It feels almost like an insult to ask for more, but here it goes
I\'d like to have more staccato/spicato samples...switching/alternating between 2 helps already, but with fast passages I find that it is still to consistent. I realise that probably Gstudio is the main culprit here because of it\'s inflexible structure for these kind of things, so here is my 2 cents on how it could be done;
-allow maestro tool to automatically switch between multiple midichannels; say I reserve midichannel 1 to 4 for staccato violins; each channel has 2 staccato samples which can be alternated between. Now when I play a line, maestrotools will alternate between 2 samples on 4 channels, meaning that effectively I can now switch between 8 staccatosamples! This would already help a great deal.
Another feature that would liven things up a bit is a randomise feature in Maestrotools; this would randomise the alternate so that it is not 1-2-1-2 but 1-1-2-1-2-2-1-1-1-2 etc. This way, if you play staccato lines with both the first and second violins, it would sound more random (and thus more real .
Anyway, like I said, my 02. cents!
Actually - depending on your sequencer - it can be done right now. In Logic Audio you can use a transformer in the environment together with the ALT88 instruments by randomly transposing notes up 2 octaves or down one octave.
However I do not think that randomised bowings would sound very natural. They could even sound funny. The difference between up and down bows in GOS is very audible and - with a few exceptions - string players do alternate between up and down. An exception i.e. for instance is the following scenario: you want to have several strong marcatos in a row, use consecutive down bows.
Instead of randomisation I think there should be an other feature for the ALT instruments. After a string part has been silent for several measures and restarts it should be default restart with a down bow.
For the moment I work around that by inserting \"phony\" notes out of the instrument range. Mr. Garritan however warns against this, as it could trigger the wrong samples.
ALT88 are an alternative, but score wise, I find them to disturbing.
Jeannot, you dont have to insert \'fake notes\' to make sure that a line starts with an up or down bow. As you probably know inserting an A0 note activates automatic alternate bowing. Bb0 and B0 can be used to select either up or down bowing. So when you want a line to start with a specific bowing, just insert a B0 before the first note in your phrase, then an A0 before the 2nd one.
Simon, thank you. I overlooked that.
I also think it is not expressed very clearly: \"B0 sets alternate sample mode on.\" So, if my last note - 5 measures early - was a down bow, B0 should give me an up bow, which I would have anyway?
A#0:\"Default mode, no alternate samples.\" These should only be down bows, or are they up bows, if the last note before A#0 was a down bow?
I guess I\'ll use my ears.
And here I thought I could get away with a simple, fairly non-committal answer. . . OK, I’ll go into my philosophy on this subject in a little more detail.
Randomization: inappropriate between bow stroke types. In other words, up and down bow strokes should never be randomized. That’s not how it works in the real world. Up and down bow strokes should be chosen deliberately, dictated by the requirements of the passage. Where variability (which is not necessarily synonymous with randomization) would be a boon is within bow stroke types, so that if you chose to play a downbow on a particular note, the actual sample played (while always being a downbow sample) would not always be the same downbow sample. Randomization within bow stroke types could be valuable IF the pool of samples was large enough. Just so your imagination does not go racing off right now, I can tell you that we do not have enough recorded samples to build a monster short bow instrument like this. But there are lesser alternatives, alternatives I’m not willing to go into at the moment (one of them quite easy to implement).
But do not overlook the tools already at your disposal. Here’s one approach: Once a passage has been recorded into your sequencer, a single articulation need not be applied to the entire passage. Multiple articulations can be loaded into GS and chosen note-by-note if desired. This can be accomplished by setting the sequencer track to pass all MIDI channel data. Then, if you have a note that needs to be marcato (for example), pick that note and change the note’s MIDI channel designation to match the MIDI channel containing the marcato patch so that that note plays a marcato sample, and so forth. It’s not fast but it does give you a means to choose precisely the articulation needed for each note of a passage. It also gives you a way to handle a passage of repeated notes that, because of repetition, reveals the same samples being used. This approach, combined with layering articulations and proper switching between up and down bow strokes in MaestroTools, can give a great deal of variablity to a passage and camouflage, or avoid, “repeated sample” problems.
Also, the more knowledge you have about how string players would actually play a passage the better chance you have to successfully mimic the sound of that passage. If you don’t already have extensive knowledge of how a real player applies bow strokes it would help a great deal to learn. “Instant gratification” has its limits, especially in Midiland.
P.S. Downbows in ALT instruments are mapped to coincide with the range of the standard instruments (exception: “88” instruments, which are an octave lower) with upbows mapped 3 octaves above that. And, just to reiterate, in Maestrotools:
Downbows are the default.
A0 = alternating bow mode
B0 = forced upbows
A#0 = forced downbows (return to the default)
Tom, one nice alternating bow feature would be to add an alternating triplet feature, with the pattern of D-U-U-D-U-U-D-U-U.
I find myself writing string triplets a lot in my current piece. I have manually gone in and done the bowings as described, but it would be nice if Maestro Tools could automate this. The triplets sound much better when the downbeats are accented (due to the down bowing).
Forgive me if this isn\'t how string players actually play.
I agree that up and down samples shouldn\'t be used randomly; i was talking more about having more than 2 staccato samples in general which could be switched between. In live played strings the difference between up and down strokes is very important; with samples however,with a repeating line that alternates between 2 strokes only, I doubt whether you\'ll know the difference between up or down after listening to it for 20 seconds
The realism with samples, I feel, comes from the fact that it is non-repetetive and therefore somewhat random; something you\'re not going to achieve with just two samples (however meticulously they have been sampled).
So my humble suggestion would be to give people for instance 8 up and 8 downstrokes, and randomly switch between the 8 ups for upstrokes etc.
Joris de Man
ps: I played the violin for six years, so I do feel I have some grasp of what a live player is supposed to sound like
[This message has been edited by Joris de Man (edited 11-20-2001).]
Acually I think the repetativenes isn\'t only from samples, but the sequencing. Straight not samples will never sound like a real section IMO. Even with multiple samples,little nuances of each INDIVIDUAL string players rhythm makes for what I hear as a majr difference between sampled sections and real players...with short bows. Not to mention pitch bendsportamento from note to note in certain passages.
I\'ve foudn that for fast lines using Tremelo patches layered underneath different short bows helps alot.
fr more variations of the \"same\" short bows, you \"can\" assign the same instrument to two MIDI channels, coarse tune one up or down a semitone and transpose the MIDI track appopirately, then you have 2 versions os down bows and 2 versionsof up bows. You have to alternate this \"manually\" in a sequencer, but it shouldn\'t sound too bad.
This file is called Titans basic or whatever I\'ve named it... Its part of the Return of the Titan\'s score. I always dug how big the strings sounded.
The piece is all GOS, no Violas yet, and I haven\'t touched it up as much as i want. It is pretty thick tho. This alone took more than 100 note poly. Lots of layers. Nearly all the violin short bows and all the cello short bows, as well as Tremelos
PS, a triplet alternater option would be AWESOME! Never thought of that, but it would work in the beginning of this piece
Really...I am an Idiot
[This message has been edited by KingIdiot (edited 11-20-2001).]