Taperssection.com

Gear / Technical Help => Post-Processing, Computer / Streaming / Internet Devices & Related Activity => Topic started by: mattmiller on September 21, 2019, 11:08:02 PM

Title: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on September 21, 2019, 11:08:02 PM
I recently came into possession of an unmixed 29-track recording of a show I attended in 2008.  The band had it professionally recorded, and because of that the venue refused to allow me to record (even though the tour manager argued on my behalf).  They never did anything with the recording, and it's one of a bunch of recordings sitting on hard drives that they recently re-discovered, and this one got passed to me to mix for the purpose of sharing on the Archive.

Having never mixed more than 4 or 5 tracks before, I'm looking for general advice on how to approach this job.  It's a three-piece band, so is it appropriate to pan all of the drums (stage right) hard to the left, the keys/bass (stage left) hard to the right, and lead guitar/vocals right down the middle?  What issues do I need to watch out for?  There are 10 tracks just devoted to percussion, obviously not isolated from one another in such close proximity.  Is it customary to not even use every track?  Basically, I'm looking for any advice or cautionary tales.

It's a three-piece band, but they're all multi-instrumentalists, so the keyboard player also plays bass and accordion on some songs, and the drummer plays mandolin and banjo, so before I do anything else I've started mapping all the tracks in Excel, plotting who is where on every song.  I'm hoping this organization will pay off when it comes time to mix everything.

If it helps, here are the tracks I'm working with:

01. Kick - Beta 91
02. Kick - Beta 52
03. Snare Top - sm57
04. Snare Bottom - e604
05. Hat - MKH8040
06. Rack - e604
07. Floor - e604
08. Ride - VX5
09. OH SR - M160
10. OH SL - M160
11. Bass DI
12. Bass Mic - PL20
13. El. Gtr - MD441
14. Ac. Gtr - BSS DI
15. Mandolin - BSS DI
16. Banjo - M160
17. Key 1 (Triton) - Radial DI
18. Key 2 (Wurlitzer) - BSS DI
19. Key 3 (M-Audio)- BSS DI
20. Key 4 (Keytar) - Radial DI
21. Accordion - MKH8040
24. SR Vox - OM5
25. Lead Vox - OM5
26. SL Vox - OM5
27. Drum Vox - OM7
28. Keys Vox - OM5
30. Audience House Left - MKH416
31. Audience House Center - MKH416
32. Audience House Right - MKH416
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: beatkilla on September 22, 2019, 12:18:16 AM
I am interested to look at this and have a go at it.

Other than that my suggestion is to route all the percussion to a master bus.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: EmRR on September 22, 2019, 08:36:22 AM
Whew, this is hard to give concise advice about....from my phone.....  This is what i do mainly, multitrack recording and I started out as a remote guy doing club recordings. 

Usually I’d start with the audience mics, but since they are shotguns it may not make a great picture.  You may find there’s too much audible delay in the audience mics, and need to time align them to a different point.  Or the may be onstage pointed out, being shotguns.

You don’t have to use every mic, but I’d try to make sense of all of them as a group rather than working on subsections too much, at least at first.  Things like the kick, snare and bass that have multiple inputs can be good to group and lock their relative levels together, same with pairing the OH's.  You’ll probably find you have to work with and around bleed that will constrain your options.  In those cases very tiny level adjustments can sometimes be the biggest perceived differences. 

Panning does not have to follow stage plot.  Sure, try it, but also try any other placement that yields good clarity between similar registered instruments/voices, and good L/R balance.

I'd try to get basic levels without any processing BEFORE you start applying any.  The only thing I might do is use high pass filters to remove unnecessary lows from various things, that tends to be the number one clean-up tool in a large live multitrack.  The DI tracks in particular may have subsonic non-musical thumps that need to be high passed.

The drums are 10 tracks but it's still one instrument.   I'd go through the toms and ID where they were struck, and automate the rest down 6dB or so to manage bleed, which tends to be cymbal wash.  Gates sometimes work, but not as well.  Do that early on, as housekeeping.  If this is like most live shows, you will find cymbals in all the vocal mics, toms, and some other things, and you'll be using less of the actual OH to balance it out. 

The mandolin and acoustic, being DI, will probably benefit from some sort of amp simulator, could be in parallel if it doesn't work well in-line.  The keys may benefit from that as well, depending on what they sounded like in the room.  The bass DI may or may not be useful against the mic signal, and may need time aligning.  I don't usually use a lot of a bass DI if there's a mic, and I sometimes turn it off. 

I'd probably find a spreadsheet redundant, I'd just lay out the tracks by person, and it'll sort itself pretty easily.  You'll probably have to break the mix down per song anyway, with level automation.  I rarely see a multitrack live show that has anything like set-and-forget levels beginning to end. 

Just realizing from the list that this was probably recorded by Steve Remote / ASL.  There are clues. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on September 22, 2019, 08:56:10 AM
I am interested to look at this and have a go at it.

Other than that my suggestion is to route all the percussion to a master bus.

I appreciate the offer.  This has sentimental value to me as it was just my second time seeing these guys (Stephen Kellogg and the Sixers), and is the only show (of 50+) that I didn't record.  I want to see what I can do with it, but I will certainly not hesitate to seek help if I feel like I'm not able to do it justice.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on September 22, 2019, 09:17:35 AM
Whew, this is hard to give concise advice about....from my phone.....  This is what i do mainly, multitrack recording and I started out as a remote guy doing club recordings. 

Usually I’d start with the audience mics, but since they are shotguns it may not make a great picture.  You may find there’s too much audible delay in the audience mics, and need to time align them to a different point.  Or the may be onstage pointed out, being shotguns.

You don’t have to use every mic, but I’d try to make sense of all of them as a group rather than working on subsections too much, at least at first.  Things like the kick, snare and bass that have multiple inputs can be good to group and lock their relative levels together, same with pairing the OH's.  You’ll probably find you have to work with and around bleed that will constrain your options.  In those cases very tiny level adjustments can sometimes be the biggest perceived differences. 

Panning does not have to follow stage plot.  Sure, try it, but also try any other placement that yields good clarity between similar registered instruments/voices, and good L/R balance.

I'd try to get basic levels without any processing BEFORE you start applying any.  The only thing I might do is use high pass filters to remove unnecessary lows from various things, that tends to be the number one clean-up tool in a large live multitrack.  The DI tracks in particular may have subsonic non-musical thumps that need to be high passed.

The mandolin and acoustic, being DI, will probably benefit from some sort of amp simulator, could be in parallel if it doesn't work well in-line.  The keys may benefit from that as well, depending on what they sounded like in the room.  The bass DI may or may not be useful against the mic signal, and may need time aligning.  I don't usually use a lot of a bass DI if there's a mic, and I sometimes turn it off. 

I'd probably find a spreadsheet redundant, I'd just lay out the tracks by person, and it'll sort itself pretty easily.  You'll probably have to break the mix down per song anyway, with level automation.  I rarely see a multitrack live show that has anything like set-and-forget levels beginning to end. 

Just realizing from the list that this was probably recorded by Steve Remote / ASL.  There are clues.

Thanks for all of this.  The audience mics are, indeed, shotguns pointed out from the stage.  And they sound like extremely directional mics pointed away from the stage.  Hopefully I can utilize them enough to get the audience fully injected into the final edit, while not negatively affecting the music.  I assume this will be fine, and the preferred method for mixing purposes, else they would have placed the mics in the audience pointed at the stage.

I can't say for certain who recorded it, but after Googling Steve Remote, I see he's NYC-based, and the two shows on that tour that were recorded were Philly (this one) and the following night in Albany, so I'd say it's a good possibility.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: EmRR on September 22, 2019, 09:25:02 AM
I added a few more points above, good luck with it.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: yousef on September 22, 2019, 11:52:40 AM
I think I'd agree with most of what is said above but my workflow is:

Start by mixing in mono

Hi-pass everything at around 100-150Hz unless the instrument needs something down there

Start with the drums kick>snare>hat(usually mixed v.low)>toms>o/h - then make a bus for these ("folder" if you're using Reaper) so you can mix the drums as one instrument. It is often worth pausing after your initial drum balance is set and trying muting everthing but the kick and overheads - mixing the others back in only if needed

Often I add some form of compression to the drum bus - a Variety of Sound Thrillseeker plugin, or a multiband comp or even a tape simulator

Add in the bass - often using an amp sim on the DI and leaving the amp channel muted. Occasionally I'll split the bass into two channels and treat one to get a good low end and the other to concentrate on higher frequencies - then bus them together as per the drums. But then I'm a bass player and probably have issues to deal with here.

Guitar, keys and other instruments in turn now. I almost always use an amp sim on keyboard tracks

Lead vox - depending on how the raw tracks sound I'll use any or all of comp, EQ, multiband comp, tape sim, enhancer, reverb, delay, even ADT-simulation on one occasion

Backing vocals, often EQ'ing slightly to thin them out to prevent clashes with the main vocal

I then set up a basic mono mix and listen all the way through several times over several days.

Go back and start mixing levels for each song, being careful to mute instruments and vocal mics when they aren't needed (volume envelopes can help here).

I add in audience mics at this point - usually just in between songs, unless they are particularly nice and add a nice ambience throughout.

I start panning now, but only ever very slight amounts - a few percent in either direction is enough to give a sense of space, any more and it starts sounding odd to me.

On the master bus add overall EQ (Baxter EQ plugin is nice here), multitband comp, reverb (Hybrid Reverb 2) and limiter (Limiter No6 is my current favourite). SOmetimes a tape sim or valve/tube sim to help everything mush together.

I then go back and listen to this mix a few times, making sure I have a pad next to me to scribble thoughts, corrections and timings. Sometimes I go back at this point and envelope/EQ burst of feedback/mic hits etc

I like the tape sim stuff as it can soften what can often be quite cold or clinical sounding digital recordings

Things to look out for:

aud mics will probably need time-shifting a little depending on where they were in the room
watch out for phase issues - I once had a keyboard track where the L+R were out of phase and rendered it inaudible until this was corrected
take lots of pauses - ears fatigue quickly and this leads to poor mix decisions
not every mic/track needs to be used - only add something if it improves the overall sound
plugins are great but there can be diminishing returns
there are no rules - one of my favourite professional producers has advised me to do completely unorthodox things in the past: if it sounds good, it *is* good
test on a variety of systems before signing off

The Sound on Sound archive has some great articles that can help. Off the top of my head the ones on the Led Zep reunion show and the Stones' Shine a Light were particularly illuminating.

Have fun!
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on September 22, 2019, 05:32:06 PM
I think I'd agree with most of what is said above but my workflow is:

Thanks for all of this insight.  A few questions:

Quote
Start with the drums kick>snare>hat(usually mixed v.low)>toms>o/h - then make a bus for these ("folder" if you're using Reaper) so you can mix the drums as one instrument. It is often worth pausing after your initial drum balance is set and trying muting everthing but the kick and overheads - mixing the others back in only if needed

I've only ever used Cool Edit Pro.  Am I handicapped in any way with that for a project this big?  Either with track grouping/bussing for the percussion (which I've never had to do, so I'm not sure of its ability), or with the availability of any of the plugins that have been referenced?

Quote
Go back and start mixing levels for each song, being careful to mute instruments and vocal mics when they aren't needed (volume envelopes can help here).

This is something I was wondering about.  So for these "noisy" mic tracks that are only used occasionally (e.g., the accordion, and the SL and SR vocal mics for when the guys step out from behind the drums and keyboard to play mandolin and bass...), having them muted when they're not being used doesn't introduce any audible change to the music when the ambient sound is suddenly faded in along with the accordion or vocal parts?  I was contemplating the possibility of needing to leave those entire tracks in there just to maintain consistency in the mix, but I'd love to be able to just fade them in and out as they're needed.

Quote
Things to look out for:  aud mics will probably need time-shifting a little depending on where they were in the room

They were on the front of the stage, pointed at the crowd.  Should this (in theory) be an extremely small time-shift in comparison to what I'm used to doing to align my mics 30 feet back from the stage with a board feed?

A general question about Cool Edit Pro, I guess, or maybe about this process in general.  Which of these actions (if any) are non-destructive to the source files?  Or do I need to be working with copies of the original files because this high-pass filter or that plugin will be necessarily saving its changes over the source files?

Also, when I've previously mixed a board feed with my mic recording, whether I've needed to or not, my first step has always been to normalize both sources to something like -6 so that when they were mixed together I wouldn't introduce any clipping.  I also always adjusted the volume sliders in such a way to never increase the overall volume.  I would raise one and lower the other, rather then raising one twice as much and leaving the other at 0.  Is it good practice or completely unnecessary to do any kind of normalizing of all of the source files to any particular peak level?

Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: yousef on September 22, 2019, 07:11:41 PM
A caveat here is that this is all my own essentially self-taught workflow - I'm happy with the results but others may get better end results and/or get there more efficiently...

Quote
This is something I was wondering about.  So for these "noisy" mic tracks that are only used occasionally (e.g., the accordion, and the SL and SR vocal mics for when the guys step out from behind the drums and keyboard to play mandolin and bass...), having them muted when they're not being used doesn't introduce any audible change to the music when the ambient sound is suddenly faded in along with the accordion or vocal parts?  I was contemplating the possibility of needing to leave those entire tracks in there just to maintain consistency in the mix, but I'd love to be able to just fade them in and out as they're needed.

There will be slight changes in ambience but you can mitigate against this by making the fades come in and out over a second or so and then checking to see if everything sounds ok. I think it will depend on how much spill there is from other instruments and how high you need the track in question to be.

Re aud mics:
Quote

They were on the front of the stage, pointed at the crowd.  Should this (in theory) be an extremely small time-shift in comparison to what I'm used to doing to align my mics 30 feet back from the stage with a board feed?

I'd say so - again it should be very obvious from listening if it is right or not. Watch out for the phase though. Note that in the Sound on Sound Rolling Stones article mentioned above, Bob Clearmountain talks of adding a slight echo to the vocals so that the extra ambience added by the room mics between songs is not too obvious - one of the best things I've ever learned  :D

Quote
A general question about Cool Edit Pro, I guess, or maybe about this process in general.  Which of these actions (if any) are non-destructive to the source files?  Or do I need to be working with copies of the original files because this high-pass filter or that plugin will be necessarily saving its changes over the source files?

All of it is non-destructive, even after you hit "render". But I always have a complete backup of the source files on a separate drive/machine before I start editing and mixing. I think I missed a quote above re using Cool Edit Pro for mixing and busing in particular - I've never used it so can't be certain but I'm sure it will be set up for this - no doubt YouTube will have a few good tutorials. I would just download Reaper though - it is powerful and intuitive, the bundled plugins are great (even if many of the presets are not) and it is completely free to trial for however long you need it with no limitations whatsoever. The license is super cheap too.

Quote
Also, when I've previously mixed a board feed with my mic recording, whether I've needed to or not, my first step has always been to normalize both sources to something like -6 so that when they were mixed together I wouldn't introduce any clipping.  I also always adjusted the volume sliders in such a way to never increase the overall volume.  I would raise one and lower the other, rather then raising one twice as much and leaving the other at 0.  Is it good practice or completely unnecessary to do any kind of normalizing of all of the source files to any particular peak level?

With my workflow, I usually start with the drum bus peaking at about -5dB and then when I start adding the other instruments in I have to tame it to something more reasonable. I think the way most software is set up you can have the cumulative levels ending up way into the red (as long as no individual track is distorting) but by pulling back the master fader all is magically unclipped and clean.

Do keep us all informed of how this goes - and of any cool tricks you come across.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on September 22, 2019, 11:10:17 PM

The drums are 10 tracks but it's still one instrument.   I'd go through the toms and ID where they were struck, and automate the rest down 6dB or so to manage bleed, which tends to be cymbal wash.  Gates sometimes work, but not as well.  Do that early on, as housekeeping.  If this is like most live shows, you will find cymbals in all the vocal mics, toms, and some other things, and you'll be using less of the actual OH to balance it out. 
Are you trying to make it sound like a completely polished professional studio mix? Or do you just want a good-sounding live show? What is your time budget for this project?
EmRR obviously knows how to do this in a top-notch fashion, but I can't see suggesting to someone who has never mixed a live show that they go over every single drum hit with automation, because gates can be clunky...
Sorry I can't be more constructive at the moment.  .  . This is a very cool project, and learning to make a good mix is a fantastic skill to have.

Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on September 23, 2019, 02:45:24 AM
Multitrack workflow is clearly a very personal subject. Lots of good advice here, but also a lot I don’t agree with. The basic start to my process (I also do live multitrack mixdowns as part of my day job) goes like this:

Time align all the tracks to the snare. This will help small phase issues that you might otherwise not really hear, but can have a great affect on the overall tone of the tape. Summation is a very powerful tool.

Pan from the get go. I understand the idea of mixing in mono to begin with, but I don’t agree with it. Planning changes the perceived balance between the instruments, so why neglect it until the end. Usually for live stuff I try to mix to stage plot, but it is not 100% necessary. I usually avoid panning beyond 60ish degrees, as I think instruments panned all the way out in live music sound odd.

Buss everything. Drum bus, guitar bus, vox bus, and so on. Output these busses to another master bus. Use the busses as a basis start processing.Mix the instruments in the bus and Apply small amounts of compression to the groups. Then balance the busses into the master. If you want to try parallel comp, you can also do it here by duplicating the busses and compressing one.

For room mics I like to time align, pan wide, remove the center if possible, and sometimes sidechain to the level of the music. That would mean that the volume of the room will come up when the music is at a lower volume. Whether or not I do this depends on whether the aud mics capture the room ambience or really just the audience noise.

I would say be very cautious about gating and automating noise in live Multis. For tracks like the accordion that aren’t being used every song, you could mute between songs, or just pull down a few dB to mitigate bleed. I’m not a fan of gating/automating drum hits in live recordings because I think it sounds unnatural. With a well phase aligned plot of mics, you shouldn’t really have to worry about too much gross bleed.

From there you get into normal mixing territory. Compressing individual mics, balancing between instruments, verb, etc...
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: yousef on September 23, 2019, 04:01:43 AM
This is a great thread... I've never once thought of using a sidechain for the audience mics - brilliant!
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: EmRR on September 23, 2019, 07:28:57 AM
Re: gating.  You should notice i said it probably wouldn’t work well.  I’d never do it to anything but toms. Then you should notice I said automate 6dB, which is the territory of non-obvious bleed mitigation.  I find that automation pass on many drummers far faster than fighting cymbal bleed in everything else. Also, none of that may be needed at all.  And it’ll be a major pain and waste of time if it’s Keith Moon.  Many drummers have toms they almost never hit at all, or hit so lightly they have to be cranked up enough the bleed  changes the sound of the entire kit, making it hard to just turn on the rare random hit non-obviously. 

I used to time align most everything within a drum kit, then I generally stopped doing so.  If it’s well recorded it’s probably unnecessary. Depending on your tools, that could be a really fast path or a really slow painful one.  The double mic’d snare and kick are contenders, relative to themselves.  Flipping polarity there too to see which way they work together, then against the OH’s. 

It’s funny, I used to bus everything too, and mostly gave that up a decade ago.  Non-bussed grouping is another similar idea to consider, i do that with things like those OH’s. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on September 23, 2019, 08:57:43 AM
As I suspected, there is obviously no right and wrong approach to any of this, as there are some contradictory suggestions here that I'm sure yield equally pleasing results.  I suppose that's the nature of any project where the quality of the end product is subject to personal preference.

Quote
Are you trying to make it sound like a completely polished professional studio mix? Or do you just want a good-sounding live show? What is your time budget for this project?

I have no expectation of being able to put anything together that is going to fool a professional into thinking it was mixed by somebody with years of experience.  I just want it (at minimum) to be generally considered to be adequately done for the sake of having it represent the band's talent well.  That said, while I am quite busy with other things and don't have a ton of time every night to devote to this, I also don't have any pressure on me to get it done by any particular date, so I'm happy to take my time with it and proceed cautiously and methodically.  My mapping of everything out in a spreadsheet is both allowing me to become familiar with all the tracks and organize my thoughts, while also preventing me from getting too far ahead of myself and playing with settings I don't fully understand.  For that reason, I've also ordered an old used copy of an apparently well-respected book on Cool Edit Pro, since I'm sure I've never used 99% of its functions in my normal workflow of cutting my mic recordings and occasionally matrixing it with a SBD feed.  For example, I searched through all of the menus last night and it wasn't immediately obvious to me even how to buss tracks together.  I'm sure the book will be quite beneficial to helping me understand some of the different approaches that have been proposed here.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on September 23, 2019, 08:58:16 AM
I used to time align most everything within a drum kit, then I generally stopped doing so.  If it’s well recorded it’s probably unnecessary. The double mic’d snare and kick are contenders, relative to themselves.  Flipping polarity there too to see which way they work together, then against the OH’s. 

It’s funny, I used to bus everything too, and mostly gave that up a decade ago.  Non-bussed grouping is another similar idea to consider, i do that with things like those OH’s.

Sounds as though you and I have developed our workflow opposite each other. I used to think time aligning everything wasn’t necessary, now I do it as a rule. The kit is actually the part I’m least worried about. Aligning vocal mics makes a world of difference IMO.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Sloan Simpson on September 23, 2019, 09:42:59 AM
For room mics I like to time align, pan wide, remove the center if possible, and sometimes sidechain to the level of the music. That would mean that the volume of the room will come up when the music is at a lower volume. Whether or not I do this depends on whether the aud mics capture the room ambience or really just the audience noise.


This is interesting, would like to try it. For removing the center are you just using a M/S plug and pulling down the mid?

Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: EmRR on September 23, 2019, 09:58:00 AM
As I suspected, there is obviously no right and wrong approach to any of this, as there are some contradictory suggestions here that I'm sure yield equally pleasing results.  I suppose that's the nature of any project where the quality of the end product is subject to personal preference.

What may seem contradictory amongst everyone's suggestions is really the nature of mixing, what works on one thing I record may not work at all on the next thing I record, even if it's the exact same technical setup in the same room.  I find this true with opening bands versus headliners, or set 1 versus set 2.  You have to be ready to change directions if a path you take isn't working, liberal doses of common sense applied to what may sound like hard and fast rules, etc. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: yousef on September 23, 2019, 11:35:05 AM
Sounds as though you and I have developed our workflow opposite each other. I used to think time aligning everything wasn’t necessary, now I do it as a rule. The kit is actually the part I’m least worried about. Aligning vocal mics makes a world of difference IMO.

I wonder if there is some scope for uploading a song's worth of raw tracks and then seeing how different people mix them? Detailing the workflow, plugins etc...
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on September 23, 2019, 01:58:08 PM
I wonder if there is some scope for uploading a song's worth of raw tracks and then seeing how different people mix them? Detailing the workflow, plugins etc...
For many years I have anticipated a way to post audio projects publicly, with a default mix, which can be remixed as desired by listeners. A more advanced version could allow users to save their mixes and share them with other users.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on September 23, 2019, 05:21:57 PM
^
I’ve done something like that with Dropbox before. Put a template session with the files in a folder. Share the folder ad everybody just saves their own version of the sessions.

I might have a set of multis somewhere that I could throw up at some point. Wish I had saved one of the tauk shows I did, but they were so damn big...

For room mics I like to time align, pan wide, remove the center if possible, and sometimes sidechain to the level of the music. That would mean that the volume of the room will come up when the music is at a lower volume. Whether or not I do this depends on whether the aud mics capture the room ambience or really just the audience noise.


This is interesting, would like to try it. For removing the center are you just using a M/S plug and pulling down the mid?

Basically, yeah. I like the center plug by waves.

As I suspected, there is obviously no right and wrong approach to any of this, as there are some contradictory suggestions here that I'm sure yield equally pleasing results.  I suppose that's the nature of any project where the quality of the end product is subject to personal preference.

What may seem contradictory amongst everyone's suggestions is really the nature of mixing, what works on one thing I record may not work at all on the next thing I record, even if it's the exact same technical setup in the same room.  I find this true with opening bands versus headliners, or set 1 versus set 2.  You have to be ready to change directions if a path you take isn't working, liberal doses of common sense applied to what may sound like hard and fast rules, etc. 

Agreed wholeheartedly with this.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 07, 2019, 09:27:32 AM
After taking plenty of time to digest all of these recommendations and others, and after waiting more than a week for my Media Mail delivery of a book I purchased on Cool Edit Pro, this weekend I finally opened all of these tracks and started playing with it for the first time.  It took nearly an hour and a half to open them all, but once they were open I was surprised how little affect the open session had on my system (which functions as my Plex server).  I was a little concerned that my 16 GB of RAM would be stressed right from the start.

My initial observations and questions (in bold):

1.  I have mixed enough board feeds with my mics to be pretty comfortable at hearing time-alignment issues and resolving them through shifting or stretching (depending on whether the sources were captured with the same or different recorders).  I've spent a couple of hours listening to various parts of this show, playing with the mix, including plenty amounts of the house mics, and I can't hear any time alignment issues.  Should I trust my ears on this, or is it impossible for there not to be any necessary adjustment?

2.  I considered both viewpoints on whether to start with panning or end with it, and I like the argument that the perception of relative levels is affected by panning, so one of the first things I did was to move things around a little bit.  Keys are 30% right, drums 30% left.  When those musicians step to the front of the stage to play mandolin and bass, they're a little closer to center, so I panned those inputs (vox and instruments) only 20%.  The house mics (on-stage shotguns) I have panned 40%, and I currently have the center mic turned off because it seems to me to be a little more polluted with stage noise.  Whether this is a sound observation and reaction/judgment, I don't know.  But that's what I'm working with for now.  This sounds good to my ears, but I'll certainly consider tweaking it as the project evolves. 

3.  I've played with the levels a lot on different songs, and have settled on some pretty variable levels depending on the song, which underscores how much automation is going to be required to get everything just the way I want it.  Which is fine -- I have the time and patience to work my way through it.

4.  It took me a while to figure out all the effects terminology, especially in the multitrack workspace.  I think I'm figuring it out, but I'm wondering if I should be applying any of these in the track edit view (destructively to the source files) rather than layering them on top of one another in the multitrack.  Are system resources a consideration here?  Should things like the 100-150 Hz high pass just be applied to the source files?

5.  On the topic of effects, it took me until the end of the weekend (late last night) to realize that I seemed to be a little crippled using Cool Edit Pro 2.1, in that it appeared that it was going to prevent me from using a lot of VST plugins.  So the last thing I did last night was to find an old free version of Audition 3, which added a lot of VST compatibility.  All I've done so far is open a few tracks to ensure that the interface was close enough to Cool Edit Pro to look familiar to me.  In the process, I discovered that the tracks that Cool Edit Pro took 2.5 to 3 minutes each to open were each opened by Audition in 30 seconds or less.  I'm not sure if there's a good reason for that, or if it just means that Audition, by default, monopolizes more of the system resources and it's going to cripple my Plex server when I open all the tracks and resume work on the session.

Are there any particular amp sims that I should consider for the DI tracks?  Are these free plugins?

Thanks for all of the advice so far.  I think this is going to sound incredible when it's finished.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on October 07, 2019, 09:59:16 AM
Glad to hear you’re having fun! Time alignment in multis like this is most likely not something you would hear. The mics are only separated by a handful of feet, so the issues are more in phase than slap back (like you would hear aligned an SBD and AUD mics.)

I wouldn’t do any destructive editing unless you absolutely have to. It doesn’t sound like your system is having an overly difficult time, so I would just leave as is for now.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 07, 2019, 01:59:22 PM
I wouldn’t do any destructive editing unless you absolutely have to.
You could consider destructive editing on a COPY of the master if you have something like a CLICK from a cable being plugged in, or the like, but don't do any major processing to the whole file destructively.
Does your new Audition take the same amount of time to re-open an existing session as it does to create one? Audio is not really a hard problem for most modern computers anymore, it should not drag down your server unless you're competing for resources directly, like disc access or internet throughput. Servers should not be overly CPU-intensive (and audio only should be when it's rendering or actively processing, and even that should not totally hose your server, though you may notice a performance difference.)

Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 08, 2019, 06:49:15 PM
Time align all the tracks to the snare. This will help small phase issues that you might otherwise not really hear, but can have a great affect on the overall tone of the tape. Summation is a very powerful tool.

What if the tracks that should need the most alignment (the on-stage shotguns pointed at the crowd) are too difficult to discern the snare, much less down to the sample level like I'm accustomed to evaluating when mixing SBD and AUD sources?  I feel like there's more potential error in eyeballing this than there is in the actual alignment.  Do you ever resort to just calculating a theoretical approximate delay and shifting accordingly?  There are video clips of the show on YouTube, so I know about where the shotguns were placed.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on October 08, 2019, 08:46:09 PM
Time align all the tracks to the snare. This will help small phase issues that you might otherwise not really hear, but can have a great affect on the overall tone of the tape. Summation is a very powerful tool.

What if the tracks that should need the most alignment (the on-stage shotguns pointed at the crowd) are too difficult to discern the snare, much less down to the sample level like I'm accustomed to evaluating when mixing SBD and AUD sources?  I feel like there's more potential error in eyeballing this than there is in the actual alignment.  Do you ever resort to just calculating a theoretical approximate delay and shifting accordingly?  There are video clips of the show on YouTube, so I know about where the shotguns were placed.

That’s a perfectly fine way to do it.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 09, 2019, 09:06:33 PM
For anybody that uses or is otherwise familiar with Audition, is there a "preferred" way to achieve a high pass filter (say, 150 Hz, as I think was recommended earlier) on all the tracks to get rid of anything non-musical?  It seems I can apply it in different ways in different places.  For example, as an effect in the effects panel (parametric my best choice here?), or two panels down in the mixer view is the EQ panel with filtering capabilities, it seems.  Best way to do this?  Is this where I use an effects bus?  See screenshot linked below.

https://imgur.com/HEIVdsF (https://imgur.com/HEIVdsF)
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 10, 2019, 12:42:50 AM
For anybody that uses or is otherwise familiar with Audition
Well, firstly, that's NOT me. But I suspect the Parametric EQ might have a way to turn the lowest filter to "shelf" mode and set that where you want, with the curve you like?
As a rock&roll mix engineer, I would suggest doing this channel-by-channel. Electric guitar can have a lot of 60 Hz in it, but should not have any 30!? Electric bass has a FUNDAMENTAL at 31 HZ in cases of a low B string, like on a 5 or 6 string bass... don't roll that off, please. Kick drum, on the other hand, needs a low frequency SLAM and a high frequency SNAP (some people don't have ANY bass speakers and they wanna dance) but you can roll a lot of the sub-bass out of that SLAM and it can still massage your sternum at proper power settings.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 11, 2019, 10:18:46 AM
For anybody that uses or is otherwise familiar with Audition
Well, firstly, that's NOT me. But I suspect the Parametric EQ might have a way to turn the lowest filter to "shelf" mode and set that where you want, with the curve you like?
As a rock&roll mix engineer, I would suggest doing this channel-by-channel. Electric guitar can have a lot of 60 Hz in it, but should not have any 30!? Electric bass has a FUNDAMENTAL at 31 HZ in cases of a low B string, like on a 5 or 6 string bass... don't roll that off, please. Kick drum, on the other hand, needs a low frequency SLAM and a high frequency SNAP (some people don't have ANY bass speakers and they wanna dance) but you can roll a lot of the sub-bass out of that SLAM and it can still massage your sternum at proper power settings.

Thanks for the insight.  I've just upgraded software for the second time in three days (CEP 2.1 > Audition 3 > Audition CS6).  Not only am I finding much better tutorials for the tools available in CS6, but it's lightning fast compared to v3.  I was surprised that v3 opened the tracks in about 30 seconds each (compared to more than 2 minutes each in CEP 2.1), but after that the program was extremely sluggish in trying to move things around and stuttered quite frequently on playback.  CS6 is flying by comparison.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 12, 2019, 09:52:27 AM
I've been reading some articles on drum panning.  There seems to be fairly consistent advice to keep the kick and snare panned dead center (along with bass guitar).  It seems popular to pan the rest of the kit not together, but in varying amounts according to the kit layout (to this point I had the entire drum bus panned L30).  At this point there seems to be two schools of thought -- relative panning of the kit from audience perspective or from drummer's perspective.  At first I thought the idea of drummer's perspective must be for studio recording, but at least one article discussed this for live mixing as well.  Any thoughts on this?  Recall that the drum kit is stage right for this show, so, other than the bass and snare, I'm trying to figure out a good starting point for how far left I should pan these relative to one another, and whether to consider audience or drummer's perspective for the relative amounts.

For reference, here's a video clip from the show that shows the stage layout fairly well.  Best look at the drum kit is a minute into the video.

https://www.youtube.com/watch?v=PcfV_OhfLuw
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on October 12, 2019, 01:35:24 PM
Drum mixing is all personal preference. I don’t like drummer perspective because I like to think I’m looking at the drummer, particularly with live recordings.  Sometimes I will mix drums off to the side, but it’s very band dependent and pretty rare. I do post work for a jazz venue that requests all their shows to be mixed as they were staged, so drums are always out house right
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 13, 2019, 06:29:37 AM
Drum mixing is all personal preference. I don’t like drummer perspective because I like to think I’m looking at the drummer, particularly with live recordings.  Sometimes I will mix drums off to the side, but it’s very band dependent and pretty rare. I do post work for a jazz venue that requests all their shows to be mixed as they were staged, so drums are always out house right
Under the auspices of The Soundscape Preservation Society, a concern I continue, I would say historical accuracy is prime. But with a multi track, that's all out the window. Go for the best damn mix you can make. One drummer? Kick/Snare center (or snare just slightly off to the side where the drummer has it placed) and the rest panned as wide as you'd like the drums to be. Are they mono and narrow or do they spread out all over the soundstage? I should probably not claim too much expertise in soundstage, cause I don't dig in very hard, but I'm aware that you can change it!!
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 13, 2019, 06:51:08 PM
I'm working through some of the recommendations (here and elsewhere) for the vocal tracks.  I'm saving EQ (to get the lead and backing vocals playing together nicely) for last.  For now, I'm addressing compressing and limiting (in that order).  A couple questions:

1.  Am I overlooking a way to know/observe the outcome of the compressor so that I can use that to determine what peaks remain to be limited away?  Or do I need to render these tracks individually for the purpose of inspecting the resulting waveform and deciding on where to limit it?

The keyboard, drum, and stage right vocal tracks contain only vocals (plus ambiance).  However, the lead vocals also contains a harmonica on one song, a kazoo on part of another song, and the last song of the night is all three band members singing and playing into it (tambourine, accordion, plus three vocals).  Meanwhile, the stage left vocal track is, for the most part, the bass vocals (when the keyboard player switches to bass for about 1/3 of the songs).  However, like any good combination keyboard and bass player, right in the middle of the show he drags a tuba from off the side of the stage and plays it into this mic.  See waveform here:

https://imgur.com/XFzGdOA (https://imgur.com/XFzGdOA)

Most of the waveform is just room ambiance while he's playing keyboard.  Most of the peaks are his vocals while playing bass (these dynamics seem tailor-made for the compressor and limiter).  Right in the middle is the tuba. 

2.  How does the presence of non-vocal parts impact the way that I should be treating these dominantly vocal tracks?
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: EmRR on October 13, 2019, 10:23:34 PM
EQ will change the behavior of compression.  You want to do that the other way round IMO. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: goodcooker on October 16, 2019, 03:10:13 PM

I've only done a few multitracks of live shows (more than 4 inputs) and I agree with the comment above that it's a personal preference and would do whatever sounds best. I like to mix the snare, hihat and kick center then pan the overheads.

I've been doing a lot of stagelip plus sbd recordings of Johnny Vidacovich lately and his kit is always stage left. I put the mics on the stagelip off center to stage right and orient them to cover the kit. This gives good stereo coverage and keeps the image like it was live while also covering the entire stage.

If you have individual mics on the toms you might want to pan them a little but to be honest if I have overheads and they sound decent I never use the tom mics. It's not a Rush show.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on October 16, 2019, 05:37:49 PM
EQ will change the behavior of compression.  You want to do that the other way round IMO.

I'm in agreement with this, generally. I like to get EQ in the ballpark prior to wrangling too much with dynamics.  Then I approach it iteratively, going back to check how things may have changed, and converging toward an optimal solution. The last thing I typically do is touch up the master output bus EQ.

For both EQ and dynamics I find it helpful to break the tasks into different stages which address different aspects.  Early on, basic gross EQ and dynamic corrections are made.  Later, subtle fine-tuning takes place.  And with respect to dynamics there is also the contrast between high level peak dynamic management on one hand and lower level average loudness management on the other - two very different things.
Simplified, consider the differences between limiting vs parallel comp.

Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 17, 2019, 02:09:26 PM
I'm working through some of the recommendations (here and elsewhere) for the vocal tracks.  I'm saving EQ (to get the lead and backing vocals playing together nicely) for last.  For now, I'm addressing compressing and limiting (in that order).  A couple questions:
1.  Am I overlooking a way to know/observe the outcome of the compressor so that I can use that to determine what peaks remain to be limited away?  Or do I need to render these tracks individually for the purpose of inspecting the resulting waveform and deciding on where to limit it?

The sort of sideways answer to your question 1 is... drumroll... EQ! It's not that will let you observe so much as control the outcome. EQ in a signal chain before compression allows you to select what frequency bands will trigger the compressor. (whatever has hot enough signal levels to hit the threshold and trigger the comp)
This can be done on a side chain, if you don't want those frequencies to actually be boosted. EQ the side chain feed to trigger the comp where you want but EQ the actual signal differently, or not at all!?


As for tracks with lots of things in them- you COULD copy the track and delete all the vocals from the kazoo one, and delete all the kazoo from the vocal one... if that's possible, and process them differently. Or not worry about it and see how far you get with that approach?
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 19, 2019, 01:09:57 PM
If anybody has an interest in critiquing a song as I currently have it processed/mixed, I'd be interested to know if I'm way off base on anything.  Obviously everybody's taste is different, but I'm looking for gross negligence -- things that everybody can agree need adjusting.  I've posted it here in uncompressed 24/48 .wav format:

https://soundcloud.com/user1965380/milwaukee/s-Mrtwq

This song contains acoustic guitar, drums, and two different keys (melody and bass, respectively).  It's their longest song because it always contains an improvisational spoken part in the middle.  I thought maybe the variety would make it a good song to critique.

Some things I've done:

EQ to the kick tracks (bussed)
EQ and light compression to the snare tracks (bussed)
EQ to the hat
fairly sizable high pass on the OHs, but then lowered these in the mix relative to the newly-treated kick/snare/hat
Nothing to the toms
Nothing to the ride
Nothing to the acoustic guitar
Light compression/limiting on the keys was because of this song, where they get a little loud at the end
No other processing of the keys (amp sims, etc)
All vox rolled off at 50 Hz, de-esser, and light compression and reverb
No additional processing of the master track

I've turned off the center audience mic because it's placed right over the stage monitor and is pretty polluted from that.  I've got the others panned fairly wide.  I've been playing with fading up the center AUD mic only between songs (or otherwise when it's mostly pure AUD sound), and I've liked the result of that.  Is there any technical reason not to do this?  I didn't do that on this mixdown, as I haven't started playing with envelopes yet.  I'm simply playing with fading it up while monitoring in real-time.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 19, 2019, 08:08:51 PM

This sounds good! The only thing that jumps out at me is that the waveform appears to have a taller (asymmetrical) "top part" - that is, the positive excursion seems to have a greater absolute value than the negative excursion.
There could be a number of different reasons for this, it might be unimportant, but you might take a glance at the individual tracks, and see if there are more than one with "tall top parts" and consider flipping polarity on one or more of them? Just to get more "area under the curve" if possible. (it will use the amplifiers more efficiently and thus might have a bit better signal:noise)
edit: OK made it to the end. #skintags <- ew
I don't think you need quite that much piano at the very end, though there may not be much else to mix. We know it's a piano solo. He said "play your piano" or whatever. It's ok, it just seems like a dry SBD at that (singular) point in the show.
edit2: >I've been playing with fading up the center AUD mic only between songs (or otherwise when it's mostly pure AUD sound), and I've liked the result of that.  Is there any technical reason not to do this?
Nope, just be sure that it sounds natural. Check your fades in headphones and on speakers and don't be afraid to fade in before the end of the song if the band is blowing it out.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 19, 2019, 09:40:03 PM
Thanks for taking the time to listen to it.  I feel better knowing that it's not an absolute mess to somebody's ears that are more tuned than mine to this kind of work.

The only thing that jumps out at me is that the waveform appears to have a taller (asymmetrical) "top part" - that is, the positive excursion seems to have a greater absolute value than the negative excursion.

I noticed that it looked that way on the Soundcloud representation of the waveform, but it looks normal to me when I open it in anything else.  Is the Soundcloud waveform supposed to mean anything?

https://imgur.com/1Smq6nA (https://imgur.com/1Smq6nA)

I don't think you need quite that much piano at the very end, though there may not be much else to mix. We know it's a piano solo. He said "play your piano" or whatever. It's ok, it just seems like a dry SBD at that (singular) point in the show.

I was hoping that the little bit of compression and limiting that I applied would have tamed that a little better.  Perhaps I just need to envelope that down a little bit.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on October 20, 2019, 08:46:54 PM
Agreed about the piano. Usually when I'm mixing a show I'll basically play through and do a live 'mix' of it, riding automation levels as if they're faders on a console.

Personally I think it could use some less overheads and more drums in general. I felt like the snare got lost in the acoustic a bit. Might be able to fix that by carving a little bit out of the gtr. I think it could also benefit from a bit more 'cohesion' between the different tracks. It's a hard time to accomplish, but is sometimes helped by sending everything through the same processing, whether that be a little bit of the same room verb on all the instruments, a little bit of master bus compression, or something else, it depends on the tape
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 20, 2019, 10:11:33 PM
Agreed about the piano. Usually when I'm mixing a show I'll basically play through and do a live 'mix' of it, riding automation levels as if they're faders on a console.

This is my planned strategy.  My goal is to come up with good baseline levels for the majority of the show, from which I can automate levels up and down as needed.

Personally I think it could use some less overheads and more drums in general. I felt like the snare got lost in the acoustic a bit. Might be able to fix that by carving a little bit out of the gtr. I think it could also benefit from a bit more 'cohesion' between the different tracks. It's a hard time to accomplish, but is sometimes helped by sending everything through the same processing, whether that be a little bit of the same room verb on all the instruments, a little bit of master bus compression, or something else, it depends on the tape

I'm surprised (but don't doubt, given your experience) that you can hear the drum overheads, as they average in the -30s on that song and only peak around -18db.  One of the things that I've read a lot about, but have struggled to implement to this point, is the whole idea of carving out frequencies to let certain instruments shine.  For example, the kick drum vs. the bass guitar.  I haven't been able to identify to my satisfaction the right place to boost/cut either of them in order to give them their own little frequency space to dominate.  Same thing with lead vs. backing vocals, and I assume the snare/acoustic conflict you hear would be resolved similarly by a professional.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 21, 2019, 12:24:29 AM

One of the things that I've read a lot about, but have struggled to implement to this point, is the whole idea of carving out frequencies to let certain instruments shine.  For example, the kick drum vs. the bass guitar
Well I know that one, roll the very lows out of the kick so the bass guitar doesn't get stepped on. But in general I'd just make certain each instrument sounds good on its own, and try not to EQ more than necessary.


>Is the Soundcloud waveform supposed to mean anything?
No idea, I've never posted there.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on October 21, 2019, 11:51:44 AM
Sounds good here.  Nice work.

What caught my ear on first listen: Acoustic guitar and keys are a bit dry at some points (sounding somewhat "soundboardy"), and drums are a bit low/distant at some points. I imagine the drum presence, sense of proximity and detail would respond really well to some parallel comp.  EQ seems good, even though I'm not using headphones I trust to really assess that kind of thing. The audience sound is good, and I see no reason not to ride the center audience mic during the breaks if you like, especially when it works nicely like it does here to increase the feel of live realism.  I sometimes do something similar with my rear-facing audience/room pair.  The sort of baseline essense of the recorded audience and room ambience is picked up by my hard-panned wide omnis.  To that I'll add some center-rear-facing mic or a stereo-pair of rear facing mics to sort of more evenly fill the ambient space between those hard-panned positions.  Those channels are generally lower in level and respond well to manual envelope or side-chain control to brings them up in level during quieter sections and between songs. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 21, 2019, 12:57:07 PM
Sounds good here.  Nice work.

What caught my ear on first listen: Acoustic guitar and keys are a bit dry at some points (sounding somewhat "soundboardy"), and drums are a bit low/distant at some points. I imagine the drum presence, sense of proximity and detail would respond really well to some parallel comp.  EQ seems good, even though I'm not using headphones I trust to really assess that kind of thing. The audience sound is good, and I see no reason not to ride the center audience mic during the breaks if you like, especially when it works nicely like it does here to increase the feel of live realism.  I sometimes do something similar with my rear-facing audience/room pair.  The sort of baseline essense of the recorded audience and room ambience is picked up by my hard-panned wide omnis.  To that I'll add some center-rear-facing mic or a stereo-pair of rear facing mics to sort of more evenly fill the ambient space between those hard-panned positions.  Those channels are generally lower in level and respond well to manual envelope or side-chain control to brings them up in level during quieter sections and between songs.

Thanks for your time.  Actually, one of the things I've done since I uploaded that sample (in addition to boosting the snare and the entire drum bus a bit) was to discover the "Guitar Suite" of effects in Audition.  I've started playing with just a little bit of a warm amp sim on the acoustic.  I'm still in the process of dialing it up and down to see where I like it.  The next task is to hopefully discover a similar amp sim suitable for the keyboard tracks, and I was also going to experiment with the bass to see if there's something I can apply to the DI track that is more pleasing than the mic track I have for that one.  Anybody have any recommended free plugins I should try, either for the keys (which may not be available in Audition natively...still need to search) or in place of the Guitar Suite I'm currently playing with?
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on October 21, 2019, 03:28:33 PM
I don't have any specific plugin suggestions, but some kind of amp or acoustic guitar sim plugin sweetening could be just the thing well on the guitar. I hear the classic "directly plugged in dry piezo-pickup quack", which is so common it's become a sort of widely recognized live acoustic guitar sound some folks actually prefer, yet it always sounds like a piezo to me.  A touch of 'verb might help as well, maybe something like a plate verb with the lows cut to it. 

In terms of dry/wet verbish-ness, I think it helps to listen to the entire mix while directing attention to each instrument individually, determining if they all seem to inhabit the same imaginary acoustic space.  That's not to say some things shouldn't be drier clearer and perceptually closer sounding while and others are more 'verby and perceptually distant, or that they should all sound spatially similar in isolation, rather that they all have some discernible sense of spatial commonality.  It's hard to describe this, but it's something that sort of helps to glue all the parts together in such a way that it isn't perceived as a multi-track mix of close-mics but rather a perfectly balanced AUD.

BTW the piano push at the end sounds appropriate to me in terms of it's loudness and timbre, but the lead and bg vox portion leading up to it after the quiet section is buried somewhat and could use some comp or a bit of bump in level.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 22, 2019, 09:21:59 PM
I'm trying to do some housekeeping in my session.  For starters, I'm setting up a reverb bus to send all the vox tracks to in place of having reverb on all the tracks.  I plan to experiment with sending many of the instrumental tracks through this as well to see if I can get a more cohesive sound coming out the other end.  I've reached a roadblock that I don't want to barge through or detour around without knowing how it's working.  I can't for the life of me find an answer to what I imagine is an elementary question.  I understand the concept of wet vs. dry in terms of the balance of processed vs. unprocessed signal.  It seems most of the effects that utilize this as a metric have a single slider with wet and dry on the extreme ends.  The regular "Reverb" effect in Audition CS6 has separate Dry and Wet sliders which go from 0 to 200% for Dry and 0 to 500% for Wet.  It's obvious how to manipulate the sliders to get one extreme or the other, but it's not clear to me what the purpose is of having independent sliders here. I'd like to understand the subtleties of this.  See the default "Room Ambience" settings below (my chosen starting point, for better or worse).  Hopefully this is similar to how it's handled in other DAWs and is easily explained.

https://imgur.com/qZrarXj (https://imgur.com/qZrarXj)

Second -- with the reverb applied to the individual tracks, I adjusted the reverb effect settings to control how much reverb I wanted.  When sending these through the dedicated reverb bus, it seems I have two choices:  1) set up the reverb like I did before and send all the signal to it; or, 2) set up a more extreme reverb and then calibrate how much each track gets via the send control knob.  Is there a best practice here?  Pros/cons?  To end up with just a little reverb in the end, is it best to give it "a little of a lot" or "a lot of a little"?
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: opsopcopolis on October 23, 2019, 09:06:52 AM
The range on those we/dry sliders is bizarre. When bussing verb, you want it to be 100% wet. You determine how much verb you want by how much of each instrument you send to the bus.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 23, 2019, 09:23:51 AM
The range on those we/dry sliders is bizarre. When bussing verb, you want it to be 100% wet. You determine how much verb you want by how much of each instrument you send to the bus.

Thanks -- that answers one big question.  Can anybody speculate on the purpose of the separate dry slider, then?  Do I necessarily just put that at 0%?  Up to this point, and in the sample posted earlier, the reverb I had applied directly to all the vox tracks was about 20% wet and 80% dry, since I had made up my mind to ignore the goofy ranges and make my selections sum to 100%, hoping that would effectively mimic a single wet/dry slider.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on October 23, 2019, 11:50:46 AM
Lost an entire post here this morning about the routing options and don't have time to retype it.. maybe later.

In essence (without the fluff and parenthetical ifs and buts):
If using effects bus routed 'verb set it to 100% wet as opsopcopolis states.  This routing scenario gives you an effective wet range adjustment of 0-50% (parenthetical hand-waving)

I can only speculate that the very odd adjustment-range of the sliders in the plugin affects positive gain for settings above 100%.. making for an output gain increase. Sort of like the make-up gain control on a compressor.  /speculation.

With regards to the second question, the other 'verb settings are what affect the character of the 'verb.  The amount applied is all about wet/dry ratio and levels.  The end effect should be the same within the available limits of adjustment regardless of where those levels are set, as long as gain staging isn't crazy off base causing noise-floor issues. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 23, 2019, 04:05:26 PM
This routing scenario gives you an effective wet range adjustment of 0-50% (parenthetical hand-waving)

Can you elaborate on this when time permits?  In Audition, the send level is not in percentages, but goes from -∞ to +15.  In addition, I'm trying to conceptualize in my head the impact of sending signals pre- vs. post-fader.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: EmRR on October 23, 2019, 05:05:25 PM
This routing scenario gives you an effective wet range adjustment of 0-50% (parenthetical hand-waving)

Can you elaborate on this when time permits?  In Audition, the send level is not in percentages, but goes from -∞ to +15.  In addition, I'm trying to conceptualize in my head the impact of sending signals pre- vs. post-fader.

A mix setting on reverb is usually for times when you use it as a plug inline on a mix channel, there it's in parallel if mixed 1-99%.  If it's bussed and sends feed it, it's in parallel with the signal channel and thus should be 100% wet.

Send levels on a whole group of vocals should probably start at unity gain on the send, and set the return fader where it sounds right.  That gives a good starting spot for other things to be lower if needed, and room to boost if something is quiet. 

You usually want send post-fader so they follow the fader. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on October 23, 2019, 06:48:10 PM
^Thanks.  I'd started this reply then got pulled away.  EmRR answered more concisely above but I'll post this anyway. Basically the same info stated differently in case that helps.

This routing scenario gives you an effective wet range adjustment of 0-50% (parenthetical hand-waving)

Can you elaborate on this when time permits?  In Audition, the send level is not in percentages, but goes from -∞ to +15.  In addition, I'm trying to conceptualize in my head the impact of sending signals pre- vs. post-fader.

Sure. The dry and wet signals are routed in parallel and combined in the mix bus, so if each is at the same level, that's 50% wet. My simplified scenario assumes a -∞ to +0dB send level range.  You can actually go more than 50% wet since the send range goes up to +15dB gain instead of 0dB (unity gain).

Consider the signal routing through a channel strip in the mixer.  Channel level is adjusted using the fader.  After that the signal is mult'd and sent two different places: directly to the mix bus (the dry signal), and the effects send (if setup to be post-fader, which is typical for mixing, not sure of the options in Audition).  Because that split done post-fader, the send level output tracks along with the channel output level to the mix bus as the fader is adjusted - such that the balance between the two (the dry/wet ratio) remains constant as the fader is adjusted up/down.   If the send level is set to unity (0dB), you are sending the same level of signal directly to the mix bus (100% dry) and to the effects bus 'verb (100% wet), making the combination in the mix bus 50% wet.  The signal paths are in parallel. This ignores any adjustment of the effect return level or gain the 'verb plugin introduces (assuming those to be unity-gain).

If you turn down the effects send on the channel strip, you reduce signal level sent to the 'verb, which reduces the combination in the mix bus to something less than 50%.  Yet since your send goes up to +15dB you can actually send more signal through the effects bus (wet) than directly out to the mix bus (dry), increasing the ratio to more than 50% wet it you wanted to do that.  ..and
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 24, 2019, 05:39:12 PM
You usually want send post-fader so they follow the fader.
^^^ I am seeing very good advice from these cats here, but I want to post and specifically agree with this part.
Pre-fade sends might typically be used to make a monitor mix for performers to hear what they are playing, mixed with previously recorded tracks. Then once that mix is set, you can keep using your faders to create the main mix without changing anything on the monitor side.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: EmRR on October 24, 2019, 07:32:28 PM
You usually want send post-fader so they follow the fader.
^^^ I am seeing very good advice from these cats here, but I want to post and specifically agree with this part.
Pre-fade sends might typically be used to make a monitor mix for performers to hear what they are playing, mixed with previously recorded tracks. Then once that mix is set, you can keep using your faders to create the main mix without changing anything on the monitor side.

Pre-fade can be good for special effects in a mix, like making a vocal disappear into a sea of reverb only. 
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on October 24, 2019, 11:42:58 PM
Pre-fade can be good for special effects in a mix, like making a vocal disappear into a sea of reverb only.
TRUTH!
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on October 28, 2019, 07:24:44 PM
I've just about completed my first comprehensive round of volume envelopes (basically, just turning instruments and mics "on" and "off" as necessary for each song).  For DI tracks, I'm not worrying about fading them in/out too much.  For the mic tracks, I'm sometimes fading in/out over a minute or more to make the transitions as imperceptible as possible.  Once I'm done with this, I'll start listening to the whole thing on various speaker systems and making notes of where to further adjust things (piano down here, lead vox up there, etc.).  At that time I'll probably also solicit opinions on a new mixdown of the track I already shared, and maybe another song with different instrumentation.  In the meantime, a couple questions:

1.  The show starts with the drummer playing banjo on a version of Seven Bridges Road.  This is the only time the banjo is used.  This track is very quiet, and I've got the fader pushed up to +13 (/15) to get it where I want it.  Am I better off adding amplification to that track via effects so that I don't have to deviate that much from unity gain on the fader?  Or is the fader applying the gain just as cleanly?  (Full disclosure:  I already added amplification to the BassDI track because it was so quiet, so your answer here will also advise me on whether to treat that track differently).

2.  On one song, Stephen (lead singer) plays the piano at stage left.  I have the piano vox panned R40 for the show.  For this song, would most of you bring the vox back to center or leave them panned?  I assume this is a personal preference thing, so go ahead and cast your vote.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on November 04, 2019, 09:06:31 PM
I've worked quite a bit more on this and have gotten things working together better by passing most of the instrumentation and vocals through the same room reverb.  The drums have their own plate reverb.  I've also boosted the drums and completed a lot more volume automation of the backing vocals.  If anybody has the time to check out a couple songs, I've posted a couple with different instrumentation below.  Comments on anything your ears do or don't like are welcomed.

https://soundcloud.com/user1965380/4th-of-july/s-PqAda (https://soundcloud.com/user1965380/4th-of-july/s-PqAda) (acoustic guitar, drums, keys)
https://soundcloud.com/user1965380/wagon-wheel/s-UNZ18 (https://soundcloud.com/user1965380/wagon-wheel/s-UNZ18) (electric and bass guitar, drums, keys)
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on November 05, 2019, 12:01:32 AM
2.  On one song, Stephen (lead singer) plays the piano at stage left.  I have the piano vox panned R40 for the show.  For this song, would most of you bring the vox back to center or leave them panned?  I assume this is a personal preference thing, so go ahead and cast your vote.
Move lead vocals to center. Other key Vox can be located in correct relative stereo position.
Congrats on getting your rough fader mix finished! That's major. I'll take a listen after I listen to what I'm listening to... As for boosting levels, since I use Audacity which has a limited range of gain, sometimes I'll just copy a really low part to a brand new track, and just run the soft parts doubled (or more!) to get more gas on that track.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on November 05, 2019, 09:26:43 AM
Acoustic guitar and drums sound and sit right, and no longer draw my critical ear on the 4th of July track.  Sounds good.  Audience reaction is good and natural sounding, but perhaps a bit overly hard-panned left/right.  I consider it is good to have audience/room-ambience portrayed very widely, enhancing the sensation of envelopment and sort of keeping from overly infringing on the musical elements in the center, but a bit more across the middle helps tie it together so it doesn't sound like two separate "pools" of audience hard left and hard right.

Nice job.

Haven't had a chance to listen to the second clip.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on November 05, 2019, 09:36:45 AM
Acoustic guitar and drums sound and sit right, and no longer draw my critical ear on the 4th of July track.  Sounds good.  Audience reaction is good and natural sounding, but perhaps a bit overly hard-panned left/right.  I consider it is good to have audience/room-ambience portrayed very widely, enhancing the sensation of envelopment and sort of keeping from overly infringing on the musical elements in the center, but a bit more across the middle helps tie it together so it doesn't sound like two separate "pools" of audience hard left and hard right.

Nice job.

Haven't had a chance to listen to the second clip.

Good catch on the audience.  I've been doing a new mixdown each night and listening to it at work the next day, making notes, and then making those adjustments.  Last night's trial was to see what it would sound like if the audience was panned 100% left/right.  I was theorizing that maybe each mic might have picked up just enough from the other side to not sound like two different audiences.  I think I originally had them at L50 and R50 respectively, so I'll return them there unless you think there's an in-between that is just right.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on November 05, 2019, 10:21:17 AM
I was theorizing that maybe each mic might have picked up just enough from the other side to not sound like two different audiences.  I think I originally had them at L50 and R50 respectively, so I'll return them there unless you think there's an in-between that is just right.

How hard to pan will depend on the audience/ambient mic arrangement and how much common info those channels share, so best to just dial it in by ear in the mix.  If I have 3 recorded ambience type channels I'll hard pan the L/R ambience pair to keep elements of that very wide, then bring up just enough rear-facing center ambient mic to to tie it all together with the appropriate smooth blend across the center.  Most of the energy remains wide, because I find having some audience/ambience cues wider than anything else helps convey a more dimensional sense "there-ness".  Yet the overall perception should be of audience continuity across the soundstage.

With 2-channels of audience/ambience you need to find the right balance between super-wide and continuous-without-a-perceptible-hole-in-the-middle.  I suspect you can go wider than L50/R50 without perceiving a hole and would push it until you begin to notice hole then bring it back in just a bit from there.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on November 20, 2019, 09:56:37 AM
I finally reached the point of calling this "good enough".  Anyone interested can check it out here:

https://archive.org/details/skellogg2008-11-21.flac24

Thanks again to everybody who spent any amount of time giving advice, answering questions, and providing feedback along the way.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on November 20, 2019, 11:33:30 AM
Sounds very good here. Strong work!

Was just now jumping around to sample a bit of it, and while listening through the In Front of the World ballad where the audience and ambience contribution is very clear I came across his comment about how distracted audience conversation makes the band feel like dogshit.  Gold.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: mattmiller on November 20, 2019, 12:16:06 PM
Was just now jumping around to sample a bit of it, and while listening through the In Front of the World ballad where the audience and ambience contribution is very clear I came across his comment about how distracted audience conversation makes the band feel like dogshit.  Gold.

Stephen has a history of not tolerating crowd noise during inappropriate times, or in any way that is distracting to him.  In fact, this show was my second time seeing them, having traveled all the way to Philadelphia from NE Ohio to do so.  My first show was the month before in Cleveland where a handful of people in the crowd triggered him to the point of him asking them to leave, offering to pay them to do so, and then offering anybody's money back who wanted to leave to escape the awkwardness that the show had turned into.  I became an instant fan, which prompted the trip to Philly, in which I was turned down for taping by the venue since the band was multi-tracking it, wishing for more than a decade to have that recording that I knew was just sitting on a hard drive somewhere, and then suddenly it appeared out of nowhere in my lap to mix.  It all worked out in the end, I guess.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: Gutbucket on November 20, 2019, 01:05:40 PM
Right on. Cool to hear the back story.
Title: Re: Mixing Multi-Track Recording (29 Tracks)
Post by: morst on December 02, 2019, 05:05:30 AM
...  I became an instant fan, which prompted the trip to Philly, in which I was turned down for taping by the venue since the band was multi-tracking it, wishing for more than a decade to have that recording that I knew was just sitting on a hard drive somewhere, and then suddenly it appeared out of nowhere in my lap to mix.  It all worked out in the end, I guess.
Patience FOR THE WIN!
Great backstory. I'll download it and add to my library, maybe throw it on my phone to show up in the shuffle, so I can really compare it blind to other stuff I like.