Become a Site Supporter and Never see Ads again!

Author Topic: Oddball microphone technique (OMT) - part 2  (Read 8500 times)

0 Members and 1 Guest are viewing this topic.

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12298
  • Gender: Male
Re: Oddball microphone technique (OMT) - part 2
« Reply #135 on: August 03, 2018, 12:58:03 PM »
Hey Kyle, I basically pulled a few setups out of the previous illustrated OMT booklet to make the PDF I just posted above.  I did simplify the images and change the wording a bit, but in essence the info there isn't really any different than what I'd already posted.   It mostly puts these setups using sideways-facing Mid/Side arrangements together and identifies that as a category. 

And as always, thanks for your interest and insights!
« Last Edit: August 03, 2018, 03:51:51 PM by Gutbucket »
volition > vibrations > voltages > numeric values | numeric values > voltages > vibrations> virtual teleportation time-machine experience
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: 12298
  • Gender: Male
Re: Oddball microphone technique (OMT) - part 2
« Reply #136 on: August 03, 2018, 02:23:32 PM »
Sorry to veer off course a bit here, but I've been thinking about four channel setups like a pair of spread omnis with a pair of directional mics in the middle.  What I'm wondering about is the relationship between the SRA of each pair to the other.  So, using that Sengpiel Audio visualizer tool, if I have a pair of omnis 60 cm apart then that gives me a SRA of 118.1 degrees.  Assuming that's the right SRA for the specific location/circumstances, should the directional pair's SRA ideally be the same as that of the omni pair (assuming something like mid/side that can be adjusted in post isn't being used)?  If the SRAs are significantly different, won't the blend of mics result in a weird stereo image?

That's one way of approaching it, and partly informed my thought process along with other ideas when working up these setups initially.  But I now don't think it applies in any straightforward way.  And think the opposite is what actually works best-  very different SRAs from the omni and center pairs

The SRA curves are empirically derived from listeners with 2-channel recordings made using various combinations of microphone spacing and angle.  It's an attempt to derive objective conclusions from subjective psychoacoustics, addressing situations where only a single pair of microphones is used rather than more than one pair.  Even in Michael William's extension of this approach to multi-channel surround recording, each sector around the playback array is always covered by no more than a single microphone pair, the intent being to have each sector line up with the adjacent sectors on either side without a gap or excessive overlap. It specifically does not address content outside the SRA or "cross-talk" from the microphones of non-adjacent sectors, which is related to what you are asking about. 

OMT differs by relying on the overlap and intentionally managing it.  It pushes the omnis wider than one would probably want without the center mic or pair to intentionally tend toward a hole-in-the-middle which the center mic or pair then fills with partial overlap. Yes that omni spacing may or may not produce a seemingly appropriate SRA angle by charts, yet SRA imaging only weakly applies to non-uniformly distributed omni pair imaging (hence the middle-hole thing), and the center pair is typically dialed in such that it provides less overall width than one would want without combination with the omnis.

I think that explains the essence of what's going on, but stated another way from experience..

With wide-spaced omnis + a center Mid/Side pair, I always end up dialing in far less stereo width from the center pair than I would without the omnis.  Say I'd choose a 50/50 Mid/Side ratio as being best when just using center Mid/Side pair on its own.  When combining with the wide omnis I'm likely to dial that back to something like a 80/20 or 90/10 ratio of Mid/Side.  If the same SRA was desirable with the omni pair and center pair, there would be no reason dial back the center pair with a more Mid-heavy ratio when combined with the omnis.

If you go to the Sengpiel Audio visualizer and choose a 60cm omni spacing the SRA is 118 degrees.  Switching to X/Y supercards it takes a 107 degree angle between microphones to produce an equivalent 118 degree SRA.   That's a combination I'm quite likely to use for an X/Y pair on its own, but not one I'd be likely to decide is best in combination with the omnis.

By contrast, here is a combination I'd be likely to use- A pair of 1.5m spaced omnis (around the spacing I typically use for OMT) has an SRA of around 40 degrees total.  A pair of X/Y supercards with a narrow 20-degree angle between them produces a super-wide SRA of 220 degrees or so.   

Empirically I've found it works best when the SRAs of the omnis and center pair are sort of tending towards opposite opposite extremes rather than close to the same.  It's probably another case of achieving good differentiation between pairs so they do not "step on each others toes", rather than trying to achieve similar SRAs across both pairs. The imaging of the omnis positioning discernable sources widely near or outside the speaker locations, while the imaging from the X/Y pair positions discernable sources more narrowly across the center between the speakers.
« Last Edit: August 03, 2018, 04:06:25 PM by Gutbucket »
volition > vibrations > voltages > numeric values | numeric values > voltages > vibrations> virtual teleportation time-machine experience
made easy- >>Improved PAS table<< | made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline heathen

  • Trade Count: (15)
  • Taperssection All-Star
  • ****
  • Posts: 1333
Re: Oddball microphone technique (OMT) - part 2
« Reply #137 on: August 03, 2018, 02:40:48 PM »
Thanks for the explanation.  I may be revising my plans for the show I'm going to tonight based on this information.
Mics: Core Sound TetraMic | AT4031s | AT AE5100s | AT853s (C/SC) | Line Audio CM3s | DPA 4061s | CA-14 omnis | Studio Projects CS5
Pre: CA9200
Decks: Zoom F8 | Roland R-05

Offline heathen

  • Trade Count: (15)
  • Taperssection All-Star
  • ****
  • Posts: 1333
Re: Oddball microphone technique (OMT) - part 2
« Reply #138 on: August 04, 2018, 04:17:26 PM »
Nothing terribly innovative here, but I'd like to share my reasoning for the setup (and find out how wrong my thinking may have been): http://taperssection.com/index.php?topic=187418.0

I've got DPA 4061s on a TV antenna, spread about five feet.  Nothing much to say about that, other than that it was heavily influenced by Gutbucket's recent comments about my SRA question.

Roughly in the middle I've got the pair of Audio-Technica AE5100s.  I've tried XY from this same location, between split omnis (for reference: https://archive.org/details/garajmahal2018-05-11.AT4031.CA14), so I wanted to try something different.  I've got a DIN/ORTF combo mount for the AT4031s that I can squeeze the AE5100s into, but I wanted to configure the AE5100s in a way that would maximize sound from the stage hitting the mics on axis.  Playing around with the Sengpiel visualizer, I ballparked that if I had the AE5100s AB about 20 cm apart, and at around a 75* angle between them, I could get an SRA that should nicely cover the stage from that distance.  My thinking was that this would give me a pretty solid center "image" to compliment the wide split of the omnis, without making the center too smooshed like XY might.  My goal was basically keeping a fairly wide soundstage once it's all mixed together.  The Rode bar I used allowed me to get the mic spacing pretty much dead on, but I had to eyeball the angle.

I haven't mixed it all together yet, but on an initial listen I think it sounds pretty good.  The DPAs will likely need some EQ to keep the bass from getting too flabby when both sources are combined, but even without EQ I'm happy with the sound.  One thing that's particularly gratifying is that both sources sound good on their own, but mixing them together is definitely an improvement.

To be clear, I'm not trying to say that I've done anything new here.  Nor am I trying to say that this was the ideal setup for this spot.  It will be interesting to compare the Garaj Mahal show with this one, though, mostly to compare the results in terms of soundstage.  As always, I appreciate any comments and criticisms (though I'm sure it will be much more helpful once I actually post the recording!).

Is there any interest in me posting a raw sample of each source so people can do their own mixing/EQ?  I'm happy to do so if anyone wants.
Mics: Core Sound TetraMic | AT4031s | AT AE5100s | AT853s (C/SC) | Line Audio CM3s | DPA 4061s | CA-14 omnis | Studio Projects CS5
Pre: CA9200
Decks: Zoom F8 | Roland R-05

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12298
  • Gender: Male
Re: Oddball microphone technique (OMT) - part 2
« Reply #139 on: August 06, 2018, 06:33:42 PM »
Should be interesting comparing your recordings made with these two similar setups using the same microphones- the first with an relatively-wide angled X/Y directional center pair and the second with a near-spaced PAS directional center pair.  Are all other variables other than the band and date generally the same?  (Venue, recording location, omnis used and their spacing, etc?)


..I wanted to configure the AE5100s in a way that would maximize sound from the stage hitting the mics on axis

That is a very appropriate fundamental goal for the center pair IMO, and there are several ways of achieving it.  Each presents different implications-

Your method of playing around with the Sengpiel visualizer variables to find a near-spaced combination of pattern/angle/spacing that points the microphones directly at the PA, is one of them.  This is the essence behind the Improved PAS technique of which you are aware, which consists of a table indicating the most appropriate spacing between microphones based on the Point-At-Stacks angle between them, such that the resulting SRA angle equals the PAS angle.  The table simplifies the process by not requiring online access to the visualizer and not requiring playing with the variables to find the solution each time.   The Improved PAS technique and Sengpiel visualizer data upon which it is based assumes we want an SRA which is equal to the orchestra angle / PAS angle.  As I mentioned in my previous post above , that's almost always appropriate for a two channel stereo microphone arrangement, yet may or may not be optimal for a stereo main microphone arrangement built upon the combination of of more than two pairs..

Another is to run the center pair as a coincident X/Y arrangement, except using a narrower mic angle than you'd typically want if the X/Y stereo recording was intended to be used on its own.  In that case the X/Y angle is equal to the PAS angle and the microphones are pointed directly at each PA.  SRA will be much wider and discrete imaging across the center will be tight and more compact.  I'd suggest trying this arrangement for your next recording at this same venue, keeping the other variables unchanged.  You'll then have all three setups to compare against one another.

Wide-angled X/Y center pair (appropriate microphone angle and SRA for an X/Y pair alone) with mics not pointed directly at the PA.
Near-spaced PAS center pair (appropriate spacing/angle and SRA for a spaced-pair alone) with the mic pointed directly at the PA.
Narrow-angled X/Y center pair (less appropriate microphone angle and SRA for an X/Y pair alone, but likely good with the omnis) with the mic pointed directly at the PA.

At that point you can try something interesting while comparing the three recordings.  Besides comparing each as mixed in a straightforward fashion, try inserting a stereo-width adjustment tool in the signal path of the center pair.  You can then play with making the center pair contribution wider or narrower (all the way down to a monophonic center) while listening in combination with the wide omnis.  Although this kind of stereo-width adjustment is intended for coincident-pair arrangements such as Mid/Side and X/Y, it can also be applied to your near-spaced pair.  The range of adjustment before comb filtering problems become audible will be more limited with the near-spaced pair (listen for it by muting the omnis and dialing the width of the near-spaced pair all the way down to mono). 

We've discussed in the past how to do this using two back-to-back Mid/Side matricies (L/R>M/S>[ratio adjustment]>L/R) and how its easier to use a single instance of a stereo width adjustment in the DAW software (most typically accessed through the panning control for a stereo channel) or a Mid/Side based stereo-width plugin, both of which do the same thing.  I can explain that in more detail again if you like.  Here's good free VST plugin with this capability - https://www.voxengo.com/product/msed/

A comparison made along with center width-adjustment will be illustrative in a couple ways- First, you are quite likely to find in an altered-width setting that works better than a straight mix of the two pairs, and this will be a good starting point setting for other recordings made using the same setup.  Second, you can better compare the different center mic-pair arrangements against each other after each has been tweaked to achieve its own optimal center width and blend with the omnis.  And third, you can mute the omnis to find how non-optimal that setting is for the center pair on its own without the omnis.

Quote
My thinking was that this would give me a pretty solid center "image" to compliment the wide split of the omnis, without making the center too smooshed like XY might.  My goal was basically keeping a fairly wide soundstage once it's all mixed together.

How smooshed or stretched is best is a large part of what you will be determining with this.  In my experience, it's much better to have the center pair contribution smooshed rather than stretched or even optimally-wide (for use on its own), since the omnis will be stretching out the center and usually don't need help in conveying additional width.  Consider the simplified setup of a single forward-facing center microphone between the two wide omnis, in which there is no stereo width provided by the center microphone at all.  It works, quite often far better than a pair of omnis alone, but can usually be improved by introducing some stereo width to the center.  The question is how much center width is most appropriate?  The answer IME is always "somewhere in between a single mono center microphone and an X/Y pair optimized for good stereo width on its own.  Mono is too narrow, optimized X/Y on its own, too wide.  The same applies to near-spaced center configs.

The bummer is not having either the omnis nor center pair fully optimized for use on their own, but that's the price to pay for achieving something superior than could be achieved without both pairs in combination.  That said, there is nothing wrong with preferring to optimize each pair on its own, which provides better redundancy should one pair fail, rather than pursuing the most optimized combination which relies on both working.

Quote
One thing that's particularly gratifying is that both sources sound good on their own, but mixing them together is definitely an improvement.


That's a good indication of being on the right track.  The trade-off I mention above is between how much of an improvement that represents, when weighed against further improvements that push each source towards not being as good on its own without the other. 

My current thinking on this is to always have a the directional center microphone or pair pointed directly at the source, to use a coincident center arrangement for that pair to minimize phase interaction problems, and to adjust the stereo width of the center contribution afterwards as appropriate.   It may seem contradictory that I'm using a near-spaced 3-microphone arrangement in the center of my 6-channel OMT setup, but the difference is that even though this 3-mic arrangement is near-spaced, it provides a direct source-pointed hard-center channel and is specifically designed to minimize phase interaction problems between the 3 channels.

Quote
Is there any interest in me posting a raw sample of each source so people can do their own mixing/EQ?  I'm happy to do so if anyone wants.

I'd like to play with samples of these two recordings (or three if you decide to try the suggested PAS X/Y variation as well) to help check my thinking on all this using someone else's recordings other than my own.  It makes for a nice opportunity when the there is only one significant variable which changes between recordings (other than the band).
« Last Edit: August 07, 2018, 09:32:10 AM by Gutbucket »
volition > vibrations > voltages > numeric values | numeric values > voltages > vibrations> virtual teleportation time-machine experience
made easy- >>Improved PAS table<< | made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline heathen

  • Trade Count: (15)
  • Taperssection All-Star
  • ****
  • Posts: 1333
Re: Oddball microphone technique (OMT) - part 2
« Reply #140 on: August 06, 2018, 08:08:20 PM »
Should be interesting comparing your recordings made with these two similar setups using the same microphones- the first with an relatively-wide angled X/Y directional center pair and the second with a near-spaced PAS directional center pair.  Are all other variables other than the band and date generally the same?  (Venue, recording location, omnis used and their spacing, etc?)

The venue and recording location are the same, but the mics are different.  The Garaj Mahal show was the XY AT4031s with split CA14 omnis, whereas this Breakfast show was AB AT AE5100s with split DPA 4061s.
Mics: Core Sound TetraMic | AT4031s | AT AE5100s | AT853s (C/SC) | Line Audio CM3s | DPA 4061s | CA-14 omnis | Studio Projects CS5
Pre: CA9200
Decks: Zoom F8 | Roland R-05

Offline heathen

  • Trade Count: (15)
  • Taperssection All-Star
  • ****
  • Posts: 1333
Re: Oddball microphone technique (OMT) - part 2
« Reply #141 on: August 06, 2018, 09:21:55 PM »
Quote
Is there any interest in me posting a raw sample of each source so people can do their own mixing/EQ?  I'm happy to do so if anyone wants.

I'd like to play with samples of these two recordings (or three if you decide to try the suggested PAS X/Y variation as well) to help check my thinking on all this using someone else's recordings other than my own.  It makes for a nice opportunity when the there is only one significant variable which changes between recordings (other than the band).

You've given me a lot to think about, as usual.  In the meantime, here's a link to download a sample of the raw sources on their own (link good for seven days): https://we.tl/89Rd74EsX5

I'll confess that part of my desire to make sure each source sounds good on its own is to continue my evaluation of the AE5100 because there aren't a lot of tapers out there using it.  I may even post the AE5100 source separate from the final four-mic mix.  I think this was only my second outing with the AE5100s, and so far I continue to think they have a lot of promise.

I would love to hear what anyone does with the samples!
Mics: Core Sound TetraMic | AT4031s | AT AE5100s | AT853s (C/SC) | Line Audio CM3s | DPA 4061s | CA-14 omnis | Studio Projects CS5
Pre: CA9200
Decks: Zoom F8 | Roland R-05

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (13)
  • Needs to get out more...
  • *****
  • Posts: 12298
  • Gender: Male
Re: Oddball microphone technique (OMT) - part 2
« Reply #142 on: August 07, 2018, 10:03:21 AM »
Thanks, I may have some time in the evenings this week before I head out of town.

I'll also say that my listening preference has shifted somewhat over the years, which is inevitably influencing my comments above.  I've always loved detail, clarity, stereo width and immersion.  These days I find I also really value a strong, solid, well-anchored center as a vital foundational element that conveys realism and naturalness.  Sort of the kernel from which everything extends and emerges except for ambience and audience reaction - best when not obviously identifiable in itself, except upon its absence or when contrasted against recordings which do not have that quality.
volition > vibrations > voltages > numeric values | numeric values > voltages > vibrations> virtual teleportation time-machine experience
made easy- >>Improved PAS table<< | made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

 

RSS | Mobile
Page created in 0.222 seconds with 35 queries.
© 2002-2018 Taperssection.com
Powered by SMF