Topic: Mixing down onstage jazz (recorded in M/S)

[acidjack]

Did my first "on stage" jazz recording last night (mics in front of the band; there wasn't really a stage).  Ran MK5+MK8 M/S and really like the sound.  Setup was like this:

        DRUMS        PERCUSSION       SAX        BASS

                                            ME


I have two mixdowns I have been playing with.   The "wider" one emphasizes the side channel more to try and reduce the heavier percussion hits of the bongo and other drum player that was right in front of me.  I think it also brings up the sax and sounds better, though I'm going to listen more. 

When I decode this, the L channel is much hotter than the R channel, since all of the loudest stuff was the to the left.  In a normal rock mixdown I would just boost the R channel to make it even, but given that I'm less familiar with recording this genre and have lots of studio recordings that are done all kinds of strange ways:

- Should I go ahead and raise the volume of the R channel to about even with the L channel, so the sax/bass-heavier channel is about as loud as the percussion-heavy channel?  Obviously, the split is more pronounced on the more side-heavy mixdown.

- Are there any other tips and tricks that people like to do in post?  Realizing every recording is unique, are there any common things, like, for example, the reduce bass/boost treble EQ move that's common for rock recordings, playing with the stereo imaging, blending the channels together, etc.?

Finally, a brief shout out to bomdiggity for encouraging me to do this. Have listened to jazz for a long time but have never really made it part of my recording efforts.  It's a lot of fun to use those nice high end mics to record actual instruments instead of just a PA system/guitar amp...

[bryonsos]

I only did a few M/S recordings, back when I had the BSCS-L and quickly abandoned it. Too much post work for my taste, I would monkey around with things to the point of frustration. Having said that, and after reading your dilema, I would try panning some of the left > right, maybe do the same with the right > left, then balance the levels. I find it to be very realistic to have things panned the way they were recorded, but for listening pleasure, I like having some of everything in both channels with enough of the image remaining to give you the "being there" feeling.

[noahbickart]

You can try doing EQ on the S and M independently.

[bombdiggity]

I'm not a M/S expert but I tend to look for balance between the instruments and leave the tone as it was (for jazz anyway). 

The panning is always an interesting question.  I do a lot of things with cards very close like that where the result provides a very distinct impression of how the stage laid out for someone sitting dead center there.  If that's really pleasing I leave it (as the spatial representation can be very nice).  Goody seems to agree with me on leaving that feel when it is right.  Other times (where there's more bleed or the balance seems off) I will cross mix or pan the channels to get a blend. 

The can sound very different in cans (the spatial representation is of course much more distinct) as opposed to speakers, so that is a factor too.  If I get to something really good in both I run with that.  If the spatial representation is still pronounced and pleasing from speakers that makes me lean towards that soundstage. 

Glad you got this one. 

[Gutbucket]

This kind of recording is my favorite thing.

Don't be afraid to leave one side somewhat louder than you normally might, but go ahead and adjust the Left/Right channel balance as you see fit.  Needing to do so is not uncommon in on-stage jazz recordings as there is often a louder instrument on one side or the other, which may be exaggerated from the mic's up-close perspective compared to what is heard from the audience.  Such an imbalance is more likely to occur on-stage than farther back in the room with or without a PA.  One thing I'll note about doing so is to suggest listening not only to the Left/Right balance of the instruments themselves, but also the entire ambience of the space- the more subtle aspects of stage reflections, reverb, crowd reaction, and room-bloom.  Too much balance adjustment and that ambient stuff gets lopsided and rapidly looses an important part of the 'magic mojo' that makes great on-stage recordings exceptional in my opinion.  I often find myself striking something of a compromise between best direct instrument balance and most natural ambience.. and often leaving one side louder sounds real and exciting when I mentally switch hats from 'picky sound engineer listening for technical details' to 'music aficionado simply enjoying the music'.  Naturalness, excitement, emotion and envelopment in the music is what maters most.

As for EQ adjustment I find it to be generally opposite of what’s often called for in my PA recordings form farther back.  Usually little need to cut bass and boost presence range or highs as much- I more often boost the low end a bit, and sometimes need to tame the brilliance of the cymbals given their proximity.. but of course that's just a generalization and there are always exceptions.  What may seem to be relatively small differences in distance to the drums and cymbals seems to effect that significantly.

[Gutbucket]

Have listened to jazz for a long time but have never really made it part of my recording efforts.  It's a lot of fun to use those nice high end mics to record actual instruments instead of just a PA system/guitar amp...

One of the many reasons I love doing this is that we can make recordings this way which simply blow away much commercial stuff.  And they needn't be overly complex, no need for SBD matrices to get enough up-front clarity.  This is the real deal.

[macdaddy]

^^^
Amen to that.

Can't wait to hear the results

[page]

- Should I go ahead and raise the volume of the R channel to about even with the L channel, so the sax/bass-heavier channel is about as loud as the percussion-heavy channel?  Obviously, the split is more pronounced on the more side-heavy mixdown.

Do the mix down with the assumption that you're only listening for where you want it in the mix (center, right of center, far right, etc), then adjust levels. If it sounds better with the stuff in the right channel amped, then as Lee notes; go for it. Once you start doing mixing and stuff, lots of "rules" that we have for OTS style taping sort of become much more bendable.

- Are there any other tips and tricks that people like to do in post?  Realizing every recording is unique, are there any common things, like, for example, the reduce bass/boost treble EQ move that's common for rock recordings, playing with the stereo imaging, blending the channels together, etc.?

Compression (either single band or multiband, parallel or serial), EQ, limiting. Each has a place in my tool box when I'm doing stage type stuff. There are others, but I think those are the big three that I'd use in this environment.

Finally, a brief shout out to bomdiggity for encouraging me to do this. Have listened to jazz for a long time but have never really made it part of my recording efforts.  It's a lot of fun to use those nice high end mics to record actual instruments instead of just a PA system/guitar amp...

 

Got a raw/pre-MSdecode sample that is representative that we could play with?

edit: clarification in the sample request.

[acidjack]

Thanks all.

I finished the mix down I liked. I should be posting FLAC/MP3 link here shortly.  Hoping to share more widely if the artist confirms he's cool with it.

In the end I did very little.  About a 1.3dB bass boost at just under 100Hz and a slight drop around 8kHz to kill some harshness in the cymbals.  A little exciter warmth. 

I ended up going with the more "mid-heavy" mix, though it's still about 50/50 between M and S.  It felt tighter and had better spacial representation to me.

I did not do any compression or limiting; maybe I will depending what folks think. I don't mind doing some light compression on rock recordings, but felt less confident doing it to music with this much dynamics. 

[bombdiggity]

Thanks all.

I finished the mix down I liked. I should be posting FLAC/MP3 link here shortly.  Hoping to share more widely if the artist confirms he's cool with it.

In the end I did very little.  About a 1.3dB bass boost at just under 100Hz and a slight drop around 8kHz to kill some harshness in the cymbals.  A little exciter warmth. 

I ended up going with the more "mid-heavy" mix, though it's still about 50/50 between M and S.  It felt tighter and had better spacial representation to me.

I did not do any compression or limiting; maybe I will depending what folks think. I don't mind doing some light compression on rock recordings, but felt less confident doing it to music with this much dynamics.

As I said (offline) to you (and as Gutbucket would probably agree) with jazz of this sort played unamplified the onstage mix is really pretty close to optimal because the band has to be able to hear each other clearly (in totality) to be able to play this music effectively.  These sorts of players are masters of tone, dynamics and interplay and they usually know how they should sound and how to adjust their natural levels to each other (if you can't hear and react to other players and don't listen very closely you won't last long in that arena). 

I agree I would not use compression.  The dynamics are critical in this music and one of the things that can give the ambient recording of it such a special feel.  Maybe a little limiting if relative proximity caused some issues on a particular instrument or passage.  I'd generally only eq if the mics are colored a little opposite the space or PA (when there is one) or if the bass is relatively weak as sometimes happens since it usually is a lot lower volume than the drums and horns where the player can really pick their spots to blast...  Most of the rest of the instruments inhabit the same natural range so you can't really "mix" things.  EQ is really only effective for overall tonal shading other than the real low end. 

Looking forward to hearing it.  The artist involved is one of the best in jazz bar none. 

[Gutbucket]

Yep, agreed on the on-stage self mixing.

And Page raises good points on dynamics which I didn’t mention earlier because one can easily get lost down the rabbit hole with it.. It may not be as loud as a big PA gig in an absolute SPL sense, yet can get wildly dynamic up close. That's especially true of unamplified drums, even with excellent jazz players who have a strong command over and awareness of the dynamics of their instrument.

The problem that creates for playback is more-or-less three-fold:
1) A handful of peaks pushes the average level down lower than would be optimal. (top end problem)
2) At reasonable playback levels which accommodate the loud parts, the really quiet stuff between numbers becomes inaudible.  Thinking stuff like musician communication between tunes, interaction with the audience if the performer is talking directly and not using an announce mic, etc. (bottom end problem)
3) The dynamics during/within the music itself can be too much.  The loud sections are deafening while the soft sections, like those featuring delicate drum brush work are too quiet to be intimate sounding and engaging. (too much difference between the top and bottom ends)

Modifying each of those things is best done with different approaches.  You can spend a lot of time getting it all managed while keeping everything sounding transparent and making sure what you are doing is making things better, not worse.  All that can be fun, a challenge and a learning experience.. or it can be way too much work and a chore.

You don’t have to do anything to the dynamics, but this kind of stuff is where it really pays off if you figure it out.  PA amped stuff is already compressed and limited to a large extent.  Working on the dynamics of something like that is easier, but also has less of a payback and less smiles of self-satisfaction when you get it right.

If you want to keep it simple, manually scanning for and knocking down a handful of the wildest drum hit transients will allow you to bring the average level up considerably without sonic penalty.  A good limiter set properly can do the same, but is tricker to setup without sonic penalty on this kind of material IME.  Page can do it better than I.

That may be enough alone, but if you want to do more, consider some parallel compression to bring up the bottom and also bring out all the hidden details and tone.  That’s trickier than simply knocking down the biggest transients, but not too difficult to get sounding good and much jazz magic lurks there.

The middle is the most difficult to get right for me without the right compressor.  I’d put that last in the list of importance anyway.  Sometimes I’ll burn a copy that sounds great at home on the big system with only the top and bottom dynamics worked on, but is still way to dynamic for any chance of listening in the car or passing out to others and I’ll go back and envelope down the loudest parts if it was a mostly a quiet show, or envelope up the quietest bits if it was a mostly loud one.. but that’s probably more work than you want to do and the hardest to do well.

[acidjack]

Ahh, the parallel compression would have been a good move... (along with your other great advice).  I should also gain a better understanding of proper use of a limiter.... I use them sometimes, but I doubt I'm using them totally correctly.

Anyway, for those interested, here are links to the set.  They expire in not too long and are limited by number of D/Ls.  The artist will also hopefully d/l them himself since I just sent them his way.

FLAC: http://www.yousendit.com/download/UW14SlIrd0FtNEpqQThUQw

MP3: http://www.yousendit.com/download/UW14SlJ3TXZ0d0VYRHNUQw

Constructive criticisms (and of course compliments) appreciated! 

The rig was not as high or as pointed upward as I'd have liked... Maybe 3-4ft pole, so it'd be below head height. 

[Gutbucket]

Thanks for the links. I'll try and give a listen this weekend.

Not a post production thing, and I’ve posted about this previously in other threads, but here's my thinking on balancing the instrumentation in terms of level, position and depth, by way of mic placement on stage, which I think bears repeating here..  maybe only because it is a slow day at work. 

Even though the musicians are balancing themselves by what they hear, some instruments are simply louder and / or project differently than others so the sound balance the musicians hear may be significantly different than what is heard at the recording position.  Guitar cabinets project significantly and are much louder and more up-front sounding on-axis than off-axis, especially down at floor level.  Pianos and bass are more diffuse and omnidirectional and don't carry or project like that, drum kit radiation is complex- generally omnidirectional taken as a whole, yet each component of the kit is like a separate instrument, with closer drums shadowing the farther ones, the farther ones sounding both quieter and more diffuse (farther away), etc.

When I setup onstage for this kind of stuff I try to take all that into consideration, especially if I have freedom of mic placement.  Simultaneously, I'm weighing those loudness and directivity aspects against which instruments I feel should sound more upfront and which can sound more diffuse and farther back.  That upfront/back – direct/diffuse thing isn’t just a practical compromise issue but what makes for some of the magic mojo when balanced well, giving the recording a lush sense of depth and dimension which I rarely hear in modern commercial releases or PA recordings where everything is equally distant sounding, be that close or far.  The third and more obvious consideration is left/right position of sources.

I ‘ll sometimes move off-center and angle the center axis of the mic array across the stage so it is physically closer to the bass, or sometimes piano, to pick up the more subtle details and direct sound from those instruments, and farther away but still sort of on-axis to a guitar amp over on the other side which carries more and is loud down on the floor where my mics often are, yet the stereo image is still balanced and the depth of things is appropriate.

I usually try to favor the snare side of the drums with a direct line-of-sight to the snare from my mic position, and don’t mind if the toms shadowed on the other side end up sounding more distant.  Unless the drums are only playing a support roll, in which case they can be more distant and diffuse sounding like they are farther in back of the other instruments, I usually want clear transient detail from the drum kit and snare especially.  It’s an instrument of transient snap which needs that clarity more than other tonally sustaining instruments that can be more diffuse sounding without suffering for it.. as long as it doesn’t become too loud.  The other instruments still need to sound loud enough in comparison even if they may be farther back and more diffuse or sounding.

It’s an art to juggle it all by eye and past experience without a chance to listen to the mic feed, and often no chance to even listen with one’s head in the mic position before the set.  But that’s all part of what makes it interesting, challenging and rewarding.

[bombdiggity]

> Even though the musicians are balancing themselves by what they hear, some instruments are simply louder and / or project differently than others so the sound balance the musicians hear may be significantly different than what is heard at the recording position. 

Very good point. 

Horns can be trouble because they project very much away from the player (whereas an amp is usually behind the player) and are typically out in the front line center too (closest to typical near-stage/on-stage mic position).  Horns are also not naturally balanced with each other (trumpets are typically the loudest thing on any stage, saxes can vary with the player, trombone is usually lower than those and then there are other horns, not to mention woodwinds like clarinet...).  An unamplified horn section is a huge challenge (and in the end you basically just get what you get no matter how you do it).  The smarter big bands at least put the trumpets in the back row!   

> Pianos and bass are more diffuse and omnidirectional and don't carry or project like that,

I agree those are the usually trouble spot in the ambient balance.  My friends always gripe about acoustic piano in almost any context (trio, whatever).  Piano really has to go through a PA to have much hope of being in the mix (unless maybe it's Larry Willis or someone who really plays hard).  Soundpeople almost never have the piano up to relative parity (and its sound will carry a little better toward the back through a PA so is deceptive at a distance).  After the last show I did (where I got to do a sort of soundcheck/level set) someone unsolicited said how happy they were they could really hear the piano "for once"...  I think I had it up at least twice as high as the house guy leaves the default setting.  There's also the issue of comping vs. solos.  The range of levels the player is at can be pretty broad relative to many other instruments. 

Contrabass is (nearly) always played through an amp (though usually a small one).  So the bassist is controlling that element and may or may not set it to relative parity.  It dissipates quickly but also tends to be coming from behind the band and therefore often is partially blocked (so may be about right on stage but lost a row or two back in the audience). 

> drum kit radiation is complex- generally omnidirectional taken as a whole, yet each component of the kit is like a separate instrument, with closer drums shadowing the farther ones, the farther ones sounding both quieter and more diffuse (farther away), etc.

Very good to keep in mind.  Drums typically "rule the house."  Some players are more aware of that than others...  I haven't really thought about strategies to work with balance among the elements of the kit.  I'm usually more worried about them in total not overpowering things or creating transient spikes. 

> That upfront/back – direct/diffuse thing isn’t just a practical compromise issue but what makes for some of the magic mojo when balanced well, giving the recording a lush sense of depth and dimension which I rarely hear in modern commercial releases or PA recordings where everything is equally distant sounding, be that close or far. 

I'd like to have you with me in Baltimore next week! 

I agree on the merits of lush ambiance.  I'm not sure if the average listener is trained by commercial releases and what they usually hear most places to be somewhat oblivious.  I think we as recorders are a lot more sensitive to subtleties that even most listeners of audience recordings have no clue about. 

We rarely have the run of the house so the magic mojo is really in eyeballing the situation (often with none of the players in hearing/sight) and then setting up where you get a nice representation of all the elements (without drawing the ire of anyone by being too pushy). 

Certain mics provide much better spatial representation than others, which is a factor in that mojo as well. 

> The third and more obvious consideration is left/right position of sources.

I'd given that a little thought lately.  I do really like when you can get soloists oriented that way.  It's been really effective when I've had two tenors or several horns and there's a blend of solos and unison.  Getting a pair onto sides gives a live feel and helps establish contrast and a natural but very distinct stereo feel.   I can also see that being a useful strategy for post mixing if you can get components somewhat isolated and need to adjust the recorded balance. 

> It’s an art to juggle it all by eye and past experience without a chance to listen to the mic feed, and often no chance to even listen with one’s head in the mic position before the set.  But that’s all part of what makes it interesting, challenging and rewarding.

+1

[acidjack]

incorporating some of the (er, easier) ideas suggested here, I did a 2nd mix which added:

- light parallel compression
- Centering R/L (by extracting to mono and then panning each channel only 85% in its respective direction)

Also fixed a tracking mistake on track 6.

I think the result is fatter-sounding, still has plenty of ambiance and separation, but is less jarring to listen to on headphones.  http://www.yousendit.com/download/UW14cHBOOW4wMEVYRHRVag

[danlynch]

I've used panning on that backyard acoustic stuff we've done over the last few years, exactly because listening on headphones (while still feeling like you're "there") is jarring.   

[acidjack]

I've used panning on that backyard acoustic stuff we've done over the last few years, exactly because listening on headphones (while still feeling like you're "there") is jarring.

Exactly. The weirdest thing I find about recordings that are ultra-panned is when official releases are like that.  A few jazz records from the 60s I have do that, and IIRC Johnny Cash "Folsom Prison" also has some really strange L/R differences.

[page]

I've used panning on that backyard acoustic stuff we've done over the last few years, exactly because listening on headphones (while still feeling like you're "there") is jarring.

Exactly. The weirdest thing I find about recordings that are ultra-panned is when official releases are like that.  A few jazz records from the 60s I have do that, and IIRC Johnny Cash "Folsom Prison" also has some really strange L/R differences.

cause mixing for stereo was both an afterthought during that time and music headphones were far and few between (early Beatles albums are similar).

I gave the first one a whirl and found I wanted to came them in around 50% off of each channel. Started at 85, and dropped it to 50 for general listening. I'm not sure I'd recommend going that far, just personal taste at the time.

There is a trick that I picked up from someone on GS that when doing midside (or mixing in general if you can place stuff in a flat mix) is to mix it so it's comfortable on loud speakers and then slowly condense it on headphones until it's not unbearable. The rough guide is you can take two hard pans and drop them to about 80% of their original and still keep enough stereo so that you don't get a monoey soundstage on loudspeakers, yet the headphones have enough sound in both that it's not jarring. It's more of a guideline than a rule, but food for thought.

I ‘ll sometimes move off-center and angle the center axis of the mic array across the stage so it is physically closer to the bass, or sometimes piano, to pick up the more subtle details and direct sound from those instruments, and farther away but still sort of on-axis to a guitar amp over on the other side which carries more and is loud down on the floor where my mics often are, yet the stereo image is still balanced and the depth of things is appropriate.

+1

I've done this a lot before. Being centered is one of those rules I was alluding to earlier that can be bent (or even outright broken) when doing stuff on stage or stage lip. Just compensate by adjusting where you are pointing. Yeah, it makes the audience seem sort of squirrelly when they clap, but the music sounds better. (and some other stuff he wrote too, but that's the big one that stuck out to me).

edit: added info.