Pete Brown
Regular Member
Hey there, VI-C,
I am looking to develop a little more efficient way of working with a DAW and VIs that are used by many on this forum for orchestral writing. Although I have a decent background as a pianist, I am comparably new to the world of DAW/VIs. As far as DAWs are concerned, I've mucked around with a few, but ultimately settled on using Cubase in Windows (I'm on version 11 Pro and although I am not even close to appreciating all of it's features, this is kind of settled for me).
What I've Tried
It took me quite some time to learn that the simplistic question of articulation sets vs key switching vs one track per articulation is more complicated than it seems. As a pianist, it is fairly easy for me to "play" notes in and with a piano sample, there is zero problem once one gets beyond latency. That is easy. A staccato/spiccato line on a violin is also pretty easy to "play" in, though here I set a track delay to account for the time that the bow touches the string (the sample start) and the start of the note per se (the start of the sound). If one is aligning to the grid, the start of the note would be what I would align to the grid. I get that and that is fairly easy too. Long notes add a new layer - that of dynamic control and although my left hand was not trained to wiggle a wheel, it is getting better and honestly I often end up tweaking my cc1 values after the fact. Legato passages add another layer because there are often two different delay offset values - one for the first note of the passage, and one for the other connected legato notes. I find that I set a negative track delay based on all of the subsequent notes, and then pull the first note start time a little (I use a mouse - there is probably a better way). I think I am pretty efficient with these sorts of passages that are either all shorts or all longs and have been using a single track per articulation. That makes sense to me.
Where things get more complicated is when a passage contains a combination of shorts and longs. What I had been doing (and still do sometimes) is open up an instrument, play the line in, duplicate the track, and convert one instrument track to the short and one to the long (or whatever the different articulations are). Then I delete the midi data that doesn't belong (e.g. delete the short notes from the longs track and vice versa). At that point, I was savvy enough to know that this was not efficient, but not savvy enough to know a better way of doing things.
Then, I took the key switching approach. Key switching with the left hand for simple things (short long) seemed reasonable enough when playing things "live." But if I snapped things to the grid or even rough quantized to match other performances with other instruments (I often do this with orchestral, but not piano, writing), I run into the problem of different delay times per articulation on a single track, and this problem varied by developer and library. Inconsistencies in key switch maps and delay times was maddening! Preaching to the choir, I know. This made VSL more attractive and I found the Synchron Player to be very useful (e.g. SSP). Wonky out of tune notes and inconsistent timing made other developer's libraries more of a challenge, even if I preferred their sound for a particular project. But I still run into the problem of multiple delay times with the key switching method.
What I am Considering
In addition to my interest in CSS (and maybe CSSS) this Black Friday, this problem brought me to the Variable Delay Compensator thread (Holy Moly! @kpmuzik seems to have hit upon something and wrote a very helpful and detailed manual - wow!). But this seems pretty complicated and I am not sure that I completely understand this. I know that purchasing CSS with all of the fabled consistencies of the library will not solve my problem. I love the sound and the consistencies between sections is just a (huge) added bonus (because even copying one section to another can be very labor intensive with other libraries). But I will still have this issue of multiple delays and different articulations. I suspect that there is a way to send certain notes from a single instrument midi to different tracks so that I can use different delays. And maybe this is what VDC is all about. It seems like Cubase might have a way to send certain notes from a "master" track that I play in to other tracks (I've only ever used instrument tracks - starting a new instrument and a new Kontakt instance for each articulation - and maybe this is what the midi tracks are about?). Maybe there is a way to use Kontakt's multiple outputs for different tracks for this (I actually don't know how to do this)? Newbie Question indeed
Finally, from what I have read, there seems to be no holy grail way to solve this problem, particularly when using multiple libraries from multiple developers. Things like TouchOSC and Streamdeck are attractive, but seem like a better mousetrap version of a key switch. I honestly have not used articulation sets because they also seem like a better mousetrap version of a key switch and none of these solve the multiple delay problem. Not that better mousetraps are not better, but they definitely seem to add a layer of complexity (though this could just be reflective of my anxiety investing time into a new "system" or workflow).
What I am Missing??
If you were going to start anew with an approach (in Cubase!), and knowing that you will sometimes have to rely on the grid, how would you address this problem? I saw @jesussaddle's recent post but it seems like the suggested solutions are switching DAWs and I am not ready to do that, especially since I know that a ton of people somehow do this work with Cubase. If nudging notes is the best way to get final result, I will accept that....if one track per articulation is the best route, I will accept that too....if there was advice for how to get complex lines with shorts and longs into different tracks and that is a smart approach, I'd accept that too...Key switching, articulation sets - I am open to it all. I would prefer not to have to surgically construct each and every line for each instrument (a little massaging and edits on cc is totally fine), but if this is reality, I'd accept that over searching for a holy grail that may never materialize. I'd rather just make music. I have used this "surgical" approach and can get a decent end product, but boy it is kind of painful. Maybe that is reality? If so, so be it. But I know enough to know that I don't know enough
I *really* appreciate any input you might have to offer!
Pete
I am looking to develop a little more efficient way of working with a DAW and VIs that are used by many on this forum for orchestral writing. Although I have a decent background as a pianist, I am comparably new to the world of DAW/VIs. As far as DAWs are concerned, I've mucked around with a few, but ultimately settled on using Cubase in Windows (I'm on version 11 Pro and although I am not even close to appreciating all of it's features, this is kind of settled for me).
What I've Tried
It took me quite some time to learn that the simplistic question of articulation sets vs key switching vs one track per articulation is more complicated than it seems. As a pianist, it is fairly easy for me to "play" notes in and with a piano sample, there is zero problem once one gets beyond latency. That is easy. A staccato/spiccato line on a violin is also pretty easy to "play" in, though here I set a track delay to account for the time that the bow touches the string (the sample start) and the start of the note per se (the start of the sound). If one is aligning to the grid, the start of the note would be what I would align to the grid. I get that and that is fairly easy too. Long notes add a new layer - that of dynamic control and although my left hand was not trained to wiggle a wheel, it is getting better and honestly I often end up tweaking my cc1 values after the fact. Legato passages add another layer because there are often two different delay offset values - one for the first note of the passage, and one for the other connected legato notes. I find that I set a negative track delay based on all of the subsequent notes, and then pull the first note start time a little (I use a mouse - there is probably a better way). I think I am pretty efficient with these sorts of passages that are either all shorts or all longs and have been using a single track per articulation. That makes sense to me.
Where things get more complicated is when a passage contains a combination of shorts and longs. What I had been doing (and still do sometimes) is open up an instrument, play the line in, duplicate the track, and convert one instrument track to the short and one to the long (or whatever the different articulations are). Then I delete the midi data that doesn't belong (e.g. delete the short notes from the longs track and vice versa). At that point, I was savvy enough to know that this was not efficient, but not savvy enough to know a better way of doing things.
Then, I took the key switching approach. Key switching with the left hand for simple things (short long) seemed reasonable enough when playing things "live." But if I snapped things to the grid or even rough quantized to match other performances with other instruments (I often do this with orchestral, but not piano, writing), I run into the problem of different delay times per articulation on a single track, and this problem varied by developer and library. Inconsistencies in key switch maps and delay times was maddening! Preaching to the choir, I know. This made VSL more attractive and I found the Synchron Player to be very useful (e.g. SSP). Wonky out of tune notes and inconsistent timing made other developer's libraries more of a challenge, even if I preferred their sound for a particular project. But I still run into the problem of multiple delay times with the key switching method.
What I am Considering
In addition to my interest in CSS (and maybe CSSS) this Black Friday, this problem brought me to the Variable Delay Compensator thread (Holy Moly! @kpmuzik seems to have hit upon something and wrote a very helpful and detailed manual - wow!). But this seems pretty complicated and I am not sure that I completely understand this. I know that purchasing CSS with all of the fabled consistencies of the library will not solve my problem. I love the sound and the consistencies between sections is just a (huge) added bonus (because even copying one section to another can be very labor intensive with other libraries). But I will still have this issue of multiple delays and different articulations. I suspect that there is a way to send certain notes from a single instrument midi to different tracks so that I can use different delays. And maybe this is what VDC is all about. It seems like Cubase might have a way to send certain notes from a "master" track that I play in to other tracks (I've only ever used instrument tracks - starting a new instrument and a new Kontakt instance for each articulation - and maybe this is what the midi tracks are about?). Maybe there is a way to use Kontakt's multiple outputs for different tracks for this (I actually don't know how to do this)? Newbie Question indeed
Finally, from what I have read, there seems to be no holy grail way to solve this problem, particularly when using multiple libraries from multiple developers. Things like TouchOSC and Streamdeck are attractive, but seem like a better mousetrap version of a key switch. I honestly have not used articulation sets because they also seem like a better mousetrap version of a key switch and none of these solve the multiple delay problem. Not that better mousetraps are not better, but they definitely seem to add a layer of complexity (though this could just be reflective of my anxiety investing time into a new "system" or workflow).
What I am Missing??
If you were going to start anew with an approach (in Cubase!), and knowing that you will sometimes have to rely on the grid, how would you address this problem? I saw @jesussaddle's recent post but it seems like the suggested solutions are switching DAWs and I am not ready to do that, especially since I know that a ton of people somehow do this work with Cubase. If nudging notes is the best way to get final result, I will accept that....if one track per articulation is the best route, I will accept that too....if there was advice for how to get complex lines with shorts and longs into different tracks and that is a smart approach, I'd accept that too...Key switching, articulation sets - I am open to it all. I would prefer not to have to surgically construct each and every line for each instrument (a little massaging and edits on cc is totally fine), but if this is reality, I'd accept that over searching for a holy grail that may never materialize. I'd rather just make music. I have used this "surgical" approach and can get a decent end product, but boy it is kind of painful. Maybe that is reality? If so, so be it. But I know enough to know that I don't know enough
I *really* appreciate any input you might have to offer!
Pete
