationship off, the mids as there is simply the most 'traffic'/our ears are the most sensitive and the hf range which can make the stereofield more wide.

___


long answer: i mostly measure the room before i set up spaced mics and then decide on their distance (although i hardly ever end up using spaced mics as mains but anyway):

let's assume you experience a 'room tone' at 100hz: you could almost cancel out this frequency by using a spacing of 83,5cm which would get you (up to) 90° degrees phase offset at 100hz while all other frequencies would have a less (or a less noticeable) phase offset.
then you set identical apf's on both tracks but invert phase on just one track until you get +90° one one side and -90° on the other side at the frequency which you want to cut down.
with multple apf's you can affect the phase relationship in multiple frequency areas - but also completly mess up things, so handle with care! :-)

i'm using a measurement mic, an fft to analyze and a speaker processors for precise alignment (and i record clave clicks right in front of the main mics). without fft, a goniometer and even a simple spl meter are also helpful (although way less precise).

___


p.s. i learned s few things from measuring and aligning large speaker systems and especially from designing and measuring subwoofer arrays: same principles on the way in as on the way out...


I'll spend some time looking at this (already have done) and trying to grasp the whole of what you're saying!

Geoff Poulton 29th October 2019 02:47 AM

Quote:

Originally Posted by chris661 (Post 14290937)
Any chance of processing it as M/S and pushing the -side channel up?

Probably the simplest way to do it, but I'd be interested to try the other methods.

Chris

Where would you start!?

(I saw the name Grimshaw and knew I was somewhere up t'north!)

chris661 29th October 2019 09:41 AM

I'd start with encoding into M/S, and then pushing up the side by 3dB and see how that goes.

That'd amplify any L/R differences your mic array has already picked up. If I want to widen a stereo sound, that's my first move. I'm certain that DeeDee's methods will work too, but I like to try the simple stuff first.

There are plenty of articles online about mid-side processing. Here's one: http://www.masteringhouse.com/master...s/midside.html


I like it up North - rolling hills of green are where I feel at home, so it's quite fortunate that I'm a stone's throw from the Peak District.

Chris

deedeeyeah 29th October 2019 09:58 AM

Quote:

Originally Posted by chris661 (Post 14291501)
I'd start with encoding into M/S, and then pushing up the side by 3dB and see how that goes.

That'd amplify any L/R differences your mic array has already picked up (...)

m/s encoding may indeed do the job; possibly less well if the seating was very wide/the mics were close/deep into the ensemble - not sure i'd bring the sides up though, maybe even bring them down but apply some image enhancing/widening tools (one of which are said apf's).

Quote:

Originally Posted by Geoff Poulton (Post 14291159)
I'll spend some time looking at this (already have done) and trying to grasp the whole of what you're saying!

no voodoo, there's quite a bit of literature available on arraying of subwoofers, array processing and alignment of arrays - apf's are mostly the last resort if things cannot get achieved by positioning (and 'bad' positioning is the basis of the analogy i made).

Folkie 29th October 2019 03:03 PM

For MS Decoding, Encoding, or in your case
“Inline” mode I would try Voxengo’s free MSED plugin. Very easy to use. As I recall “Inline” mode
is the default setting ( button on left side of GUI).
https://www.voxengo.com/product/msed/opinions/

Geoff Poulton 30th October 2019 05:52 PM

Quote:

Originally Posted by Folkie (Post 14291956)
For MS Decoding, Encoding, or in your case
“Inline” mode I would try Voxengo’s free MSED plugin. Very easy to use. As I recall “Inline” mode
is the default setting ( button on left side of GUI).
https://www.voxengo.com/product/msed/opinions/


Thank you for letting me know about this, I'm not sure how I'll get on with it as I'm running Sadie, but I'll give it a shot! Some other Vexengo software has functioned properly on download.

Geoff Poulton 10th November 2019 09:57 PM

Quote:

Originally Posted by chris661 (Post 14290937)
Any chance of processing it as M/S and pushing the -side channel up?

Probably the simplest way to do it, but I'd be interested to try the other methods.

Chris

Hi Chris,

I'm down at the south end of the Peak District and when I travel north I to Manchester etc., I often give the M6 a miss and travel up through the Peak District.

When you convert your LR stereo into M&S, how do you go about accomplishing that. I used M&S sometimes on occasion, but getting LR to M&S is a bit of a brain teaser.

Folkie 10th November 2019 10:20 PM

Quote:

Originally Posted by Geoff Poulton (Post 14315045)
Hi Chris,

I'm down at the south end of the Peak District and when I travel north I to Manchester etc., I often give the M6 a miss and travel up through the Peak District.

When you convert your LR stereo into M&S, how do you go about accomplishing that. I used M&S sometimes on occasion, but getting LR to M&S is a bit of a brain teaser.

The Voxengo MSED plugin I recommended in my previous post can do this as well.
i.e. “Encode” (LR to MS) in addition to “Decode” (MS to LR) and “Inline” (LR to MS-allowing manipulation of stereo width, etc and then back to
LR).
If you use this plugin in “Inline” mode as an insert on
any LR stereo channel you can easily adjust the
stereo width and still have a LR stereo channel.

chris661 11th November 2019 06:02 PM

Quote:

Originally Posted by Geoff Poulton (Post 14315045)
Hi Chris,

I'm down at the south end of the Peak District and when I travel north I to Manchester etc., I often give the M6 a miss and travel up through the Peak District.

When you convert your LR stereo into M&S, how do you go about accomplishing that. I used M&S sometimes on occasion, but getting LR to M&S is a bit of a brain teaser.

Nice route to take. Some lovely pubs on the way, too. Good food.


Mid = L + R
Side = L - R

You can do it manually using something as basic as Audacity.

Chris