I just finished a pleasant Workshop for which I was invited by the University of Arts Graz largly occupied with the way I handle VSL-Samples:
-
VSL in Workshop at the University of Arts Graz
-
To give an Idea what we talked about here is a short (german) description of my concept of a universal Expressionmap for all VSL Winds & Strings, which makes always only all "relevant" variants available while working in the sequencer on the concrete Mididata, and allow widely to copy Midievents with all important information about the chosen articulation between parts of very different instruments even if they provide very different sets of of articulationsvariants. (I always thought this would be interesting for the VSL developpers to improve the usability of their constantly and impressivly growing variety of available instrments and patches)
I am curious what you guyss here think about.
http://klassik-resampled.de/S-Fahl-12-2016-Orchesterspielweisen-vom-Sequezer-gesteuert.pdf
-
Hi Steffen, Alas my German reading skill, or lack of any such skill to be honest, is not up to the task of reading your document. Nevertheless, I'm very interested in this subject as I've been developing something similar, I think, in the context of the computer-assisted composition and performance system I've been developing. It features a query system that allows different instruments to interpret the same musical passage in their own idiom, using the appropriate articulations. It's based in a common naming convention that encodes the salient features of each articulation that can be searched and scored for matches of those features to get a "closest fit". If you have time, would you mind giving a summary of your approach in English? I have a feeling it will spawn an interesting discussion. Thanks, Kenneth
-
Hi Kenneth;
Thank you for your Interest. Here I tried just to translate the linked pdf in english. For me it is just a working System how I organise my application of articulations keeping all reasonable Variants available.
http://klassik-resampled.de/tutorial-Orchestra-en.pdf
This is just my System which make things for me very efficient and helps me to avoid a lot of very tedious programming-procedures which would incresae with each new Library or sampleset provided by VSL
I am curious what you think about.....
-
Thank you so much for this! I look forward to reading it and benefitting from your wisdom and expertise. I'm always grateful to discover individuals making strides to improve workflow when dealing with the myriad of technical challenges virtual orchestration and sequencing present. All the best, Dave
-
Hi Steffan,
Sorry for the delay in getting back to this. I wanted to give out a useful chunk of design thinking around this that could be useful to others. It took this long. I'm attaching a pdf of the piece I wrote with some thoughts on your template design which I think has some useful approaches. I'm going to share some thoughts here on the bigger perspective of what we might be trying to do in this curious form of virtual composition, orchestration and performance. Anyone who wants to get more into the technical design details can have a look at the the pdf.
There are are a lot of overlaps in our approaches since the overall objective is the same: how to make the best use of the extensive VSL libraries. The difference in my case is that I want to leverage the power of computation to speed the process even more by automating some aspects of the articulation process which entails encoding some of the decision making that goes into making those choices by hand.
I completely agree with your emphasis on providing ways of making "musical decisions". It's got reflect the language of music as a prerequisite if it's going to function at anything resembling a musician's level of understanding of the medium. Music is a sophisticated language that operates in multiple dimensions all at the same time. We need tools that map as much as possible to that multiplicity.
The part of my design process that maps most closely to yours is the work I did over the last 18 months providing an answer to a basic question in two parts. It goes like this: In the case of the solo violin for instance, "Out of the almost 200 unique articulations to choose from in this amazingly rich library, why would I choose x, or y? and how do I maximize the variation in articulation in a musically satisfying way to breath the maximum amount of "life" into the resulting performances using this library?" I know that one answer that we hear a lot on this forum is that it's what sounds good and that is, of course, the final arbiter of any choice made. But I also think that there's always a lot of craft and embodied knowledge behind how we can even "know" what's good. So my project is a lot about trying to look deeply inside and drawing out intuitions and hidden patterns that seem to support my own sense of what "sounds good". That's in itself a complicated thing because there are so many facets to that understanding: biology, evolution, metaphor, cognition, cultural convention, style, taste, exposure to the right exemplars to use in comparison, and on and on. So I cherry pick from the literature in all those fields for bits of theory and insight that make sense in relation to what I seem to know from the inside. Then try to build a tool that produces a musical output. In relation to the VSL libraries that's a performance problem. Focus it there and the answers to question — why this articulation? and how to make a rich set of choices? — get reigned in enough to be able to think of some answers that could be tried out. It's still devilishly complex: performance is itself another of those multi-dimensional balls of complexity that good music seems to thrive in. Interpretation of the temporal patterns of the music, its rhythm in relation to a metrical structure that guides the formation of the sounding materials - figures, phrases, periods, sentences, sections, movements, works, to cite one traditional example. Or a mapping of the evolving spectral features of a single Trombone pedal tone to look at a more contemporary example. Either way there's pattern, form and conventions for mapping from features in those patterns and forms to an expressive performance. I call it the information that's encoded in a melody or phrase for instance. With a little thought, its' easy to identify ten or more salient features that a melody is likely to have, each of which could be mapped to a performance feature. Even with those two dimensions at work—meter/rhythm and the features of a melody, there's ample complexity to work with. How to reconcile the disagreements these two will have about how to treat the music. Which demand to best answer at any given moment? How to evaluate where the priority is greatest? That's what's keeping me busy listening, composing, listening to the composition, going back and reworking it with another approach. Looking for those moments where something unexpected emerges and demonstrates yet again that music is an endless wellspring of creative potential and we've only scratched the surface of a couple of corners of the whole design space it's made out of.
Kenneth.
TimbreU0020andU0020HierarchyU0020-U0020MeaningfulU0020SelectionU0020ofU0020ResourcesU0020inU0020aU0020LargeU0020Database.pdf.zip-1696254291032-1bxb0.zip
-
To come to the point shortly:
"It’s in the (musical reasonable) definition of the (hierachical) sub-types that the most power lies." 😉
It is true there definitly seem to be some important parallels in our approaches due to the fact, that we both want to make use of the full power provided by VSL as musical reasonable as most efficiently.
I still hope the VSL-Devellopers might keep in mind that there is still a great potential in smart and musical reasonable organizing the sampleapplication with pretty few things to do for them.
I think they are already working on a new VI-Version, and I hope they manage to incorpoate improvements in this direction.
Thank you for you detailed discussion
Steffen