Livesets

@seagreg

Would this be for cases where you load up several Unify layers, but then only use one or two at a time? This seems to be a popular technique in worship music.

In one of his videos John loaded an additional layer while he was playing . That’s nice - specially if it happens seamlessly - but normally you don’t have the time to look for another layer if you’re playing live. You have to be fast - in any kind of music I think. So tweak before and than solo or mute. 

@seagreg

What John has shown in videos is basically a method for building up multi-layered patches using embedded Unify layers. Neither he nor I consider this adequate for live performance, but now that we can do this, it points the way to more advanced approaches.

As far as I know, people use other programs like Gig Performer or Cantabile in either of two main ways:

1. Load several layers at once, then selectively enable/disable them, and/or adjust the layer mix. This is common among church musicians.
2. Create a list of patches, then switch from one to the next. (A patch list could either be a set list, or a sequence of patches used in one song, or some hybrid of the two.) The software can then use the list to load patch N+1 silently in the background while patch N is being played, then execute a seamless transition in response to some user input, by switching input MIDI over to patch N+1 while allowing patch N to continue sounding (e.g. delay and reverb tails) until its output level drops below a threshold.

If anyone knows of other approaches, I'd like to hear about them.

Posted by: @getdunne 

If anyone knows of other approaches, I'd like to hear about them.

Approaches with Gig Performer and Cantabile or live in general?  Mainstage is also an extremely popular tool for church musicians.

As I mentioned in another thread, I have used the MIDI Channel approach.  My background is in hardware synths so it was a familiar way for me to work live.  When I started moving to software, I continued with the MIDI Channel approach as that is how I think.  I load up all my layers/sounds at once and assign MIDI Channels.  Then program my keyboards appropriately.  Unfortunately, most MIDI controllers these days don't really work quite like the hardware workstations do, which is one reason I've remained with workstation type of keyboards/synths.

I basically use a DAW just for my sounds.  I would definitely use Unify standalone in this way as it can do exactly what I need.  The issue being, my needs have also changed once I added the Native Instruments S61 MK2 (and I can no longer simply stick to assigning MIDI Channels).  It is a beautiful combo with Ableton Live and loses that functionality outside of using Live.  The S61 is now an integral part of my setup.

I'm not really a fan of program changes as I can't afford to have any hiccups (load times) when switching patches.  I need immediate access to a sound when I press it.  I also can't have the previous sound suddenly cut off when switching to the new patch.  What I may use is if there was some type of holding tank for ALL sounds in a set, and I can swap them in/out using program change.  Then I can continue using the S61 and assign the buttons to program changes.  I will lose out on the S61 Mixer view (which is actually one of my favorite parts of the integration with Live).  All sounds would share the same fader, which may not be ideal either.

Major things when I play live is:

• All sounds transition seamlessly.  That is, no breaks/cuts when switching patches.
• When I switch to another patch, I need it there and ready to go.  Can't afford a load time.
• The fewer taps I need to get to a sound, the better.  Ideally, one tap.

I don't know if this helps, but it's how I prepare and create my set assignments.  I know others who use a completely different approach.  I've now been using Unify live for several months and haven't had any issues.  I mentioned today during the live stream I had a piano/pad layer I was using earlier this week and it was crackling.  I moved the Omnisphere pad into Unify and the issue disappeared.  I used that patch live this morning without any glitches.

@ssquared

These are the kinds of insights I'm looking for; thank you.

What you call the MIDI Channel approach is a variant of my approach #1, where MIDI channel selection serves as the switching mechanism.

most MIDI controllers these days don't really work quite like the hardware workstations

Can you elaborate a bit? I'm not very familiar with traditional MIDI techniques, but I have long suspected that exactly what you said is the case.

@getdunne

I should note, I started using softsynths five years ago, so I still feel brand new.  I stayed with hardware for years because it was reliable.  I was unsure of the hurdles and technical issues I may suffer with computers/software.  I also did not want to add yet another piece of equipment I had to remember to bring and setup (which I did forget last week and had to drive back home and get it).  I was looking to simplify my setup.  For several years I used an iPad (before switching to a computer) which had proven itself and greatly increased my sound palette.

Hardware is mainly based on MIDI Channels.  You assign your hardware (rackmount, drum machine, even effects for example) a MIDI channel for listening and anything sending on that MIDI Channel will control the device.  Workstations, at least the ones I've used, tend to provide a feature for programming patches using this concept of MIDI channels.  Korg called them Combis, Yamaha calls them Performances.  Roland might be Performances, too.  This option lets you combine multiple MIDI channels into what you can consider a patch.

For example, pressing BankA Patch1 can bring up a grand piano on MIDI Channel 1.  Pressing Patch 2 can bring up the Grand Piano on MIDI Channel 1 with a pad assigned to MIDI Channel 2.  You now have a piano/pad layer.  Patch 3 calls up a layer using sounds assigned to MIDI Channels 2, 3, 4.  Patch 4 a polysynth on MIDI Channel 5.  And so on.  The point is, with workstations, you can simply tap a button to call up your sound and even give it a nice name.  Even in this mode, it still has a concept of patch selection.  Each can have its own set of layers and splits.  I typically name them based on the song and sound (PNO, PAD, ORG, etc.).  My typical hardware setup I used the voice/patch mode on rackmounts and set it to a specific MIDI Channel, using Program Change to switch sounds from my workstation controller.

What I have found with MIDI Controllers is they completely miss this concept of programming patches.  They have templates, but it's not the same.  Some let you do layers and splits, others don't.  You sometimes have to scroll through templates as opposed to quickly selecting what you want.  Many don't allow you to name anything, other than remembering a number (01, 02, 03, etc.).  And most have a limited number of templates.  In my research, they seemed better suited for in the studio production, than live use.  At least for my needs.  I still use basic MIDI Controllers.  Just depends on whether it will meet the need.

I have carried over this concept of MIDI Channels to how I use my DAW when playing Live basically because that is how my workstation controllers work.  Though, these days, a lot of the layering/splits I now program directly on the DAW side of things.

I haven't kept up with the latest trends in controllers.  I know everyone is trying to come out with their own software to make DAW integration better and maybe some do a really great job of it.  The last time I researched this (six or seven years ago) I ended up realizing nothing met my needs, so bought the Yamaha MOXF6 (I believe you have the MODX) which so perfectly met all the qualifications I was looking for.  Two years ago I did look again at 61 key controllers, but didn't need much research before I settled on the NI S61.  The integration with Ableton Live and its mixer view is perfect for my needs.

I hope this was somewhat clear.  Ask if you have any questions.  I do enjoy discussing this, learning and seeing how others use the software and play live.

@ssquared

Thank you, that's totally clear and very helpful.

What I see happening with MIDI controllers is that they are gaining new features (like DAW integration), while many other features (like split/layer assignments and presets of all kinds) are migrating into the software they connect to.

Another trend, which isn't quite as obvious, is that the whole notion of MIDI channels is slowly migrating out of the controller and into software. Our PolyBox is a perfect example: notes can come in on any MIDI channel, and PolyBox splits them out to multiple monophonic layers with a different MIDI channel for each layer.

The biggest issue which remains (as I see it) is that most new MIDI controller devices use USB rather than the original 5-pin DIN MIDI connections. Most software (including Unify) aggregates all incoming MIDI into a single stream, so the ONLY way to separate the input from multiple controllers is to set each one to a distinct MIDI channel. This used to be easy to do on older keyboards, but for new ones, I find I have to pore through the user's manual to figure out how to do it. This really is too bad, because it means at least some users will give up and not bother.

[ BTW I don't have a MODX, but I certainly did drool over them at NAMM, and I live in hope that Santa will bring me one someday. 😎 😎 😎 😎 ]

@getdunne

Regarding the single stream, the DAWs I've used seem to know I am using different hardware and I can assign a track to ALL devices or select a specific input device.  Then each hardware can have up to 16 MIDI Channels.  This only works for USB devices, I believe.  Older 5-Pin devices can only be separated by MIDI Channels, not by individual device.  If necessary, I can do a little investigation into this.  It's kind of like the "Active MIDI Input" settings in Unify where it lists all the connected hardware and you can check/uncheck which ones are active.  I'm wondering if there is a way to have an option where you can select which device a layer listens for (All or a specific device selection).

Posted by: @getdunne

This used to be easy to do on older keyboards, but for new ones, I find I have to pore through the user's manual to figure out how to do it.

Do you mean older keyboards were a lot easier to change MIDI Channels?  Yes.  I agree. That's part of my struggle with pure controllers these days.  The ones I use have such limited programming capabilities.  One I have to press two buttons down at the same time, press a keyboard key representing change MIDI channel, press other keys representing numbers, then another key to signify I am done.

Oh, I'm sorry.  I thought I saw you write about the MODX on the live chat.

@getdunne

I think if you can nail seamless "patch" or "set" changes - where you can sustain the previous instance- and when you switch- the next note on's trigger the new patches- that would open it up as a MainStage replacement (and for PC and Mac users).  Simple start would be your N+1 idea.  You could call it Uni-Cycle lol.  Mainstage works under a "concert" for the entire collection of "Sets" and then individual "Patches (that contain multiple channel strips that could be instruments, Effects, playback tracks.

It would be great if a "song" or "Set" could have all the underlying paches load up to make for seamless switching between sounds in a set - then leave it up to the user to decide how deep they want to build their patch change arrangement depending on cpu power.

Posted by: @jasonschoepfer

It would be great if a "song" or "Set" could have all the underlying paches load up to make for seamless switching between sounds in a set - then leave it up to the user to decide how deep they want to build their patch change arrangement depending on cpu power.

That's how I currently work my live sets.  I have the entire set loaded in Ableton Live.  Everything is loaded and ready to go making for a seamless transition between patches (like you mentioned).  Seamless changes are super important to me.  To throw a twist into the N+1 idea (which I think is a totally valid and great concept), I work a little differently in that I may decide to use something different on the fly depending on where things are heading or how the team as a whole is playing.  I also do padding and background and decide right then and there what sound to use.

Our sets are small (five songs), and I haven't run into any issues so far (even on much larger/longer sets).  And I don't consider my computer all that powerful (Microsoft Surface Pro 3).  I typically have between 8 and 12 sounds per set, most of which contain several layers.

Clearly, some people prefer the set-list approach, and others prefer the "load everything at once and choose" approach. We hope to support both options, with flexibility to combine them.