Not sure if ya'll think topics related to realistic emulation of instruments belong here or in the DP forum. Part orchestration, part technology. Since we moved the "Adapting Orchestration Styles to VI" thread here, I'll assume this is the place for such discussions even though it's not strictly in the forum's statement of purpose.
My question touches on both that thread and the one which mentions the Samuel Adler set. There's a video demonstration of a guy playing a simple violin scale which I looked at recently since trying to get more into the "realism" thing. When I listen to just this simple scale i feel at a loss as to how to even approach getting a sound like that out of VIs. He isn't using much variety in terms of articulation (legato, marcato, etc). But every note seems somehow to come from a different place. Is it slight tuning variations (intonation)? What it seems like is as if there's different amount of harmonics on each note. I don't know if I'm explaining it right, but the effect is of something "wilder" than I hear with any VI. It's almost akin to the overtone wildness of some old upright piano when compared to a VI violin. I would love to figure out how to come as close as possible to this simple scale because it seems like it would be a good way to fine-tune sensitivity to detail. (I'm pretty sure posting the video would be illegal since it's copyrighted material, so I'll have to hope maybe some of you already have it and can understand what I mean.) Do you think it's possible to get close to that vibe? And if so, how close and using what trickery?
Using VIs to emulate real instruments (belongs here?)
Moderators: Frodo, FMiguelez, MIDI Life Crisis
Forum rules
Discussions about composing, arranging, orchestration, songwriting, theory and the art of creating music in all forms from orchestral film scores to pop/rock.
Discussions about composing, arranging, orchestration, songwriting, theory and the art of creating music in all forms from orchestral film scores to pop/rock.
Hey Zaster:
I think your topic is eventually what most of us are aiming for, so to me it seems fitting. It's certainly not OT, and on some level it's not even *that* closely connected to DP issues. It's all related on some level, so it's going to be impossible not to talk about how-to's without mentioning DPs features or the features of various instrument plugins at some point.
With that said, I would first recommend looking at the approach a little differently. Instead of trying so much to bring virtual tracks to the level of real instruments, it may be an easier process to at least *start* by working on eliminating undesirable sonic elements in the virtual track with mechanical or technical qualities that take away from the listeners ability to remain focused on the music itself. Otherwise, you could spend the rest of your life and half another lifetime getting a violin patch to sound exactly like the real deal. I do believe, however, that you can get close.
Some VIs for strings allow for more vibrato control than others. The problem is, vibrato is not as steady from one note to the next or from one player to the next as most VIs would have you believe. Some vibrato control is better than none at all, and that's where most VIs drop the subject. Even VSL is having trouble at the moment with the larger string sections smoothly x-fading from non-vibrato to full vibrato-- except on the solo strings, which really do this quite well.
But when listening to your violin scale, each note has a different vibrato (or a different non-vibrato, as the case may be) and a different set of "imperfections" which could be analyzed and at least emulated (where imitation is beyond feasibility). Understanding just how the vibrato wavers irregularly might involved some pitch/wave analysis, and each note would have its own set of parameters. This digs deeper into the techno uber-geek underbelly than probably necessary, imho, where the pursuit of technical reality overshadows the pursuit of musicality.
Upbows and downbows. These are ultra-critical, even on long notes. Some strings VIs don't distinguish between the two, so it's quite easy to have an entire track of downbowed tones. A good player can match upbows and downbows, but that's another topic. For the sake of sequencing with samples, the variety of bowings contributes mightily to the lack of ear fatigue- having many up bows and many down bows per articulation comes in ultra handy. Otherwise, prretty soon the ear picks up on a sameness with every note-- the attack, the vibrato, the envelope, etc. Having as many variations of any one single articulation as is reasonably possible itself gets closer to emulating reality than just about anything else. Reality would render a unique *sample* for every note played, whereas a good VI would allow for 3 or more round-robin staccatos (meaning 3 or more staccatos which are randomly or regularly swapped with each attack to prevent the machine gun effect with repeated notes).
I'm finding that with the best string VIs, you can still tell it's not the real deal. Once again, I think the better approach is to not worry so much how real your sequence sounds as much as it is to focus on how un-fake you can make it. Hardly anyone is going to be fooled into thinking, "oh, what orchestra is that"? It could happen, but for the most part there will be some obvious difference. The less obvious the difference, the more successful the mock up.
I can't even say which string VI is best for one endeavor or another because it also presupposes a certain lack of feasibility where the prices of these VIs vary so much. So, at least for now I won't praise one VI or put down another.
I will say that you need to start with a library that you feel sounds great in quality. Don't worry so much about character, because YOU (the maestro) will breathe the essential life into the instruments later on. The next thing to consider is how extensive the library is. To buy a library that just has patches called "violin legato" and "violin staccato" requires further investigation. Up and down bows again are important for sonic variation to avoid a flat, stagnant quality. But understanding string articulations in general will help assess what a library has to offer. There are spiccato and detaché, among many other playing techniques used that the ear at least viscerally identifies with realism. There is a difference between notes played "on" the string from notes played "off" the string. The list goes on and on, but understanding how these bowing techniques impact the sound from attack to sustain to release will help give you a better idea of how to emulate these important subtleties where a VI may fall short of offering them to you right out of the box.
One thing you might want to try is to fly in a track of a solo or section violin part into DP and just focus on a measure or two. Find a set of patches that cover the essential articulations in that one- or two-measure excerpt and see if you can "spank" your VI samples into some kind of submission to get as close as possible with continuous data. You may want to even automate filters or even EQ settings as the passage goes along (very slightly) because samples will tend to have the same brilliance or dullness on every note. This is another thing that comes across as artificial.
Ultimately, the question to ask yourself is "how much time do I have to deal with this"? It's not a quick process to say the least. But the end result can pay big dividends where the music itself is well-served by the extra effort.
There are some really decent mockups floating around the internet, and I will try to collect some mp3 links for you just for the sake of listening and discussion.
One thing you will hear again and again is that all of this has as much or more to do with the programmer as it does with the software. Some online demos of great sample libraries simply sound as if the person doing the sequence didn't have enough time or didn't care enough. Entire libraries have been passed by because of lousy demos and careless programming. And yet there are wonderful demos of less expensive libraries. Not that they sound more real, but the programmer started with a great piece of music and spent a lot of time really caring for the musical intent and the expression itself.
There's NOTHING at any price that replaces the real thing, but maybe enough users will be willing to share their tips on how to come away with an end result that needs no apologies.
I think your topic is eventually what most of us are aiming for, so to me it seems fitting. It's certainly not OT, and on some level it's not even *that* closely connected to DP issues. It's all related on some level, so it's going to be impossible not to talk about how-to's without mentioning DPs features or the features of various instrument plugins at some point.
With that said, I would first recommend looking at the approach a little differently. Instead of trying so much to bring virtual tracks to the level of real instruments, it may be an easier process to at least *start* by working on eliminating undesirable sonic elements in the virtual track with mechanical or technical qualities that take away from the listeners ability to remain focused on the music itself. Otherwise, you could spend the rest of your life and half another lifetime getting a violin patch to sound exactly like the real deal. I do believe, however, that you can get close.
Some VIs for strings allow for more vibrato control than others. The problem is, vibrato is not as steady from one note to the next or from one player to the next as most VIs would have you believe. Some vibrato control is better than none at all, and that's where most VIs drop the subject. Even VSL is having trouble at the moment with the larger string sections smoothly x-fading from non-vibrato to full vibrato-- except on the solo strings, which really do this quite well.
But when listening to your violin scale, each note has a different vibrato (or a different non-vibrato, as the case may be) and a different set of "imperfections" which could be analyzed and at least emulated (where imitation is beyond feasibility). Understanding just how the vibrato wavers irregularly might involved some pitch/wave analysis, and each note would have its own set of parameters. This digs deeper into the techno uber-geek underbelly than probably necessary, imho, where the pursuit of technical reality overshadows the pursuit of musicality.
Upbows and downbows. These are ultra-critical, even on long notes. Some strings VIs don't distinguish between the two, so it's quite easy to have an entire track of downbowed tones. A good player can match upbows and downbows, but that's another topic. For the sake of sequencing with samples, the variety of bowings contributes mightily to the lack of ear fatigue- having many up bows and many down bows per articulation comes in ultra handy. Otherwise, prretty soon the ear picks up on a sameness with every note-- the attack, the vibrato, the envelope, etc. Having as many variations of any one single articulation as is reasonably possible itself gets closer to emulating reality than just about anything else. Reality would render a unique *sample* for every note played, whereas a good VI would allow for 3 or more round-robin staccatos (meaning 3 or more staccatos which are randomly or regularly swapped with each attack to prevent the machine gun effect with repeated notes).
I'm finding that with the best string VIs, you can still tell it's not the real deal. Once again, I think the better approach is to not worry so much how real your sequence sounds as much as it is to focus on how un-fake you can make it. Hardly anyone is going to be fooled into thinking, "oh, what orchestra is that"? It could happen, but for the most part there will be some obvious difference. The less obvious the difference, the more successful the mock up.
I can't even say which string VI is best for one endeavor or another because it also presupposes a certain lack of feasibility where the prices of these VIs vary so much. So, at least for now I won't praise one VI or put down another.
I will say that you need to start with a library that you feel sounds great in quality. Don't worry so much about character, because YOU (the maestro) will breathe the essential life into the instruments later on. The next thing to consider is how extensive the library is. To buy a library that just has patches called "violin legato" and "violin staccato" requires further investigation. Up and down bows again are important for sonic variation to avoid a flat, stagnant quality. But understanding string articulations in general will help assess what a library has to offer. There are spiccato and detaché, among many other playing techniques used that the ear at least viscerally identifies with realism. There is a difference between notes played "on" the string from notes played "off" the string. The list goes on and on, but understanding how these bowing techniques impact the sound from attack to sustain to release will help give you a better idea of how to emulate these important subtleties where a VI may fall short of offering them to you right out of the box.
One thing you might want to try is to fly in a track of a solo or section violin part into DP and just focus on a measure or two. Find a set of patches that cover the essential articulations in that one- or two-measure excerpt and see if you can "spank" your VI samples into some kind of submission to get as close as possible with continuous data. You may want to even automate filters or even EQ settings as the passage goes along (very slightly) because samples will tend to have the same brilliance or dullness on every note. This is another thing that comes across as artificial.
Ultimately, the question to ask yourself is "how much time do I have to deal with this"? It's not a quick process to say the least. But the end result can pay big dividends where the music itself is well-served by the extra effort.
There are some really decent mockups floating around the internet, and I will try to collect some mp3 links for you just for the sake of listening and discussion.
One thing you will hear again and again is that all of this has as much or more to do with the programmer as it does with the software. Some online demos of great sample libraries simply sound as if the person doing the sequence didn't have enough time or didn't care enough. Entire libraries have been passed by because of lousy demos and careless programming. And yet there are wonderful demos of less expensive libraries. Not that they sound more real, but the programmer started with a great piece of music and spent a lot of time really caring for the musical intent and the expression itself.
There's NOTHING at any price that replaces the real thing, but maybe enough users will be willing to share their tips on how to come away with an end result that needs no apologies.
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52
This is a really important point. Having worked with orchestral sampling for a while now, I have found that the goal for me has shifted from trying to nail the real thing to just making it sound musical. I still emulate reality, but I have given up competing with it. It's a futile task, one that will end with fistfuls of hair removed from one's head.Frodo wrote:With that said, I would first recommend looking at the approach a little differently. Instead of trying so much to bring virtual tracks to the level of real instruments, it may be an easier process to at least *start* by working on eliminating undesirable sonic elements in the virtual track with mechanical or technical qualities that take away from the listeners ability to remain focused on the music itself.
You can still be expressive without the benefit of all the variables real performance has to offer, but you do have to make concessions to the limitations of your samples. I have heard many mockups (I hate that term - it assumes strict emulation and implies certain failure!) that try so hard to sound "real" they just sound like parody. Then there are others that work within the confines of the articulations available and sound very musical. Those are the ones I admire the most.
I think the trick is to think of a VI as an instrument in itself, rather than as analogous to the instrument it models. Like physical instruments, VIs have their own performance quirks and restrictions. They take a while to master. When I write with orchestral samples, I go through the same growing pains I did with my natural instrument. There are times I'm happy with my progress, and times I know I can do better. But it helps to constrain the goal first and foremost to sounding musical, not just "real."
Frodo- what a well written and thoughtful post- thanks for the insight! you have been quite appropriately pegged as moderator material for this sector of Unation! I'm still pretty new to all this (a few months at most sinceI started playing around with VIs more seriously than "sketchpad" mode) and there seems to be so much to learn! One thing I feel I pick up on intuitively is this aspect where synthesized parameters can accomplish something that samples can't really scratch the surface of- for example vibrato rate. Obviously it would be ideal approaching infinity if a sample library could have authentically produced vibratos at many varying rates and intensity, but even the best (largest in GB and most expensive) libraries will have only a pittance compared to the real thing. So it seems like, as far as vibrato goes and taking it as an example parameter, it might be better to have a non-authentic but more flexible synthesized vibrato working in conjunction with the high quality samples. The harmonic variation I brought up is another place where synthetic modeling seems better suited than repeat notes or some kind of sample-based approached. Not that having all those extra round robins is detrimental, but in a broad-stroke way, if you could CC control some more fundemental qualitative change, such as brightness, attack, it seems to me more likely to make the whole more dynamic than the sum of parts. This weird meeting place between sample and synthetic simulator seems to me like an analogy to something I've been familiar with via working with graphic design software for many years. For a while there was the pixel end of things and the vector end of things. The former was a more sample (hence RAM) intense, photographic approach and the latter (like these Wallander instruments or Synful) of endlessly scalable resolution, but completely computer generated, synthetic, on the other hand. In other words, low RAM, more CPU oriented. By and by, the line between the Photoshop and Illustrator worlds started to blur and the real potential in the field seems to lie somewhere in between, in mixing it up or, more to the point, knowing which approach to rely on where. Mixing up the real and synthetic to the point where the realness is no longer an issue seems to me to be the main goal to strive for. In Citizen Kane's "big" scenes, such as the political speech scene, most of the "crowd" was painted models. (Also with Titanic's huge crowd scenes- mostly CGI "people" plus some real extras.) Enough real people were woven in to moot the issue, which is what I'm also hoping to learn to do. There's an analogue to, say, painting a forest landscape- you don't articulate EVERY branch, since the eye doesn't see things that way, but you balance the detail against the mass.
Anyway, It's late and I'm rambling- thanks again for your contribution to all this!
Anyway, It's late and I'm rambling- thanks again for your contribution to all this!
A pleasure, Zas. Great post by you and Blue, btw. Very cool.zaster wrote: Anyway, It's late and I'm rambling- thanks again for your contribution to all this!
Ya know-- I's a long-winded cuss, ain't I?
Gosh-- there's so much to this and mastery is slow going.
It's almost 2 AM and I'm zonked, but here's one mockup I wanted to share-- solo cello:
Ernst Bloch "Prayer" MP3
It's not a matter of its reality which is always up for debate, but its "unfakeness" averts many of the pitfalls associated with virtual mockups.
Another issue we've yet to get into is the mixing process, but that's a different matter than conveying musical expression virtually.
more to come...
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52
Blue- I didn't see your post when I posted mine before (I think you posted while I was "composing"!) Your advice is great and may indeed save me some fistfuls of hair as I return to this all today!
Frodo- that link was VERY well done! was it yours (not the composition, the "mock-up")? Isn't the piano real? I actually can't hear anything in it to suggest otherwise so you must let me know!
Frodo- that link was VERY well done! was it yours (not the composition, the "mock-up")? Isn't the piano real? I actually can't hear anything in it to suggest otherwise so you must let me know!
It's all Memorex, Zaster-- both cello and piano. For the sake of discussion, I won't say just yet what the library is.zaster wrote:Frodo- that link was VERY well done! was it yours (not the composition, the "mock-up")? Isn't the piano real? I actually can't hear anything in it to suggest otherwise so you must let me know!
No, I didn't do this one, but I thought that the person who did it showed great understanding of the cello's bowings, great clarity of countless degrees of vibrato control-- different types of release samples (that connective material which does more to convey legato than just overlapping MIDI offs to MIDI ons, etc. You can actually hear where a cellist's change of hand position necessitates the same finger be used on successive notes--- resulting in a tiny little gliss of sorts only between certain notes and only at specific places where a real cellist might do the same change of hand positioning. Different cellist? Different fingering, bowing, and interpretation-- (and different amount of that little gliss effect). Incidentally, that glissando is probably more accurately referred to as a portamento-- but in sequencing you use whatever sounds the most organic.
Of course, there is a wonderful sensitivity with expression that doesn't sound merely like someone moving a volume fader up and down. There is a genuine range of dynamics used from pp to f, maybe a ff once in a while.
In addition, there is a certain pysycho acoustic phenom happening with the way the two instruments sit in a recital hall and not a concert hall-- and how the piano sits "behind" the cello. This is where the mixing process becomes inseparable from the sequencing, although it is a different discussion that goes on ad infinitum.
Moreover, the programmer showed a great love, tremendous respect and depth of understanding for the music itself. By avoiding the unfake (rather than imitating the real), I think they actually got closer to emulating reality than a lot of people might have.
Would you believe that some of the super geeks have satiated themselves by discussing what's wrong with the mockup? I say, why bother? I've referred a few cellists to this same mp3. Casually, they'd asked why I'd sent it (all of them knowing the work very well). None of them paid too close attention when they heard it because they are listening for different things than we do. The fun part was seeing the looks on their faces when they learned it was all done on a computer! Upon second hearing it was more obvious to them, but they found it more remarkable. One of them could only respond with a single swear word.
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52
Hey-- this mp3 showed up in another thread, but I thought it would be fitting to have it here as well. It's a pretty darn good orchestral mock up, imho-- more of an in-your-face cinematic mix rather than a concert hall sort of setting.
But, it's great writing for a start. This guy really understands the workings of an orchestra and how instruments function together and individually.
Where most people would write single notes, observe where the composer/orchestrator chooses double bows per note in the strings right from the very top. By doing so, he firmly establishes a rhythmic momentum that sets the stage for the rest of the composition.
Again, many have tried to pick this mix apart-- but taking the "unfake" perspective-- it's a FAR cry from GM!!
"MOJO MADNESS" by Thomas J. Bergersen
He is part of a group who is developing their own custom orchestra library somewhere in central Europe, so the library he uses is not commercially available.
But, it's great writing for a start. This guy really understands the workings of an orchestra and how instruments function together and individually.
Where most people would write single notes, observe where the composer/orchestrator chooses double bows per note in the strings right from the very top. By doing so, he firmly establishes a rhythmic momentum that sets the stage for the rest of the composition.
Again, many have tried to pick this mix apart-- but taking the "unfake" perspective-- it's a FAR cry from GM!!
"MOJO MADNESS" by Thomas J. Bergersen
He is part of a group who is developing their own custom orchestra library somewhere in central Europe, so the library he uses is not commercially available.
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52
Frodo! You must tell me! I would have to guess it's VSL cause the link was off the VSL site and I know you're a big VSL-er. But none of the piano demos on the VSL site sound like that (and piano isn't really one of those "tweak the mock-up" instruments where the programmer's skill makes or breaks it). I haven't checked out your 2nd mp3 yet but I will shortly. The cello on the Bloch mp3 was so detailed that whew! how does one attain that level of skill? Wish there was a MIDI document to look at or a screen shot like on the beat kaufmann pages where he shows you which samples he's working in where. (If you know of any such study resources please link us!)
Mwwwoooo-ha-ha-ha-ha--- Nope.zaster wrote:Frodo! You must tell me!
Okay.
It is VSL solo cello.
This was done before the VSL piano was available, so I'm gathering more specific data and info for you.zaster wrote: I would have to guess it's VSL cause the link was off the VSL site and I know you're a big VSL-er. But none of the piano demos on the VSL site sound like that (and piano isn't really one of those "tweak the mock-up" instruments where the programmer's skill makes or breaks it).
Actually, you may be in luck-- hang on>>>>>>>zaster wrote: I haven't checked out your 2nd mp3 yet but I will shortly. The cello on the Bloch mp3 was so detailed that whew! how does one attain that level of skill? Wish there was a MIDI document to look at or a screen shot like on the beat kaufmann pages where he shows you which samples he's working in where. (If you know of any such study resources please link us!)
Okay:
BLOCH PRAYER MIDI FILE
This zipped folder contains a MIDI file which opens right up in DP-- conductor track and all. As you can see, the piano was indeed a VI-- but I'm not sure quite what it was. My best guess is that it was Ivory or similar. I'm still looking for any notes on how it was mixed and will pass that along once I track it down.
Now, the zip folder also contains an .fxp file which probably won't work unless you had the same libraries installed used in the sequence. My guess is that Logic was the original DAW used-- or Cubase (not sure). I know that DP was not used. There's no guarantee that any cello patch from any other library will work with this sequence. In fact, I wouldn't even attempt it. But such things as continuous data as they relate to the longer and shorter notes will be an interesting study. Keyswitches will be confusing to see, and there may be some other VSL-specific data which won't make any sense to you.
I still think 90% of the battle with mastering this virtual world is in understanding the music first and as much as possible about how it's to be expressed-- being able to recognize the cresendos and decresendos (over phrases as well as over individual notes) as well as how the vibrato detailing coincides or contrasts with the bowing-- how dynamics work in conjunction with bow speed---
Ugh. It's a lifetime of work that can give you a headache when tackled in a single day.
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52
So you think it's VSL solo celo from the new VSL VI series or an older set? That would seem to make the articulation maneuvering even more impressive since I hear the new VI set takes care of some of that more primitive manual labor (though, short of winning the lottery, I'll probably never know fir$thand!)
But that piano! I have ivory but as good as it is i've never heard it sound like that! I'd love to know for sure what it was and how it was done.
And thanks for the MIDI file- I will delve into it and see what secrets come to light!
But that piano! I have ivory but as good as it is i've never heard it sound like that! I'd love to know for sure what it was and how it was done.
And thanks for the MIDI file- I will delve into it and see what secrets come to light!
Last edited by zaster on Thu Jun 21, 2007 2:49 pm, edited 2 times in total.
It's from Vienna Instruments collection, not from the earlier Pro Edition.
I think to some extent the brand new Special Edition might work with this as well. It's the same console, but I'm not totally up on which samples used in the Bloch might also be available in SE. For the most part, the SE has nothing new that the full set of Vienna Instruments doesn't have since SE is a subset of the larger collection-- same samples and all.
So, since we've put VSL up, it's only fair to put up at least one example from other libraries.
Still, I hope not to spawn a discussion on a "this library" vs "that library" except where musical issues are concerned. For example, I have Miroslav Philharmonik and like it, but it doesn't have string section trills. There are some trills on some solo strings and limited trills available on other instruments. Neither have I been able to find any muted strings except on a small violin section, so scoring something with a complete section of muted strings is simply impossible. That's more of a quantitative judgment rather than a qualitative one. What I do like about Miro is that I can mix larger string sections with smaller ones (4 or less) plus add in solo strings for another layer of complexity. That alone keeps the ear guessing and is a good thing.
One thing to consider, too, is in how these libraries work. What most of us are used to is changing tracks and MIDI channels for each articulation or sample. VSL is so tightly integrated that everything you need for a single instrument can be programmed onto one single track (as you may have seen from the MIDI file). With Miroslav and EWQLSO, I tend to have 8 or more tracks of first violins. With VSL, I only have one track for most everything except where I might have the Appassionata Strings loaded as well (different expansion set) or the Chamber Strings loaded. Even here, there are never more than 5 tracks per string sample set (v1, v2, va, vc, and cb).
more to come...
I think to some extent the brand new Special Edition might work with this as well. It's the same console, but I'm not totally up on which samples used in the Bloch might also be available in SE. For the most part, the SE has nothing new that the full set of Vienna Instruments doesn't have since SE is a subset of the larger collection-- same samples and all.
So, since we've put VSL up, it's only fair to put up at least one example from other libraries.
Still, I hope not to spawn a discussion on a "this library" vs "that library" except where musical issues are concerned. For example, I have Miroslav Philharmonik and like it, but it doesn't have string section trills. There are some trills on some solo strings and limited trills available on other instruments. Neither have I been able to find any muted strings except on a small violin section, so scoring something with a complete section of muted strings is simply impossible. That's more of a quantitative judgment rather than a qualitative one. What I do like about Miro is that I can mix larger string sections with smaller ones (4 or less) plus add in solo strings for another layer of complexity. That alone keeps the ear guessing and is a good thing.
One thing to consider, too, is in how these libraries work. What most of us are used to is changing tracks and MIDI channels for each articulation or sample. VSL is so tightly integrated that everything you need for a single instrument can be programmed onto one single track (as you may have seen from the MIDI file). With Miroslav and EWQLSO, I tend to have 8 or more tracks of first violins. With VSL, I only have one track for most everything except where I might have the Appassionata Strings loaded as well (different expansion set) or the Chamber Strings loaded. Even here, there are never more than 5 tracks per string sample set (v1, v2, va, vc, and cb).
more to come...
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52
I think that some of it may be the reverb used-- not sure what that was. It just captures the pre-delays of the hammer strikes quite convincingly.zaster wrote: But that piano! I have ivory but as good as it is i've never heard it sound like that! I'd love to know for sure what it was and how it was done.
Sure-- (it's actually cool to see you soak all this stuff in so voraciously!)zaster wrote: And thanks for the MIDI file-
Not really. Some people are pianists and others are programmers. I think it could be a combo of two things: Having unique control of left hand and right hand expression (which is something to consider where such expressive requirements work independently of one another) and also-- I remember seeing some data about mic positioning-- so whatever was used may have had some sort of separate MIDI tracks for near/far info.zaster wrote: EDIT: So far I see- 4 piano tracks? It's almost broken up into SATB. Strange. (Is this some kind of typical technique?)
It could have just been easier for the programmer to play parts in one note at a time. There's no saying that he didn't step-enter the notes and sweated the continuous data editing on the back end.
Yuppers-- lots of stuff going there. But I think the MIDI file contains all the MIDI data used in the sequence.zaster wrote: Expression and Pitch Change are the only controllers operating on the Cello track. The velocity data is all over the place. (BTW- I sure wish there was some better way to see veocity data, I hate these miniscule little ys!) There's gotta have been more going on than what's on this MIDI file, dont you think? Ooh wait a minute- keyswitches- whoa! 2 octaves worth! Well I guess now at least I can try to figure out which articulation goes with which keyswitch by comparing the MIDI to the mp3! Fun fun fun!
I'll see if I can get the project and the 'fxp' loaded in Logic in case there is other self explanatory info about how the matrices were set up. If I can transfer that data somehow to DP we can sort out how to get the DP-format patch config file to you if you thought it would help. Otherwise, I can just make a list of patches organized by their matrix positions.
So much of this, as mentioned earlier, won't make complete sense unless you have VSL up and running.
Last edited by Frodo on Thu Jun 21, 2007 3:04 pm, edited 3 times in total.
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52
So far I see- 4 piano tracks? It's almost broken up into SATB. Strange. (Is this some kind of typical technique?)
Expression and Pitch Change are the only controllers operating on the Cello track. (Are there more within the VI interface that you can't see on the MIDI track?) The velocity data is all over the place. (BTW- I sure wish there was some better way to see veocity data, I hate these miniscule little ys!) There's gotta have been more going on than what's on this MIDI file, dont you think? Ooh wait a minute- keyswitches- whoa! 2 octaves worth! Well I guess now at least I can try to figure out which articulation goes with which keyswitch by comparing the MIDI to the mp3! Fun fun fun!
I understand your point about the not wanting to start the library debates. I assure you that I for one am not even looking at this that way. But it's good to know the source because the way the articulations are put together varies so much from library to library. (BTW- I discovered that the Garritan mutes I'd been using in a track and grown fond of were actually just an EQ setting within Kontakt. I was able to approximate the same sound on violins from othe libraries by dropping a similar EQ in! Maybe you could do that for your mutes- on the plus side, it would allow you to use all your articulations!)
EDIT: Oooh! Everytime I'm writing a post you are already posting something else! So read this one as though it was between those last 2!
Expression and Pitch Change are the only controllers operating on the Cello track. (Are there more within the VI interface that you can't see on the MIDI track?) The velocity data is all over the place. (BTW- I sure wish there was some better way to see veocity data, I hate these miniscule little ys!) There's gotta have been more going on than what's on this MIDI file, dont you think? Ooh wait a minute- keyswitches- whoa! 2 octaves worth! Well I guess now at least I can try to figure out which articulation goes with which keyswitch by comparing the MIDI to the mp3! Fun fun fun!
I understand your point about the not wanting to start the library debates. I assure you that I for one am not even looking at this that way. But it's good to know the source because the way the articulations are put together varies so much from library to library. (BTW- I discovered that the Garritan mutes I'd been using in a track and grown fond of were actually just an EQ setting within Kontakt. I was able to approximate the same sound on violins from othe libraries by dropping a similar EQ in! Maybe you could do that for your mutes- on the plus side, it would allow you to use all your articulations!)
EDIT: Oooh! Everytime I'm writing a post you are already posting something else! So read this one as though it was between those last 2!
LOL!!zaster wrote:EDIT: Oooh! Everytime I'm writing a post you are already posting something else! So read this one as though it was between those last 2!
Yes-- Now you have some idea how passionate about orchestra VI programming I've really been. I'd promised Zed that I'd get going on Fab Four today. I've been more or less on U-nation since 8 this morning (it's now after 3 PM and I've yet to get to the studio!!)
Indeed-- there are all sorts of tricks to combining sounds from different libraries, and EQ balancing is indeed unavoidable where blending and sculpting is concerned.
But the nice thing about muted strings is that you can get away with a whole different library without worrying about your "main" library. If I need muted strings, I can fly in sections from VSL or EWQLSO and then use some other library for the "senza sordini". In a way, this doesn't make total sense because I'd more than likely rely more on EW or VSL for the string sounds in general-- but you probably see what I mean. (I know-- bad example!
)Ah. It's 95 degrees outside-- not a cloud in the sky-- and I'm sitting indoors dressed like a CNN correspondent! That's going to change in a big hurry-- and it's almost Miller Time!!
Hey-- isn't today (or wasn't yesterday) the first official day of summer?
6,1 MacPro, 96GB RAM, macOS Monterey 12.7, macOS 10.14, DP9.52