I’ve been experimenting with the Vienna Symphonic Library, which includes some amazing orchestral samples with many types of articulation, played via the Synchron Player plugin.
I was wondering if anyone else is using it, and if they have found any good way of managing patches with it? It seems to be set up to switch sounds via note presses outside of the normal playing range,
which takes you through a hierarchy to set up your sound, so for example you might press C#1 to choose Long Notes, then G1 to choose Non-Marcato, then E7 to select Sforzato. Three note presses to set up your sound. This might be ok for a performance where you want to switch things manually, but not very helpful when you want Cantabile to be in control. I’ve looked into using bindings to trigger these notes sequentially, with limited success. You can use program changes instead of note messages, but you still have to send a series of program changes to navigate through the style/articulation choices. I could make a rack to encapsulate this sequential approach, but it’s pretty horrible.
I’ve tried the “Whole Bank” approach, but when switching patch like that I invariably get a big audible bang coming through as it switches. I’ve tried countering that with a timed binding that mutes the plugin for a period after patch changes, but it hasn’t been massively successful.
I’m at the point where I’m wondering if I’ll just have to have a different rack for each type of Vienna Symphopnic sound I use, and never actually change the sound within each plugin instance. But then there’s a memory usage concern.
I’m just wondering if anyone knows of a good way to use this plugin, such that Cantabile state changes can change sounds in this library, without big bangs/pops?
I guess @Neil_Durant has already tried the basics - he’s one of our resident gurus, so when he comes here to ask questions, things must be reeeeeally difficult…
What about a rack with an embedded Media Player?
You could trigger short MIDI sequences to send to the plugin, using the binding “On state load” => Play sequence from start.
Just an idea…
I don’t own Vienna Symphonic, so can only speculate. But I guess the key memory issue is not the plugin, but the samples. So if you manage to configure two separate racks with everything but the relevant articulation un-loaded, the memory impact of having two racks should be manageable.
Unless the keyboard switching actually unloads one sample set and loads another - but that would be prohibitive to use within a song, so I assume that’s not the case.
So having two different racks for the articulations might actually be the safest way to go…
I wondered about that too. It would probably work, but really annoying in terms of maintaining the rack, adding new states etc., as each new rack state would require a new MIDI sequence. This is why I thought using a set of bindings, each configured to fire a certain duration after the rack state change might be better.
Yes, I think you’re right, that may be the most solid solution. The memory aspect is a concern because the whole library is rather huge. But yes, constraining each rack to only use a certain sample set would probably work out ok.
Set state behavior to “Target”, then you can set different notes for every song/rack state - should be automating the switching nicely - @Neil_Durant, I assume you’ve already tried that?
Thanks @Toaster. Yes, that’s how I do it for all my other plugins/racks. The issue here though is that the selected articulations don’t appear to be “exported” to Cantabile as plugin state, as far as I can see, so I can’t stash it in rack state. Using the “Whole bank” approach does work, so does appear to include the selected articulations etc. - it’s just that when switching whole bank, you get this massive bang through the audio, which would not really be very good when playing live! So it looks like I’m stuck with the rather clunky approach of driving the built-in sound changing mechanism using a sequence of notes or program changes for each articulation change.
Thanks @Torsten. Yes, I tried that and it kind of worked. I think I need to put a bit more time into tweaking the time delays to make it reliable (as the plugin seems to need enough time to reconfigure itself between each one).
Depending on the sound, it could require up to 5 (I think) sequential notes/program changes. The sequence of notes is essentially describing a tree structure of possible articulations/variations.
I just reached the point of thinking to myself, “Do I really have to resort to doing it like this?!” and then decided to tap into the collective expertise of the Cantabile Forum
@Neil_Durant, if I understood correctly, the plugin has no way to remember the selected articulation (no preset system?), so you have to use “Whole bank” to store all its parameters in a Cantabile program? but then it will produce a big noise when you switch programs? Is this correct?
Correct. There is a preset system, but that seems to be more about which samples to load, rather than the various articulations/styles to apply. Those presets don’t hold any articulation state. And actually, I haven’t found a way to drive that preset system from Cantabile either, again except via “Whole bank”.
So it really is feeling like the articulations (which make a huge difference to the sound) are something that it only supports dynamically via its own articulation interface, rather than via any controllable preset system.
Interestingly, the mix of all the orchestral instruments is exported as plugin state to Cantabile, so I can control the balance of brass/violins/cellos etc. via rack state. Just not the articulations.
You probably already tried this: you can check the list of parameters “exported” by the plugin by clicking the “faders” icon in the top left corner of the plugin window. Then you can look for some “articulation related” parameter…though, from what you said, I guess there will be none.
Ugh - that doesn’t sound reliable enough for live use, for my taste. In that case, I’d really try to split this up into multiple racks and avoid all the fiddling - leave the switching entirely to Cantabile.
I try to avoid large sample libraries wherever I can - and for those cases where I can’t avoid it (typically piano / e-piano libraries), I follow a “load and forget” approach, leaving the sample library completely static after loading and maybe changing some processing afterwards (different EQ settings or FX parameters) for different rack states.
In my experience with articulation switching it was meant (made) to be used manually in real time. This means the VSTi is expecting the time between note presses in the articulation keys to be as long as it typically takes to press the key, release and go to the next one. So if you do use bindings to perform the note sequence out you should go with larger delays between notes, in effect fooling the VSTi into thinking it’s receiving manual note inputs. Say 1 sec or so for starters. So it may end up that in the cases where you use it your delay before being able to hit notes could be some number of seconds. If the articulation stays the same for the whole song than it should be OK for use on a normal song change time for a band. So the bindings solution Torsten posted with longer delays is probably the best bet as far as Cantabile goes. Just my 2 cents.
Dave
Yes, I tried that, and there’s really nothing that looks related to articulation or similar. I’ve also tried checking all the boxes in the plugin state behaviours list, but articulation appears to be not persisted/changed between states there - only with “Entire bank”.
I agree, I follow the same principles. Even if it appears to work fine at home, the fact it hinges on timing is unsettling. What if samples load slower one day, and the delay isn’t long enough?
So yes, I think just setting up some “static” racks sounds like the more robust option. I’m not planning on using most of the articulations anyway, so I think that should be fine.
Yes, you’re probably right, extra-long delays would probably make it robust, as it’d be no different to a user doing the articulation changes by hand. But then for some sounds that would mean 4-5 seconds just to get the right sound. Probably ok for general use, as I’m not likely to be changing these much during songs.
But still, the amount of “machinery” involved just to dial in a sound is unsettling to me for live use, so I think I’ll probably end up going with the thing @Torsten is suggesting - make a few solid, reliable racks for the main sounds I use from that library, and don’t fiddle with the articulations.
