It’s also a question of where you need smooth sound transitions.
TBH I have only ever needed it in songs between parts, not across songs.
But in that I include albums like Dark Side of the Moon where in the Floyd band I was in, we played side one first and then side two from start to finish with no interruption, but in that case, whilst we split side one up to give the guitar player a chance to get ready for slide on “Great Gig”, I treated side two as one continuous song with parts in it. Cantabile works a treat in that context.
So it really depends on what you want to do and how you set things up, and TBH even hardware units with SST usually have limitations that you need to plan for,
Thanks for the suggestions of workarounds @Sausagefingers and others in the thread. I will try that. Often in a set I may only have two songs where I need a seamless transition so hopefully there’s a way to load a song into a state so I don’t have to “rebuild” the song in a state.
@timc you’re exactly right, and I’ve thought the same myself. Cantabile should be able to handle this, ESPECIALLY since it’s a powerful and flexible app.
For example, if seamless transitions work between song states, then why couldn’t I choose an option for an individual song, to seamlessly transition to the next song? It would be a flag, and heck, they could add a warning about watching out for high CPU usage, etc. if necessary.
I’d much rather have the option and choice available than not have the feature available because some people, in some rare scenarios (too many hungry VSTs, too slow a CPU) might have problems. If it works using song states, then the app and your CPU can definitely handle it.
OK, so this workaround is doable, but you can’t “link” a song into a song state as far as I can tell. So I have to copy all the objects from the second song and pasted them in the first song, then disable routing in the second state to instruments that should only be there in the first song, and vice versa.
That’s a lot of work for every scenario I’d run into. I thought I could somehow copy the second song into a song state and go from there.
It’s fine if you use all the same instruments for both songs, but that’s almost never the case for me.
Also, this second song in State 2 means that the first song always has it from now on, but that’s almost never going to be needed for another set, with different songs. So I have to:
Save Song 1 as a separate (combined) song file
Remove the first song from the set
Edit the song to add the second song state
Copy all the instruments, effects, etc. from Song 2
Go paste them into Song 1, State 2
Go to Song 1, State 1, disable routing to song instruments from Song 2.
Go into Song 1, State 2, disable routing to song instruments from Song 1.
Remove the standalone Song 2.
So I guess I’m back to hoping a feature gets added where I can simply check a box in Song 1 (Enable sustained sound transition) so that Song 2 can start even while notes are fading out or being held for a bit from Song 1.
It works when building the big complex workaround as described above, with no hitches, no glitches, so I guess I’m left wondering why this couldn’t be implemented between two songs.
The only surefire way to realize this would be to keep the current song loaded and active when loading the next one. Then, on the next change, swap out the oldest song for the next. I wouldn’t want song loading or un-loading going on during a song, only on changes (avoid CPU and storage overloads).
But this would mean having two songs active at the same time, creating double CPU load. Definitely something I would only want to be optional - given the complexity of my typical songs, I would probably keep this set to OFF.
Would be interesting if @brad considers this feasible…
There is no SST/SSS between songs.
You can hold notes between Song States, as others have described.
However, here’s my workaround:
Each song is a song, as i tend to use song states for parts or different sounds within a song.
FX like Reverb and Delays are in the background rack. All instruments are dry and are routed to the background rack for reverbs and delays.
When i switch between songs, the sound of the instrument stops but the reverb tail or delays keep going, because the background rack doesn’t change between songs.
This way you stop playing when switching songs but there is no sudden ending in the sounds.
The Setlist is always preloaded.
@Steffen Interesting. So I always preload the entire set with no issues. But you’re saying that you have a standard reverb in the background rack for all songs and everything gets routed there? I have a standard compressor and mastering VST (God Particle) in every song, however the settings have to be often customized per song depending on the synths I’m using, so one setting for all songs might be hard to do.
I guess the other thing is that usually I have at least one pad in play, so if I cut it off and just reverb tailed it still would sound very noticeable.
@Torsten My CPU and RAM are quite strong, even on a laptop. I often have anywhere from 4-10 instrument VSTs and a 2-6 FX VSTs in play for one song. So having two songs loaded and active would be no problem at all, especially since the second song’s VSTs shouldn’t be doing much while waiting to receive notes once I move to that song.
However I don’t know what you mean by having 2 songs active at the same time. I pre-load all the songs when I load the set, but what do you mean by swapping out Song 1 for Song 2 as opposed to switching to Song 2?
So in your scenario, say Song 1 is ending and you have the sustain pedal held. You switch to Song 2 and the sounds are still being held as you hold the sustain pedal, but any new notes are using Song 2. You let go of the pedal, Song 1’s sounds trail off while you’re already playing Song 2. Is that functionally how it works?
Pretty much - not sure I held the sustain beyond the first beat of the new song…but meant that I could switch songs while the last chord of Song 1 was being held and all the patches I needed for the first beat of Song 2 were there ready for me to play on the beat.
Been I while since I’ve needed to do it TBH…(it was a jukebox musical with a segue between two numbers) so my memory’s a little hazy!
The solution I described in my earlier post solves this problem and requires none of the steps you listed. To make a new song set, I just load my master song, right-click on each rack and select “Replace Rack” to replace it with a different song’s rack, and I’m done. No reprogramming. It takes me about 30 seconds to create a new set from my collection of song-racks, and each song is sustainable with pedal as I start to play the next.
OK, I have a better idea of what you’re doing there. I guess you need to separately note things like song tempo, time signature etc. as that won’t be stored in a rack, only in a song. And for anyone using other song information that is typically saved in a song, such as global transposition, notes, etc. those can’t be stored in a rack either.
The other thing that’s missing when you do things like this is that it’s harder to see at a glance notes/comments (can’t remember the name of the column) for each instrument/FX layer in a rack for quick adjustment or memory.
For example, I may have a pad that uses CC17 for filter cutoff, two EPs whose volume is controlled by CC20, an FX layer (stutter/glitch) controlled by CC21. I put those in the notes column so if in the middle of playing I forget what knob on my controller I’ve assigned to the EP volume I can glance at the computer screen an the song has all that info. And I can quickly reach over and do things on the computer if I need to.
If this is all in a rack I’d need to expand each rack… maybe they all stay expanded when there are multiple racks? Not sure.
So I can see how this might work in scenarios where songs have very consistent VSTs and settings, and you have a separate place you’re referencing tempo and time signature, setting that up each time but it seems that’d add a lot to my workflow and limit more visibility.
But if I have to do this for only 2 songs in a 5-song set, maybe that’s a decent workaround/option.
That’s true, if tempo/meter isn’t served by an external clock, one must set them for each song state. I’ve found that to be very quick (adds another few seconds per song prep time).
I think you can set each song state to auto-open a particular rack when the state loads (e.g., using a binding). Like you, my VST controls are not consistent so I need notes, but I personally prefer forScore for my sheet music, in which I have notes about which buttons to press when; so I only refer to my rack comments when prepping. I’m also loving my Presonus Atom SQ controller because it lets me color-code the buttons per song, which really helps me remember which is which. (See my other thread on the merits of that controller.)
I guess I should also mention that there’s actually a nifty way to “store” tempo/meter/transpose settings in racks that I sometimes use for more complex songs: I have my master song programmed to send a special CC message to the “current” song-rack whenever it becomes the “active” song. This allows all my song-racks to optionally listen for that CC message and set the global tempo/meter/transpose to prescribed, song-specific values. The tempo-setting bindings live in the rack, so it’s a way to “store” such things in racks.
@Lowell Background rack is a very powerful feature in Cantabile.
To keep things simple, i control a few general things here. The bindings for changing Songs and Song States are stored here so they are always the same buttons. Also Tap-Tempo and Panic!
There is a very good post from @dave_dore on how to use the background rack for send effects. Background Rack
I have three reverbs (Plate for Brass, Mid Hall as a general Rvrb, large Hall for FX) and a delay in my background rack.
You could even have instruments loaded in the background rack. I have never done it that way, but it would have the advantage that the sounds are not cut off when changing sounds.
My point was that for true SST, it would be necessary to have two songs active for processing, with held notes (and sustain pedals) from Song 1 still being processed, while Song 2 is already ready and active for any new notes. This is a capability that Cantabile currently doesn’t have - it would require @brad to re-work the whole aspect of loading and un-loading songs.
An interesting aspect of this: this could be problematic with shared racks: Song 1 would have Rack A active in rack state 1, while Song 2 would have Rack A active with rack state 2 - this would mean that Cantabile would have to have that rack loaded TWICE - conflicts with the concept of shared racks and pre-loading.
So this is definitely not something that @brad can build over a couple of afternoons, so I guess we’ll have to live with workarounds for a while…
Note: even if a setlist is pre-loaded, this doesn’t mean that all plugins are being PROCESSED - switching songs still means disconnecting the current processing chain and connecting a new one, then starting the actual processing with the new setup. There is always a period of re-arranging, switching instruments and connections off and others on.
Also, about the assumption that “VSTs shouldn’t be doing much while waiting to receive notes” - I strongly disagree. A typical song in my setup has a baseload of around 30%. Duplicating that would put significant strain on a system. Don’t underestimate the baseload of plugins just being active for processing - not only effect plugins, but also instrument plugins can have significant baseload.
So on my Church team we might want a background pad in D behind 3 songs, I could easily setup an M1 to play “Generic Moody Background in D” and keep that in all my Song Parts, with a “Part” assigned to each tune we’re playing.
The key thing to remember in any “How do I…?” Type questions is to focus on the END RESULT, not on the step. The end result you want is a sustained sound across several “Real Songs.” The trouble I see you running into is that Cantabile Also has a thing called a “Song” but it’s NOT a “Real Song” . . It’s just a file that keeps certain information organized. You can sustain between Real Songs all you want, by putting them together into a Cantabile Song
So you have one “Cantabile Song” for your set of “Real Songs.”—Remember a “Cantabile Song” is just a file, with one or more “Cantabile Part” for each “Real Song.”
I might do this if I were doing “So Fine” into “Living Thing” by ELO; The B side of “Abbey Road”, or some of the more modern musical theatre plays where the music never stops.
Give it a whirl and let us know how it goes, but the idea is straightforward.
In my original prog band, we often have a sound effect or other sound that plays during a song change. What i have done is put a Media Player into a linked rack, shared by the two songs. Starting it at the end of Song 1, it will continue to play while i change to Song 2. So if you can make an audio file of the sound you want to sustain over the song change, that might do it for you.
OK, but regarding the VST load… suggestions here are often recommending putting all the VSTs from both songs into one big song so there’s no audible transition. So in this case you already have say, twice the VSTs active and loaded, and if you’re running into CPU load problems with that, then this is not a viable workaround.
Regarding issues with people using shared racks and wanting SST; I’d suggest that in designing the feature, if you want SST between two songs and the same shared rack is used between the two songs a warning could pop recommending you convert the shared rack from one of the songs to an embedded rack and then you can proceed. That’s far less work than building a new combined song with 2 parts.
As well, a warning could pop telling you that you may experience CPU issues if you use SST between the two songs, then it’s up to me to decide whether I want to move forward with SST in that case. I can move ahead; I’ve been forewarned, and I can test very easily, slim down any VSTs if I need to, and tell the warning to not show again because I will remember that for next time.
I’ve tried the option of copying the layers from one song into another and using song parts as suggested here. It’s clumsy, and I overwrote Song 1 by accident (I have backups so no big deal), but it works and at least in the case that I tried, absolutely no problems with CPU load. But it’s such a manual process and only applicable for the one performance as it’s very unlikely that we’ll have that combination of Song 1 and Song 2 again.