Topic: Mixing 5.1 surround in Wavelab

[goodcooker]

So just for shits and giggles I want to start mixing some surround mixes. Curious if anyone is doing this and can point me in the direction of some samples to show how different techniques produce results.

I have enough mics and gear to accomplish it - possibly on a single stand - and don't want to go too far down a rabbit hole with the possibilities if it involves hauling really wide split bars, lots of mounting gear or taking up a lot of space. I'd like to keep everything on one stand and not have a giant array flying. I found that in the last few years more than ever people (staff and patrons) are not fond of mic setups that take up too much space.

I'm thinking something along the lines of this with everything going into a DR680 - external preamps and associated power supplies optional depending on the size of the bag

Line Audio CM3 (wide card) > Wmod UA5 = front LR
AKG C522XY (XY) > Aerco = center
single (or a pair - curious on thoughts here) Senn e614 (hyper) = rear (could use the CA11 cards here too to minimize size and weight)

Sometimes I get soundboard feeds so that could come into play I'm guessing likely as the front LR or center - I've not had to do it but the DR680 can take the digital input and record it onto the stereo mix track giving enough tracks for all this at once.

I'm looking at this for smallish gigs where I could be on the stage, stage lip or at least very close to the performers.

[rigpimp]

I have not done it yet but I have a pair of mk8 and a pair of mk5 so I had Scott 3D print me a Schoeps active double MS mount.  I have a plan to dabble some using the free Schoeps double MS plugin and can report back once I've done it.  I think that with the plugin the workflow will be a bit different than with others.  I know that Marcus uses an ambisonic mic I saw him use at JRAD a couple years ago.  He sent me his file set and it's everything you'd want to create ANY polar patter in post.

[goodcooker]

I have not done it yet but I have a pair of mk8 and a pair of mk5 so I had Scott 3D print me a Schoeps active double MS mount.  I have a plan to dabble some using the free Schoeps double MS plugin and can report back once I've done it.  I think that with the plugin the workflow will be a bit different than with others.  I know that Marcus uses an ambisonic mic I saw him use at JRAD a couple years ago.  He sent me his file set and it's everything you'd want to create ANY polar patter in post.

I am considering getting a Sennheiser Ambeo mic now that they can be had for a decent price used.

[Gutbucket]

I do surround. Will try and keep my input short and succinct and hope it is welcome.

Given your description above, think quad. Best to ignore the center channel given your constraints, or maybe just put SBD in there.  Try adding a rear-facing pair to whatever 2-channel arrangement you are currently happy with using for stereo. The rear facing pair needs to be pretty directional, say supers rather than cards, and pointed more or less directly away from the PA/stage to minimize direct pickup of it much as practical into this pair feeding the surround channels.  This may provide just what you are looking for.

Beyond that, if like the sound of good coincident microphone stereo recordings (X/Y and M/S), a single point ambisonic microphone is likely to best fit the requirements you describe. Small, compact and very cool that it doesn't matter how you point it, only where you place it since you choose direction, pattern and inclusive angle afterward.  You do need to take the time afterward to dial in those choices while listening, but that part is fun informative and rewarding.  It won't produce the ultimate surround experience, and still best to think of them in terms of quad, but simply adding a front/back dimension to the stereo listening experience is a huge change which when done right is particularly well suited to live music recordings.

I've not used it but the Senn Ambio is likely to be a good fit for you.  I have a TetraMic which is in this same general category and learned a lot using it, although I haven't used it in a while and should to get it recalibrated at this point.  Beyond that, something like a Double Mid/Side setup is essentially the same thing done with 3 individual microphones, but is less compact, less convenient, requires a fig-8, a closely matched response from all 3 mics, and gives up vertical angle adjustment.  More tweaky.  A dedicated ambisonic mic is a lot simpler.  Dual/M/S makes good sense for rigpimp though since he has the Schoeps mics and can use the Schoeps D/M/S plugin designed for them.  He only needs a 3-d printed mount and 3 recording channels to do it.

[Gutbucket]

Why quad and not 5 channels?

All of these coincident ambisonic arrangements can produce surround by deriving virtual microphone channels pointing backward. There is no physical limit to the number of channels you can derive originating from the single point in space where the microphone was placed, pointing in whichever direction you like (horizontal plane only for D/M/S) .. but there is a practical limitation which is that the more channels you derive pointing in different directions, the tighter the pattern of each needs to be, and the pattern can only be at tight as a super/hypercardioid. The take away of this is that a 1st order ambisonic microphone (recorded to 4 channels), or a D/M/S setup (3 channels) does a good job at producing four virtual outputs across the horizontal plane.  Five is pushing it.  But four in quad can be enough. I think of 1st order coincident ambisonic arrangements as being the ultimate 2-channel coincident stereo microphones, that a very convenient at producing passable surround. 

A higher order ambisonic microphone allows for tighter patterns, making for finer slices, which translates to more output channels that have enough differentiation between them to work effectively from a single point in space, yet requires twice as many or more recording channels.  If serious about recording live music in surround while remaining constrained to a single-point coincident microphone rather than a spaced array, something like the OctaMic which requires 8 recording channels might the answer.

Easiest with existing microphones and capable of producing superior results in my opinion, are near-spaced multichannel microphone configurations, which are in essence extensions of typical 2-channel near-spaced configs.  Using your currently preferred standard 2-channel near-spaced configs works well for the to two front channels of quad without a center channel. The same basic issue as before applies here - the greater the number of output channels, the tighter the microphone patterns need to be.  The difference this time is that even though we don't have patterns tighter than super/hyper available to us, we can use more spacing between the microphones to compensate.  But that works against keeping the microphone array compact.  For instance, for playback across 3 front speakers using a center channel (as in typical 5 channel surround) the microphone array generally needs to be twice as wide as a 2-channel stereo near-spaced microphone configuration for playback over two channels.  This  doesn't really fit with your requirements.  However, if a total L/R spacing of 2' total is acceptable, that can be enough and fits comfortably on a single stand.  If not its probably best to stick with quad and ignore the center channel.

Generally I find level difference is the important thing differentiating the front/back axis. Microphone pattern and angle achieve that rather than spacing. Timing differences help greatly across the left/right axis, particularly for playback of over more than 2 speakers in front.   This tends to leads to microphone arrays that are shallow front/back yet wider left/right.

[Gutbucket]

Line Audio CM3 (wide card) > Wmod UA5 = front LR
AKG C522XY (XY) > Aerco = center
single (or a pair - curious on thoughts here) Senn e614 (hyper) = rear (could use the CA11 cards here too to minimize size and weight)

I think that can work nicely. Since the pattern of the CM3 is wide AND there will be an X/Y pair or single microphone in the middle between the them, it will help to angle the CM3 pair very widely, even directly sideways in order to achieve better differentiation in L/C/R speaker playback.  The wider you can space them the less widely you need angle them. This very wide angle between the pair will be less of a problem when onstage or at stage lip than it would be from a position farther back in the room.   That X/Y pair in the center will be most useful in a 2-channel mix.  It can be summed to mono to feed the center speaker but then the combined pattern becomes much wider than either microphone of the X/Y pair on its own, so if 5 channel surround is the primary intention and your ability to space the array is limited, it may be best to use just a single microphone in the center.  One reason I like using a M/S pair in the center is that the Mid channel retains the same pattern whether or not the Side fi-8 used. 

A stereo pair facing backwards is best, but a single channel mult'd to all surround channels works quite well.

Sometimes I get soundboard feeds so that could come into play I'm guessing likely as the front LR or center - I've not had to do it but the DR680 can take the digital input and record it onto the stereo mix track giving enough tracks for all this at once.

I found early on that surround playback works best when there is some audience/ambience in both the front and rear speakers, possibly just side and back speakers if playing back over 7 speakers.  Otherwise there is a audible spatial disconnect between the front and back. Good immersion requires a more even distribution of the ambient reverberant around the room.  The center speaker can be all dry SBD though.

Can confirm that using the digital input works reliably for recording 2 additional channels for 8 total into the DR-680.

[goodcooker]

Gutbucket I was hoping you would join in this discussion since a lot of the things that are an unknown to me are something you have done a lot of.

After I've had some time to think this through a little more I'm sure I'll have a bunch of questions.

Couple things - I've been using my main pair much as you describe with a wider spread and less included angle than the typical 90 degree near coincident patterns but I tend to stick to less than 30cm depending on how far away from the source I am.
Closer to source= less mic spacing more angle
Farther from source= more spacing less angle.

Not real big on coincident patterns in the past but recently picked up the AKG C522 to try something different. The capsules in this mic are not true XY and have a space of about one inch between the center of the caps. So the thought of using it in the center if I didn't have a board feed to put there made the most sense to mix down to mono but having a single mic there is already mono so less work. But if I wanted to bite off a bigger chunk of mounting gear/post work the capsules can be adjusted mechanically to 180 degrees so it would be easy to add a cardioid or hyper and do MS for the center. Or turn it around and use it at 90 or more for the rear channels. Seems like XY in the middle would make me want to push the spread of the LR pair farther apart.

Center mic - say there's a single hyper in the center (what I would usually add to the wide card spaced pair). I would usually move that mic forward by several inches. Not sure why it fits in the mix and sounds better to me but in practice it does. Any thoughts on the line up of the center mic in this scenario in relation to the spaced pair?

[Gutbucket]

I think the discussion of surround in 5ch, 7ch or whatever format is really two separate discussions.  I'm ignoring the .1 which is a whole different discussion and can of worms best ignored with music.  What is going on in front is one part, and what is going on in back, or rather in the entire room is really a completely separate thing.  The first is all about recording for playback over 3 verses 2 speakers across the front - all about the direct sound arrival, stereo imaging, etc.  The other discussion is the surround part, what is happening diffusely in the room, the mostly indirect arriving ambient/audience stuff.

So the front first-
Its certainly simplest to operate as one is already doing with 2-channel stereo in front, which corresponds pretty much directly as the front half of quad.  This also helps keep the mic array from growing overly wide.

With three mics feeding three front speakers, the L/R mic pair benefits from being angled more widely than one would otherwise do when using just two microphones or using three microphones feeding two speakers.  The L/R pair also benefits from being angled more widely if the spacing can't otherwise be made very wide.  The problem is that sideways facing L/R mics tend to work best outdoors, in a really good sounding room, or up close where the source is close and wide. 

For 2-speaker playback I like the L/R pair angled less and spaced more widely, as that keeps the L/C/R mics more on-axis with the source.  With 3-channel playback a wider angle between mics better suits the speaker arrangement, but frequently doesn't fit the recording situation as well. So its something of a conundrum. 

I find the question of L/R on/off-axis mic angle, along with the choice between a single microphone verses a coincident stereo pair in the center are the two main things involved with how mic'ing effects playback over three verses two speakers in front, regardless of what is going on with the surround channels.

Couple things - I've been using my main pair much as you describe with a wider spread and less included angle than the typical 90 degree near coincident patterns but I tend to stick to less than 30cm depending on how far away from the source I am.

You can keep things the same as you are currently doing for the front of quad.  You can of course also try playback over 3 front speakers using the recording methods your are already using for 2 channel stereo, such as routing the center hypercardioid you've mention sometimes using to the center speaker, and/or routing mono SBD to the center speaker, without changing anything.  Always best to try these things, hear it and decide for yourself what's working well and what isn't, or rather, if it works well enough to make it worthwhile.  Using more L/R spacing with a center channel microphone is simply a more optimized arrangement, but you may find it works well enough without increasing the spacing you are currently using.

[Gutbucket]

Seems like XY in the middle would make me want to push the spread of the LR pair farther apart.

-and-

Center mic - say there's a single hyper in the center (what I would usually add to the wide card spaced pair). I would usually move that mic forward by several inches. Not sure why it fits in the mix and sounds better to me but in practice it does. Any thoughts on the line up of the center mic in this scenario in relation to the spaced pair?

Yes, same here.  I like having the center microphone or coincident pair placed a bit forward of the others.  In the past I've mostly just sort of done that by feel, instinct and what I hear from the results, but lately over the past year or so I've been slowly working on a 3-point version of the 2-channel Improved PAS technique where the ideal spacing between microphones is determined by the PAS angle between them, and that process has provided a bit more insight into the 3rd center microphone position.  The improved PAS thing is based on the fundamental principle you describe:

"Closer to source= less mic spacing more angle
Farther from source= more spacing less angle"

With three microphone positions (using 3 or 4 mics, four if the center position is a coincident pair) the same fundamental principle applies, but the spacing between L and R positions grows to something like twice as wide as it would be in 2-channel Improved PAS due to the addition of the center microphone or pair.  This is unfortunately not conducive to a compact microphone arrangement, and is one of the reasons I think quad surround without a center speaker is likely to fit better with your constraints than 5-channel. The reason I mention all this is that in working through the process of deriving a 3-microphone position PAS table which tapers can easily refer to in order to determine what the 3-point PAS spacing should be, I run a whole bunch of configurations through the Schoeps Image Assistant and the Sengpiel Audio visualizers and in all cases the center microphone position is always somewhat forward of the spaced L/R pair.  Exactly how far forward varies somewhat, but most of the time everything lines up best in those visualization apps with the center microphone positioned about 20cm forward of the L/R pair, so that's what I've been recently suggesting to folks asking about forward spacing of a center mic or pair.  That's about 7-8".  In my current setup the center mic isn't quite that far forward, but I do have things spaced more widely in the L/R dimension. In general it seems good to push the forward spacing out a bit more (closer to 20cm) when the L/R microphone positions are not spaced as widely. When I redo things next I'm going to try to push the center position out to 20cm.

[Gutbucket]

Perhaps ironic to be concentrating on the discussion of microphone arrangements in your thread titled "Mixing 5.1.." But recording for surround is mostly about the microphone arrangement and playback sides of things. There is a lot less, if any actual mixing of channels together going on for multichannel playback in comparison to 2-channel stereo, as the microphone channels are generally routed to their own individual speakers rather than being summed together in the stereo buss.  Still need to balance levels between channels of course, and EQ, compression, and all that jazz still applies, although some if that stuff can be approached differently.

[Gutbucket]

.. if I wanted to bite off a bigger chunk of mounting gear/post work the capsules can be adjusted mechanically to 180 degrees so it would be easy to add a cardioid or hyper and do MS for the center. Or turn it around and use it at 90 or more for the rear channels.

The X/Y facing backwards might work well, but don't use much X/Y angle so as to keep both mics of the pair facing more or less directly away from the front, attenuating pickup of direct sound arriving from the front as much as possible.  A single rear-facing channel feeding all surround speakers works pretty well, and the addition of even just a touch of stereo difference should improve on that.  Limiting the pickup of direct-arriving sound originating from the PA and stage is more important in a rear facing pair than stereo width.

Cool idea to re-purpose that stereo X/Y mic as a pseudo fig-8 for potential Side channel use.  You may be aware all of this already, but if not.. to set that up to best effect, orient both capsules of the mic directly forward, invert polarity of one channel, sum their outputs together, and while listening carefully adjust their input trims (or fader levels) in search of the deepest cancellation null you can achieve.  Best result will be likely be found outside (as that's a pseudo anechoic space) using a sound source placed in front of the microphone directly on-axis with the null-pane.   Then to use the stereo mic as a mono fig-8, rotate the capsules 180 degrees apart, but leave the carefully set trims or faders where they previously were and record both channels. Afterward, invert polarity on the right channel before summing the two to form a single mono fig-8 Side channel. The next time you want to record using that mic as a fig-8 you'll need to be able to return to those exact trim/fader settings again.  This procedure should produce the best fig-8 pattern able to be achieved with that particular microphone.  The inch or so spacing will work against achieving a smooth fig-8 pattern off the null plane at high frequenceis, but when used as the Side channel of the Center M/S pair in a stereo mix in combination with your other L/R pair, it may be good enough as long as the Mid mic is centered and closely aligned with it in the same vertical plane.  I only use some of that Side channel in the stereo mix, and don't use it at all for surround where the individual L/C/R mics are routed directly to their own speakers, but I very much like what it does in a stereo mix where it often serves as sort of the secret sauce.

[goodcooker]

The AKG C522 is a weird mic. It has two cardioid capsules that are mounted in a "pointed forward" end address arrangement with about an inch separating the center of the capsule diaphragms. There's a screw on the outside of the housing on each side that can adjust the capsules mechanically from an angle of 90 degrees in detented stops at 120 and 150 to 180 degrees independently of each other.

The idea of using it for MS would work but there is more spacing between the diaphragms than a more typical LD figure eight mic that I would normally use. Doable though. Just mount a forward facing mic (I'd probably use a card or hyper) right on top of the stereo mic in line with the diaphragms after being set to 180. Record 3 channels then polarity invert as necessary in post.

After setting this up on a stand at home I think the most I'm really willing to invest on this experiment at a gig is flying four mics. I'd probably go with two pairs - one with 3 mics facing forward and one rear or 2 pairs - one forward one rear. I may do this at an upcoming gig outdoors for Robert Randolph. May not work since at high volume there is some slapback from the rear which is very perceptible in the venue but not with forward facing directional mics from my usual location. Doesn't make sense to haul extra gear and get my expectations up only to realize I'm just accentuating the shortcomings of the acoustic space.

[Gutbucket]

Tried looking up the C22 online and everything seems to point to the AKG Aria, without any specific mention of C22 (although some retail links say "AKG Aria C22").  From what I can tell, the Aria looks to be a USB podcasting microphone with individual fixed front and rear facing cardioid capsules, along with a combined mode for omni pickup.  That seems different from what you describe, but might still be able to be usage-hacked as discussed above to serve as a fig-8 Side mic.  Don't wish to chase a wild goose too far, but willing to discuss this potential option further if you like.  If so probably best if you can post a photo or two of the thing and its capsule arrangement.

If wanting to stick with four total microphone channels and at least one of those is going to be rear-facing, I think it best to forgo Mid/Side for now anyway.  If you were recording 5 or 6 channels total, with 4 "forward-facing" channels, then yeah, Mid/Side in the middle is great.  But good options for 4 "forward-facing" channels using a mid/side pair doesn't leave any channels available for rear-facing mics.  The partial exception is spaced omnis with a Mid/Side (or X/Y) pair in the center.  In that case you can use the omnis to feed the surround channels. The only problem with that arrangement in amplified concert situations is the omnis really pick up too much front-arriving direct sound from the stage and PA.  Because of that, I find I can't increase the level of the surround channels as much as would otherwise be desirable without the transient stuff from the front getting pulled too strongly into the surrounds. Works pretty well for non-amplified performances though.  Other than that, using spaced omnis to feed the surround channels works really well - big, lush, open, spacious enveloping surround.  That caveat gets back to the thing about needing sufficient front rejection in the rear-facing microphone channels for them to be more useful - which applies to both surround and 2-channel mixes.

[Gutbucket]

After setting this up on a stand at home I think the most I'm really willing to invest on this experiment at a gig is flying four mics. I'd probably go with two pairs - one with 3 mics facing forward and one rear or 2 pairs - one forward one rear. I may do this at an upcoming gig outdoors for Robert Randolph. May not work since at high volume there is some slapback from the rear which is very perceptible in the venue but not with forward facing directional mics from my usual location. Doesn't make sense to haul extra gear and get my expectations up only to realize I'm just accentuating the shortcomings of the acoustic space.

I think that's a good approach. Do whatever you would usually do in front. If that leaves you one extra channel to point backwards that's fine, if two that's fine too.  Space the rearward facing directional microphone a bit back from the forward facing mics as is practical, but don't worry too much about precisely how far.  As far back as your forward facing stereo pair is wide is probably good enough.  A foot or so is good if you can do that. 

Afterward, treat the "front facing" channels the same way you usually do for stereo playback. That way the front playback in quad (5 channel without the center speaker being essentially the same) won't be any different from how you typically use two or three "forward facing" mics.  Sure, if you do record 3-front mic channels, go ahead and try playback over 3-speakers in front.  It might work fine. You might like it better. But generally that's where increased spacing and/or angle between the three front microphones starts to pay off.

When considering things like slapback from the rear wall, give it a careful listen with the surround channels at various levels (same with how much of that rear-facing stuff you include in a stereo mix) before coming to any strong conclusions about how useful the rear facing content is gong to be.  Sometimes that kind of audible slapback works out really nicely, providing a great sense of depth and space, as long as its used at the right level.  I've now gotten to the point where I can more easily assess when that kind of echo from behind is likely to be good or not by listening to it during the performance with my head turned sideways so that one ear faces forward and the other backward.  But I never really know until I get home and listen to it later while playing with the levels how useful it will actually end up being.