Hi there,
having loads of string libraries, I was utterly disappointed that the authors recorded all legato intervals, except for perhaps the most crucial one - perfect 1st.
How can you compose a musical phrase without it? It's not loure, just bow change on the same note.
Spitfire string patches and libs are out of question, not possible at all
LASS tried but it's still triggering the same sample
Adagio has Loure, but that's not legato and it's very cumbersome

It's very easy to record and program (not that I know how to script it ), just record a few bow-changes on every note and then let them be triggered through playing with sustain pedal, good idea would be to record several RRs to avoid overusing single sample when there are multiple notes.



Up bow/down bow, is not the same (though quite close) as triggering those samples will disconnect the phrase (as with Spitfire)

Is there any composing technique that would help me to get that interval to sound as close to reality as possible?

Thanks!

 

a lot of useful scripts here:

http://www.vi-control.net/forum/portal.php?page=7&sid=6850d3b5497043a549f7e2ac2955d643

http://nilsliberg.se/ksp/scripts/sips/sips.htm

and on this Kontakt forum

http://www.vi-control.net/forum/viewforum.php?f=65

maybe we can find some intresting for us

Also, VSL library has lot of bowe up/down articulation with modwheel, but it has not legato with G-player. It must be used with VSL Performance Tool, and that's the other story.

Kontakt Factory VSL legacy has legato patches of every single instrument, and I see in script editor description "kontakt performance tool for legato" but I'm not having time to exeriment with this yet.

 

Thanks nikon for your answer,

The problem is not with legato transitions, those are well recorded in all mentioned libs.
Only the perfect 1st is NOT recorded. Such legato transition should be simple bow change, so I doubt SIPS would be any of use here (it's great for winds though).

I thought of a solution, not a bright one, but still great.
In Mural samples, the bow change is recorded. By extracting that portion of a sample, one can recreate such effect, but that'd have to be inserted as a WAV sound to DAW, which would complicate things a lot, and the result might not be worth the extra hassle.

Still strange though, perfect 1st is an interval like any other, but no company thought of sampling it.

 

I remember I see in some lib 1st perfect (perf unison) bowe change with modwheel (up/down), but where I don't know !?!!!

 

Wow, Andrew, you're inside my head, haha. I was thinking about this just yesterday when reading the other thread where nikon was outlining the VSL stuff. This always bugged me but I never got to say it out loud. Thanks for taking the step. *yes*

I wouldn't call it 'legato' necessarily. More like a repeated, tied note. Or "re-bowing" (with controllable/stretchable bow changes). Something like the tool that Cinesamples used for Cinestrings. Just not for dynamics but repetition. For me that's the biggest problem for strings that's never been addressed (triggering the same note, that is). I also think it's partly due to Kontakt's design where you can't "slide" through the keys but have to hit them one by one. PLAY has that (for example, you can start with one note and then "slide" the mouse on all keys for a continuous sound) and although it isn't a "re-bowing" (or true legato - leaving the script aside) sound, it doesn't cut off as abrutply as Kontakt libs do. It gives a more tied sound but with weird attacks when hitting the same note. But it's not nearly as bad as in Kontakt (you can also tame it with some automation). This was mainly one of the things I tried to showcase with my fast strings test (the lags in transitions/repetitions).

Of course, it should be possible to fix it through scripting... If I was making a lib this would be my 'selling' feature. Hey devs, I'm providing you with ideas to steal here!

I also notice that in staccato patches where some (like Symphobia) have nice releases at parts that make transistions sound more realistic, others cut off really quick in a way you can feel it's separate one-shots in succession. I mean, ok, that's what they are but they could be treated in a way that could mask that. Now, I'm not saying a software would be able to completely mimic a performance but I don't think it's been pushed as much as it could (if at all).

 

ahh, yes I remember now, It's in Sonivox library:

V1 Legato 3-Lyr Down MW_XF
V1 Legato 3-Lyr Down
V1 Legato 3-Lyr MW_Bow
V1 Legato 3-Lyr Up MW_XF
V1 Legato 3-Lyr Up
V1 Legato Down Bow
V1 Legato Down MW_XF
V1 Legato MW_Bow (I think that's the one)
V1 Legato Up Bow
V1 Legato Up MW_XF
V1 Script Legato Down Bow NR
V1 Script Legato Down MW_XF NR
V1 Script Legato Up Bow NR
V1 Script Legato Up MW_XF NR

 

Yeah, or the "duplets" and "triplets" scripts of the old Kirk Hunter strings. But still, those can give you wrong values or playing non-existent notes. It's tricky.

 

In EastWest SO also exist in 18Violins and 11Violins with keyswitches A0 (up) and A#0 (down) but it not work as I see. A#0 no response at all, but I have not 0 octave on my midi controler so I must try with DAW, maybe it's combination with A0 + A#0 etc...

 

In Sable and Mural there are 2 RR for Longs, I'd say that's upbow/downbow, but that glue in between is missing

I stand corrected
Friedlander violin has perfect 1st

 

Yes, this is a great sounding solo violin!

I think that developers are having a difficult time devising the best way to approach these big orchestral libraries. Most are more concerned about how good the mock up will sound when it's an edited sequence in a DAW. Less are focused on realistic playing techniques that let you EXPRESS the capabilities of the instrument in REAL TIME, live playing. As a musician (vs. a 'composer/engineer/etc') I prefer to be able to PLAY these sounds, even if it means mastering pedal/mod wheel/cc change techniques.

 

True, but one would expect the expressive abilities from Spitfire, known for including incredible amount of articulations for strings.
Sadly, that's not the case

 

Really? How can you tell? But, even if it is:

- Is it up/downbow?
- Is it looping (that means up/up re-trigger for example)?
- Is the bow transition programmable? (Free Time-Stretching)
- Does it adapt to tempo (optional)?
- Can you move it around?
- Is it integrated in the sample? If it is, it won't trigger at your interval but when you reach its point in the sustained sample.
- Maybe it's just a result of RR? RR would be a different sample, with a (slightly) different attack but that doesn't necessarily mean it's "connected", rather emulated for better or for worse. Listen to this (Friedlander):
https://www.firedrive.com/file/B585934EA552587B

See how it lags on repetitions. It can make pretty good trills though. Way better than LASS solo violin.

Different approaches. All viable but with varying results.

Another example here:
https://www.firedrive.com/file/FF3582C3BED877A7

LASS solo, Friedlander solo, Both (LASS [L] - Fr [R]) and normal recording with up/downbow. Friedlander has better overall transitions (audible "glue", although I'm not sure about its nature, but it sounds rather good) methinks but it lags a bit. It's these little details that can make a big difference.

As to how it could be solved, here are some thoughts:
- Alter the dynamics a bit
- Alter the vibrato intensity a bit
- Put on some automated (de)tuning (Try the Humanizer script)
- Slight attack time alteration or note event start delay
- Alter the volume a bit
- Use a slightly different patch
- Layer a noisier or shorter sample at the note change/re-bowing point
- Change to Time Machine and experiment

But it's messed up overall. In a normal performance I doubt it would all be actual "sustains" (would be shorter, noisier, late, fast, ... Also depending on context). The performance (programming) is definitive and quite random so... Just try your best.

 

Here's my attempt to rebow using Friedlander: http://www.firedrive.com/file/E4A3163580F65D70
It's a bit choppy (2 minute sketch), but not bad *yes*

 

in EWQL I found this, martinele up/down in short articulationa and legato RR patches

 

Unfortunately "Legato" patches in EWQL SO aren't true legato, it's just another name for Sustains

Wow, martinele ? That must be very loud
Even martéle is quite pronounced

And while we're at it, we can discuss double stops

 

It seems to be just upbows with RR but it's a good try, especially at the beginning. I can hear some lags again after that.

The Butter legato and QLegato from EWQL have some potential but yes, they're not really legato (SIPS?). They just have different attacks. In short notes, it's mostly the releases that matter for that connection. Having the Up/Downs (for movement mostly) and RRs (more variable character) can definitely help. EW is very responsive and varied overall.

 

It appears that the only library capable of accomplishing this is Live Orchestra

 

Right.

(But seriously, there is a line. There are times when I think: Is this (programming hours) maybe too far? The next step would be to record a live orchestra... But it's a bit different story when you don't have that access to live players and it's your final release. I mean, you should try to make it as good as possible. Nowadays, the word "mock-up" is partially losing its meaning I think. Anyways...)

 

ha ha

wow martinele type mistake