What's new

Samplemodeling Solo & Ensemble Strings Released

I wanted to share another resource in case anyone else may find it helpful. These videos are classics and you've all probably seen them before, but at least for me, I found it enlightening to review them:




I went back to these videos mainly as a refresher on playing techniques, as it's been quite some time since I actually touched a string instrument. But in watching them again, I was struck by a few things that might be helpful when learning all the controls on the SM strings, particularly the tone-shaping controls.

Granat demonstrates different kinds of bowings and how they affect tone production, in a somewhat dry context (not dry as in reverb, but in the sense that when he plays, he's not trying to be expressive, and the examples aren't melodic - just isolated notes and scales).

For me at least, when referencing a recorded performance it's too easy to get seduced by the music, and end up focusing quite a bit on that even when trying to focus elsewhere. For anyone who suffers the same affliction, the examples in these videos will be valuable precisely because they are in isolation, sans expressivity. (Note: I doubt it would be useful to try to match this recording precisely -- at least on my system, it sounds a bit too mid-rangy and seems to suffer from some close-mic "presence effect." Other recordings will be better for precise side-by-side matching. But hearing the tonal differences he produces in relation to each other, outside of the context of a musical line, I think could be very valuable when learning to shape the timbral controls, at least for people like me who aren't wizz-kids at it.)

When he demonstrates sul tasto, it occurs to me that with enough experimentation, this must already be effective and attainable using the built-in timbral controls. (Ponticello is probably a different story; that sound is so chaotic and ever-changing, we probably need Samplemodeling to work some magic for us in a future update.)

Another thing I realized from these videos that I'm surprised I didn't notice before: the sympathetic resonance of the open strings ringing out whenever the bow or the left hand doesn't attempt to damp them. Is there a way to do that with SM strings?

And of course, there's far more value in these vids than what I've mentioned here. If anyone reading this hasn't already seen them, you're missing out!


(p.s. The two vids listed above are the instrument-specific ones. Part 1 in the series doesn't discuss the instrument, but has some delightful little anecdotes from Granat's career.)

(p.p.s. Has anyone out there purchased the Itzhak Perlman masterclass? It's on my wish list, but for the foreseeable future, "spare money" = 🦄)
 
It's a hommage to the depth of SM strings that folks seem far more demanding than usual! (Mind you, I only have Embertone, Chris Hein, Kirk Hunter, and Xsample.) In fact Xsample has more bells & whistles than most: ringing open strings, Bartok pizzcato, bowing behind the bridge....
 
I had a hard time deciding whether I should get this or not. Some of the Ensemble demos sound just unbelievably great while others put me off completely. However, the SM brass is my favorite of all libraries ever made on this planet. The AM woods and strings sound okeyish (great playability but often with an unnatural synthetic sound quality to it) and it was my hope that SM could surpass the achievements of AM in the sound quality domain.

After reading this thread, I decided to give it a try as the most convincing parts of the sound demos are the things that conventional string libraries struggle with (dynamics that feel human, runs, general playability). And I must say I'm super happy that I put my doubts aside. This thing is awesome. I just played it for some minutes and am blown away by the playability and expressiveness.

I also experienced some things that I didn't find convincing at all, but I'll use the next days to study the manual and get more insight into how this library works and how it's supposed to be played.

However, as @Giorgio Tommasini is watching this thread, I have some feedback already that I wish SM to consider for the next update. To this post attached are two files. "SM Pizzicato Bugs - MoreVariationsPlease" is a simplistic rendering of some pizzicato notes on a g and a c# played on the solo cello. It sounds like the instrument uses just a single pizzicato sample per note without any variations. It seems to me there are 3-4 dynamic layers in use, but within each dynamic layer a different velocity just leads to a different volume of the same sample. This makes the sound of multiple pizzicati in a row quite machine gun like to me. I think this would greatly benefit from adding variation samples!

Also attached is the "SM Pizzicato Bugs - StrangeSustains", again the solo cello playing pizzicato. This time I'm striking with an increasingly higher velocity each keystroke. What I hear is that the first g in the range of velocity 90, the c# starting with velocity 50 and the last g in the range of velocity 90 sound more like a bad piano than cello pizzicato. This seems to come from a kind of sustain layer fading in too strong - at least that is how it sounds to me. I guess this needs more subtle adjustment? Also irritating with the last g repetitions is how playing the lowest velocity the pizzicato is quite rich of high frequencies, around velocity 50 these high frequencies vanish suddenly and around velocity 90 the high frequencies are back.

Normally, SM instruments are extremely good at transitioning from one point to another (for instance ppp to fff or non-vibrato to vibrato). To me, the pizzicatos are not at the level I'd expect from them.

The slapped pizzicato of the solo cellos g in the file "SM Pizzicato Bugs - StrangeSustains" sounds like a g# instead of a g. I guess this is just a sample placed at the wrong key?

Anyway, I'm happy I got the library and am thrilled to explore more of it, tomorrow.
 

Attachments

  • SM Pizzicato Bugs - MoreVariationsPlease.mp3
    114.3 KB · Views: 73
  • SM Pizzicato Bugs - StrangeSustains.mp3
    264.8 KB · Views: 62
Last edited:
Without a proper walkthrough video displaying and showcasing the library's functionality I find it very hard to justify the purchase. The examples so far aren't as ground breaking as I would have hoped for and the technical aspects that might set this library apart from the competition are obscure.
Finaly, the gui of the library is way below the today's standard. I would have expected something more towards audiobro 's genesis and modern brass gui' s. IDEALLY, a clear and coherent interface with most of the functionality fully accessible from the main page and with player count, position and distance graphically displayed.
I hope that future development and a proper walkthrough will convince me to buy the library, I own all of the sm (audiomodeling nowadays) strings but I am not satisfied with their tone at all so, I had so many hopes from samplemodeling to pull this off.
 
Without a proper walkthrough video displaying and showcasing the library's functionality I find it very hard to justify the purchase. The examples so far aren't as ground breaking as I would have hoped for and the technical aspects that might set this library apart from the competition are obscure.
Finaly, the gui of the library is way below the today's standard. I would have expected something more towards audiobro 's genesis and modern brass gui' s. IDEALLY, a clear and coherent interface with most of the functionality fully accessible from the main page and with player count, position and distance graphically displayed.
I hope that future development and a proper walkthrough will convince me to buy the library, I own all of the sm (audiomodeling nowadays) strings but I am not satisfied with their tone at all so, I had so many hopes from samplemodeling to pull this off.
As another SM/AM owner of all their other products, I completely agree with this.

I don't think developers can expect to release libraries these days without a full exposition of it, particularly given what their competition make available. Compare this to Spitfire for example, typically comprising Paul's walkthrough, Oliver/Homay's 'in action', and Christian's contextual accompanies all of the major releases, in addition to 4-5 audio demos. Of course I acknowledge that Spitfire are a much bigger company with more resources and personnel, but what would it take for SM staff to produce a simple walkthrough of each instrument & patch, and show the DAW contents for one of the demos that were created for the library. I was extremely impressed by the audio demos they provided (though others apparently weren't), but this is not sufficient for me to buy a library, as this doesn't show me what the library is likely to be like in MY hands. If SM let their intro price period expire without uploading a detailed walkthrough video, I think they will lose a lot of potential customers. I hope they don't make that mistake, because I think they deserve for it to be a success, because from what I hear of it so far, it's a wonderful instrument.

I also made this comment on pg 1 of this thread, and SM haven't responded to this point, so I hope the fact that others have expressed similar opinions, will prompt action from the developers. This is your opportunity to really showcase what you have achieved!
 
Last edited:
Okay, here's an example of default vs. realtime timbral shaping. The default example is the violin ensemble the way it loads initially, except that it's been switched to breath controller mode. The MIDI data is the same for both examples. I performed it with the timbral shaping then removed the timbral CC data for the default example. It's mostly unedited except to fix wrong notes. The timbral shaping example has the high overtone control linked to the low overtone control instead of to the middle overtone control as I describe in my previous post on this. This linking also doesn't seem to work correctly in the engine -- if you set two timbral groups to the same CC, only one of them reads the CC data, so I ended up having to copy the low overtone data onto a separate CC for the high overtones.

There's still work to be done on this, for sure. There are sections in the timbral shaping example where it sounds too much like a filter being applied, and I think this is mostly because I'm pulling the middle overtones out too much and/or too abruptly. I think constraining the CC data to a smaller range will help combat this. But to me, the timbral shaping example sounds overall a lot more three-dimensional than the default example.
 

Attachments

  • pmcrockett - sm strings default.mp3
    825.6 KB · Views: 163
  • pmcrockett - sm strings timbral shaping.mp3
    837.2 KB · Views: 154
Finaly, the gui of the library is way below the today's standard. I would have expected something more towards audiobro 's genesis and modern brass gui' s. IDEALLY, a clear and coherent interface with most of the functionality fully accessible from the main page and with player count, position and distance graphically displayed.

I don't know... I just need the GUI for setting up the instrument and the current one gets the job done. Once I'm done with that, the rest happens through controllers and MIDI editors. No need to constantly open up Kontakt.

Without a proper walkthrough video displaying and showcasing the library's functionality I find it very hard to justify the purchase.
Something like a video walkthrough would have been great, indeed.
 
From what I hear in most of the demos I actually think that the overall sound is quite beautiful. After all, all of this is virtual. Even a recorded live performance is not the real thing but virtual in some way.

I want a convincing outcome. I think that, depending on the composition, a lively expressive performance could be more convincing for listeners than the perfect natural "tone" in a lot of occasions.

I am looking forward to a full walkthrough and more demos to see if it fits my specific needs.
 
Your efforts towards plumbing the depths of these VI's are greatly appreciated.

Would you consider printing one of these with more of a dynamic vibrato thing happening?

Okay, here's an example of default vs. realtime timbral shaping. The default example is the violin ensemble the way it loads initially, except that it's been switched to breath controller mode. The MIDI data is the same for both examples. I performed it with the timbral shaping then removed the timbral CC data for the default example. It's mostly unedited except to fix wrong notes. The timbral shaping example has the high overtone control linked to the low overtone control instead of to the middle overtone control as I describe in my previous post on this. This linking also doesn't seem to work correctly in the engine -- if you set two timbral groups to the same CC, only one of them reads the CC data, so I ended up having to copy the low overtone data onto a separate CC for the high overtones.

There's still work to be done on this, for sure. There are sections in the timbral shaping example where it sounds too much like a filter being applied, and I think this is mostly because I'm pulling the middle overtones out too much and/or too abruptly. I think constraining the CC data to a smaller range will help combat this. But to me, the timbral shaping example sounds overall a lot more three-dimensional than the default example.
 
OK, so I'm having some luck, but not there yet. Speaking only about the violin, it is fairly accurate to say I'm aiming for a sound/feel that is most similar to the violin in the Anne Dudley Poldark Theme. To say I'm botching that would be also accurate.

Also, I've been composing with Maschine, mk3. I use the Maschine pads and also my S88 keyboard to write and fiddle about. What I NEED to do is to move my creation process back over to ProTools ... Maschine's MIDI implementation and editing is horrible, and while the setup is conducive to creative flow, it makes it next to impossible to fix things without breaking other things, including, but not limited to that fact that it's not saving some of my input for some reason. It is also why you will hear dropouts in the violin. (the rest of the flaws are just me being lame ... it's a work in progress, and I'm very slow).

So here are two iterations of the same song: One with Fluffy Audio's Trio Bros Violin, which is what I used when writing. The other version is of course the SM Violin, which for the life of me I cannot make it NOT sound like a clarinet.

THe piano is NI Noire, and the cellos are Chris Hein.

I WISH WISH WISH that the makers of SM Strings had included several style presets - even as broad as "Classical" "Gypsy" "Contemporary" that were playable out of the box. I also wish for a way for us to share snapshots or presets ... is there? So far my attempts to save those always come back incomplete ... something is always missing, including, but not limited to, Timbral Shaping presets ... ???

Here is the SM Violin version:

Here is the Fluffy Audio version:
 
Your efforts towards plumbing the depths of these VI's are greatly appreciated.

Would you consider printing one of these with more of a dynamic vibrato thing happening?
There's a bit of vibrato in the examples I posted, but I find the vibrato in general to be pretty subtle except at very high values. Not a very musical demonstration, but here's a note played with increasing vibrato with the built-in reverb off. The vibrato rate is set to 77, the intensity increases from 0 to 127 over the first five notes, then the sixth note adds expressive vibrato set to 74. Then the whole example repeats with the reverb turned on.
 

Attachments

  • pmcrockett - sm strings vibrato.mp3
    972.7 KB · Views: 102
OK, so I'm having some luck, but not there yet. Speaking only about the violin, it is fairly accurate to say I'm aiming for a sound/feel that is most similar to the violin in the Anne Dudley Poldark Theme. To say I'm botching that would be also accurate.

Also, I've been composing with Maschine, mk3. I use the Maschine pads and also my S88 keyboard to write and fiddle about. What I NEED to do is to move my creation process back over to ProTools ... Maschine's MIDI implementation and editing is horrible, and while the setup is conducive to creative flow, it makes it next to impossible to fix things without breaking other things, including, but not limited to that fact that it's not saving some of my input for some reason. It is also why you will hear dropouts in the violin. (the rest of the flaws are just me being lame ... it's a work in progress, and I'm very slow).

So here are two iterations of the same song: One with Fluffy Audio's Trio Bros Violin, which is what I used when writing. The other version is of course the SM Violin, which for the life of me I cannot make it NOT sound like a clarinet.

THe piano is NI Noire, and the cellos are Chris Hein.

I WISH WISH WISH that the makers of SM Strings had included several style presets - even as broad as "Classical" "Gypsy" "Contemporary" that were playable out of the box. I also wish for a way for us to share snapshots or presets ... is there? So far my attempts to save those always come back incomplete ... something is always missing, including, but not limited to, Timbral Shaping presets ... ???

Here is the SM Violin version:

Here is the Fluffy Audio version:

I think the main trouble is note attacks, which need to be spikier and more varied, particularly in sections like at 1:20. I'm not sure how exactly you have MIDI input set up, but if you're using the standard CC11 for expression and aren't using a CC11 controller that gives you a high degree of control over spiky attacks (e.g. a breath controller), you may need to play around with Vel. to Dynamics (Controllers 2 page), which lets you dial in key velocity to control attacks (higher means the key velocity has more influence). Attack Time (Controllers 1 page) sets the length of time that it takes for attack expression set by key velocity to go back to the CC11 value, which isn't explained too clearly in the manual (EDIT: It's explained in the Playing Techniques/Detached Notes section, not the controllers section). I think high attack time is probably better than low for most things, but I haven't played around with it much because I'm using a breath controller and have Vel. to Dynamics disabled.
 
Last edited:
I think I'd agree with that ... my attacks are actually "missing". I am using TEControl Wind controller .. is there a significant difference between that and a breath controller?

I don't have Vel. to Dynamics disabled ... hmmm.

The "clarinety" sound I complain of is during sustains. No matter how I fiddle with (I know .. the pun, the pun!) the Timbral shaping, it is always there.

I think the main trouble is note attacks, which need to be spikier and more varied, particularly in sections like at 1:20. I'm not sure how exactly you have MIDI input set up, but if you're using the standard CC11 for expression and aren't using a CC11 controller that gives you a high degree of control over spiky attacks (e.g. a breath controller), you may need to play around with Vel. to Dynamics (Controllers 2 page), which lets you dial in key velocity to control attacks (higher means the key velocity has more influence). Attack Time (Controllers 1 page) sets the length of time that it takes for attack expression set by key velocity to go back to the CC11 value, which isn't explained too clearly in the manual. I think high attack time is probably better than low for most things, but I haven't played around with it much because I'm using a breath controller and have Vel. to Dynamics disabled.
 
I think I'd agree with that ... my attacks are actually "missing". I am using TEControl Wind controller .. is there a significant difference between that and a breath controller?

I don't have Vel. to Dynamics disabled ... hmmm.

The "clarinety" sound I complain of is during sustains. No matter how I fiddle with (I know .. the pun, the pun!) the Timbral shaping, it is always there.
TEC controllers are actually what I'm referring to when I say breath controller. I'm using a TEC as well. So I'd definitely suggest trying it with Vel. to Dynamics turned down to 0. I have Breath Controller mode turned on (Breath controller page), which changes expression from CC11 to CC2 and more importantly auto-disables Vel. to Dynamics and changes it so that notes cut off completely when you're not breathing through the controller. It's a matter of taste, really, but I prefer it.

As far as playing technique goes, especially on the solo instruments I rearticulate with a tongued tah on the controller for any note that should be a bow change. I only use a continuous air stream if the notes are on the same bow. Holding the sustain pedal also creates a velocity-based bow change on the next note, though I haven't used that much so I'm not sure how it compares to rearticulating with the controller.
 
Here's Saxer's example processed using an effects chain that's pretty similar to what I often use on other dry instruments. (Saxer, if you don't want me to post this, let me know and I'll take it down.)

I gotta say, I'm liking how this sounds.

For anyone wondering, the specific effects I'm using are (in order):
Waves TransX Multi (transient shaping to emphasize attacks on upper frequencies)
Type Writer X (stereo imager/reverb to slightly widen the stereo field and fuzz the sound up a bit)
Flux Spat (early reflections from projecting the sound upward)
EW Spaces 2 (reverb, Abravanel Violins-Celli FR TS 3.0s)
Zynaptiq Intensity (harmonic exciter/compressor, sort of -- brings out the character of the sound a bit)
Overloud Gem EQ550 (hardware modeled EQ)
I much prefer the raw sample..this sounds way too processed for my taste..
 
Apologies ... I had my terms messed up. Yeah .. "Breathcontroller" is what I'm using. Can't seem to make it deliver the same attack unless my eyeballs are bulging. Vel to Dynamics is off. Not starts at the slightest breath, and stops when breath ends.

Unfortunately, too, another detail I dislike has come up ... exactly the same as in the Brass, try to taper all the way to silence ... can't do it, huh? Note tail does not taper, but cuts off near CC11 1-7. The fix? fade it with Volume. Oops, that's a fail. How many instruments in real life can taper a note all the way to silence without dropping off? Yeah .. .most.

I still do not like the sound of these instruments. Here I am literally begging the makers to come up with some presets as starting points that actually sound/feel like real instruments. Having all the control is great, and desired, yes, but one should not have to build the whole sound! Again, I'm comparing not just to real instruments as heard live, or by the player personally, but mainly, based on recordings. Recordings of real instruments. That's what I want them to sound like.

Reminds me of the digital character modeling world. With few exceptions, digitally sculpted trees, water, animals, humans, etc. tend to look and move like .. well ... animated models. Even some of those films that used real actors as bases for the sculpts, there is always some giveaway that it isn't real (credit of course given for artistic or creative and skillful mockups ... it takes TREMENDOUS skill to pull off.) Is this not an apt comparison?

Yes, these instruments do sound almost real and the music made by the masters who make them sound almost real is awesome ... sheesh, I wish I were that good. But all too often it is more ensemble/symphonic sound, not naked solo.






TEC controllers are actually what I'm referring to when I say breath controller. I'm using a TEC as well. So I'd definitely suggest trying it with Vel. to Dynamics turned down to 0. I have Breath Controller mode turned on (Breath controller page), which changes expression from CC11 to CC2 and more importantly auto-disables Vel. to Dynamics and changes it so that notes cut off completely when you're not breathing through the controller. It's a matter of taste, really, but I prefer it.

As far as playing technique goes, especially on the solo instruments I rearticulate with a tongued tah on the controller for any note that should be a bow change. I only use a continuous air stream if the notes are on the same bow. Holding the sustain pedal also creates a velocity-based bow change on the next note, though I haven't used that much so I'm not sure how it compares to rearticulating with the controller.
 
Okay, here's an example of default vs. realtime timbral shaping. The default example is the violin ensemble the way it loads initially, except that it's been switched to breath controller mode. The MIDI data is the same for both examples. I performed it with the timbral shaping then removed the timbral CC data for the default example. It's mostly unedited except to fix wrong notes. The timbral shaping example has the high overtone control linked to the low overtone control instead of to the middle overtone control as I describe in my previous post on this. This linking also doesn't seem to work correctly in the engine -- if you set two timbral groups to the same CC, only one of them reads the CC data, so I ended up having to copy the low overtone data onto a separate CC for the high overtones.

There's still work to be done on this, for sure. There are sections in the timbral shaping example where it sounds too much like a filter being applied, and I think this is mostly because I'm pulling the middle overtones out too much and/or too abruptly. I think constraining the CC data to a smaller range will help combat this. But to me, the timbral shaping example sounds overall a lot more three-dimensional than the default example.

Thank you so much. I like the tone too! I'm curious as to how one might go about writing a divisi line where the divisi'ed strings interact (e.g. if two divisi lines were to "merge"). Would you recommend increasing the section size for the note where they merged? Or would it be sufficient to simply write a note with Divisi 1 and Divisi 2, without it sounding phasey? @Giorgio Tommasini - any tips perhaps on how to go about this and have it sound fluid?
 
Clarinety sustains?

This comes from un-modulated sustains which a violinist just cannot do!
Even without vibrato, our bowstroke, plus the intermingled non-static resonances of the violin body cannot rival the clarinet's sustains.
Two solutions: baked-in irregularities as in Chris Hein's non-vib samples,
or SM's "perfect" sustains plus CC28 (attack detuning) and CC33 (dynamic modulation of tone and pitch).

Through a closed door, I have even very briefly mistaken a sax for a student's violin: odd + even harmonics, unlike the clarinet's odd-only ones. The basic timbre is similar, as the AM viola shows...

But I should like a gentle randomisation of vibrato amount and rate.
And I agree about stylistic presets to seduce the most demanding ear before we get going..
 
Last edited:
OK, so I'm having some luck, but not there yet. Speaking only about the violin,...

I WISH WISH WISH that the makers of SM Strings had included several style presets - even as broad as "Classical" "Gypsy" "Contemporary" that were playable out of the box. I also wish for a way for us to share snapshots or presets ... is there? So far my attempts to save those always come back incomplete ... something is always missing, including, but not limited to, Timbral Shaping presets ... ???

Here is the SM Violin version:

Here is the Fluffy Audio version:


Hi Gene,
in addition to the mentioned points, I would also say that in the Fluffy version Vibrato enters earlier and more intense. The SM sequence is lacking the Vibrato at an extent that my feeling is it's not yet programmed in the same way. And you know how critical is a good vibrato for realism and expression... If you share the MIDI file, I can adjust it and send it back if you like.
 
Top Bottom