Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Forum Jump  
Real release samples and convincing attacks on strings
Last post Mon, May 02 2011 by DG, 10 replies.
Options
Go to last post
Posted on Fri, Apr 29 2011 05:35
by Hicks
Joined on Sun, Feb 28 2010, Posts 48
Hi all,

first of all thank you for all those kind libraries.

I am trying to solve an issue I have since a long time and not only with VSL but with all libraries I am using, but as VSL is my favorite...
As I am a violin player, I am very careful about those samples and the issue with ALL strings samples are the release.
With short notes, all the note length seems to have been recorded, so on Staccato, Spiccato, Pizz, harsh, detache short (I am speaking about chamber strings here), the sound is full and beautiful.
However, on long note, I can't hear the strings players stopping their bow on the string at the end of the note.
For example, on sustain, when I release the keyboard note, the note cut abruptely. Ok The release function is adding some trails, but it's more like reverb to get a smooth cut than a real sample of a player stopping his bowing.
And this is he main thing which makes me tell when listening to a music if it has been done by samples or orchestra. It happens everytime and everywhere.
The srangest thing is that it also happening on detache long, but it should be a full length sample? The note is finishing like a flute (a very quick decrescendo like someonoe stopping to blow), not a string instrument.
So my question is there a way to add real release of bow samples? MAybe I am using wrong the samples.
It is not only VSL but all libraries, but maybe it's a hint for next gen libraries!
Other question is regarding attacks. With legato perf, if two notes are not overlapped (depending on ms defined in VIP) a sample with attack is heard.
Great. But sometimes I prefer an attack from another sample. I mean, on a phrase C D E in 8th note. I want a slow attack to start and then all in one bow.
If I am using only legato perf, the attack on the first note is heard but little two quick in the passage I want (nothing wrong with samples, but matter of taste), but D and E are perfectly connected.
So I wanted to use a sustain sample for the C and legato for the rest. But even if the keyswitch is trigger during the C, the D note (even if overlapped with the C) is heard with the attack from the legato patch. The E is ok.
Is there a way to make the legato perf patch start by a slur note if it is overlapped even if the note is the first one of the patch.

Thanks a lot for your help.
Posted on Fri, Apr 29 2011 09:43
by DG
Joined on Wed, May 12 2004, Posts 8608
  1. When a string player stops the bow, it is never abrupt; there is always a release in pressure first, no matter how small a timescale. With the VSL samples what you hear is more like what would happen if a player lifted the bow off the string very quickly at the end, but even this is not what it souds like in the real world. So what you need to do is to do a quick xFade down at the end of your sustain, which will give the impression of the bow stopping much better. You also may need to reduce the volume by a quick, probably parabolic curve (using Expression) so that the release sample is not too loud.
  2. This should not be the case. If you keyswitch whilst playing the initial note, when you play the second note, you should hear the transition material and then the note itself. There should certainly be no start note for the legato second note. Maybe you could post an example, so that I can hear what you are talking about.

DG

Nuendo 6.03, 4.3
2 x Intel Xeon x5675 3.07GHz Hex Core
48GB RAM
Windows 7 (x64)Pro
RME Multiface II
Intensity
ATI HD5400 series graphics card
Posted on Fri, Apr 29 2011 10:35
by Hicks
Joined on Sun, Feb 28 2010, Posts 48
I will try to use velocity xfade to do that way. But are the release samples really "release samples" of players stopping, cause it sounds to me, especially on sustain for example that this is more like a sample which is cut. Instead of reducing release sample volume, I would like to increase it. But In VIP, when putting the faders higher, I only get the impression that release samples are some reverb artefact (on sustain I mean).

I will post something when I will see how I can record my screen (to show piano roll at the same time).
Posted on Fri, Apr 29 2011 10:45
by DG
Joined on Wed, May 12 2004, Posts 8608

To do it the way you want would mean that you would have to play the note much shorter, and the release sample would have to be much longer. it would also require multiple velocities and speeds of release sample. Even though I understand what you're asking for, and up to a point agree (although there are many more important things for me), I'm not sure that I would like to adapt my way of playing so drastically. Obviously I can't speak for the practicality of all this.

DG

Nuendo 6.03, 4.3
2 x Intel Xeon x5675 3.07GHz Hex Core
48GB RAM
Windows 7 (x64)Pro
RME Multiface II
Intensity
ATI HD5400 series graphics card
Posted on Fri, Apr 29 2011 11:03
by Hicks
Joined on Sun, Feb 28 2010, Posts 48
Hum, I understand what you mean.
But I think release samples can be used within a script like for the legato patches for only a few ms.
When finishing a phrase, you stop playing and the release sample will load.
It would make strings so more convincing. But I understand that velocity should be the sample as of the last sample played, but it works for legato, isn't it.

Even with some full length samples it is happening. If you have chamber strings, detache long doesn't sound like full note. The end of the sample is quite strange acting like a flute.
I have never heard players played like this. I think the samples have been tweaked for some unknown reason.
That's strange.
But I must say I have the same issue with LASS (regarding release samples).


Posted on Fri, Apr 29 2011 12:04
by DG
Joined on Wed, May 12 2004, Posts 8608

I think that the detache is just a single recorded note, whereas in the real world it is usually connected (even though usually a separate bow) to another note, thereby not creating the release that you hear in the sample.

FWIW I very rarely use the detache articulation for exactly the reason you describe, as there are better alternatives, but I have found on occasion that it is useful for layering to give a bit of extra definition to a legato or sustained note.

DG

Nuendo 6.03, 4.3
2 x Intel Xeon x5675 3.07GHz Hex Core
48GB RAM
Windows 7 (x64)Pro
RME Multiface II
Intensity
ATI HD5400 series graphics card
Posted on Mon, May 02 2011 08:55
by Hicks
Joined on Sun, Feb 28 2010, Posts 48
Hi DG,

I can tell you that a detache played alon doesn't end like this. It must has been edited. Strange.
However, as you said, stacked to keep only the attack makes pretty good results.

I have also solved my issue regarding sustain patches following by legato patch. The issue is that a even on a legato patch, the transition from G3 to C4 is weird if the G3 is the starting note. If I play F3 G3 C4, I don't hear the strange legato transition.
It is happening on some notes, but very few. I guess when recording a library, some transitions were better than others. But there are very few with strange transitions and there is always a way to get rid of this.
For example, on this example, I wanted a strong attack on G3, so I started the G3 with marcato perf interval and C4 with leg perf interval. The transition was really smooth.

I have another question,
I am using a lot the spic perf interval. Are the transitions recorded to get the "air" between the notes?
I mean, this patch is sounding really great and would like to know if each patch called "perf" got transitions recorded?

Posted on Mon, May 02 2011 09:13
by DG
Joined on Wed, May 12 2004, Posts 8608

 A single detache normally starts and stops with the bow on the string. The VSL one almost sounds as if the bow is lifted.

A suggestion for you, when talking about notes, please use C4 as middle C (as VSL does) because anything else is just confusing.

Playing G4 to C5 is not the same as playing F4, G4, C5. In the first example the G will be a start note, and in the second one it will be a legato note.

All Perf patches have transitions.

DG

Nuendo 6.03, 4.3
2 x Intel Xeon x5675 3.07GHz Hex Core
48GB RAM
Windows 7 (x64)Pro
RME Multiface II
Intensity
ATI HD5400 series graphics card
Posted on Mon, May 02 2011 11:26
by Hicks
Joined on Sun, Feb 28 2010, Posts 48
Strange,

on my previous message the word "S P I C" as been changed by "***".
Is it a german insult?

So even spiccato perf intervals are recorded with transition between notes!
That's great.

btw, a noob question, why VSL is using a figure up than standards. That's true that a C3 is named C4 within VSL convention.

Regarding, I know regarding legato perf interval, that starting my phrase with G4 will not use the same sample than the G4 used in the middle of it (starting sample vs leg transition one).
However what I wanted to say is taht my transition from G4 to C5 is far more realistic with "leg transition" G4 than "Starting note" G4.

Thanks DG for your answers. I am very happy with Chamber strings (at 85% to be true. I would have loved more than 2 vel layers for perf and the little issues on some legato notes).
I am planning to buy Appassionata or Orchestral to get a more lusher sound simulating a big orchestra.
Posted on Mon, May 02 2011 12:54
by DG
Joined on Wed, May 12 2004, Posts 8608

There are two main standards for Middle C; C3 (Yamaha) and C4 (Roland). However, Scientific Pitch Notation considers C4 to be Middle C.

DG

Nuendo 6.03, 4.3
2 x Intel Xeon x5675 3.07GHz Hex Core
48GB RAM
Windows 7 (x64)Pro
RME Multiface II
Intensity
ATI HD5400 series graphics card
You cannot post new threads in this forum.
You cannot reply to threads in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.