Scoring Central

Full Version: VCSL
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
(06-11-2018, 03:13 PM)Michael Willis Wrote: [ -> ]Hey Sam, I have another sampling question. Let's say an instrument is sampled in whole steps. Do the results tend to be better if the gaps are filled in by tuning up lower notes (C -> C#, D -> D#, etc) or tuning down higher notes (D -> Db, E -> Eb, etc), or does it not really matter?

If the answer is that neither tuning up nor tuning down are inherently better than the other, then my next question is why bother sampling in whole steps? Why not minor thirds, in which case each sample can cover three notes by tuning both up and down a half step?

As BC said, tuning down is generally considered better, as it typically more closely resembles the acoustics. Keep in mind that repitching a sample is semantically the same as using a shrink/enlarge-ray on an instrument- it makes the entire instrument, including the resonant body, larger or smaller. That's why low repitched brass is all blatty like a tuba and why if you repitch a trombone by an octave it makes a rather convincing trumpet. However, no instrument in real life works like this, as they all have a constant (or near constant) size. Even trombones, which use a slide, have a consistent diameter of bore, so when the slide is farther out, the actual ratio of tube diameter to length of the air column becomes lesser, giving it a brighter, more narrow tone ever so slightly.

With minor thirds, you have three notes with identical attacks when you play a chromatic scale as opposed to just two with wholetone. Keep in mind that the advantage of chromatic sampling in playing realistic quick lines with clean attacks is quite significant, as each note you play in a quick line will be guaranteed to have a slightly different attack. Without chromatic sampling or round robins (to pick up the slack), someone playing something like 'flight of the bumblebee' is going to sound pretty bad.
Well, now I'm mighty tempted to sample chromatically! I'm going to give it another go with the clarinet. Last time I tried holding each note for quite a long time, but then I was out of breath in between, which made it kind of a pain, and I didn't even get a consistent amount of time on each one. I'm thinking of picking a fixed number of seconds and playing each note for that long. I'll see how high in the range I can go, it can be difficult to get pleasant sounding, in-tune notes in the altissimo register.

Oh, one more thing, I noticed some discussion about octave numbers earlier. Just to clarify, is 440Hz considered A3, A4, or A5?
(06-11-2018, 07:39 PM)Michael Willis Wrote: [ -> ]Well, now I'm mighty tempted to sample chromatically! I'm going to give it another go with the clarinet. Last time I tried holding each note for quite a long time, but then I was out of breath in between, which made it kind of a pain, and I didn't even get a consistent amount of time on each one. I'm thinking of picking a fixed number of seconds and playing each note for that long. I'll see how high in the range I can go, it can be difficult to get pleasant sounding, in-tune notes in the altissimo register.

Oh, one more thing, I noticed some discussion about octave numbers earlier. Just to clarify, is 440Hz considered A3, A4, or A5?

Keep in mind that temporally, recording chromatically with 1rr and wholetone with 2rr take the same amount of time and, in a situation with a run, sound similar. However, 2rr gives you the advantage of better repetitions, while chromatics gives you a little more accuracy to the original instrument- which may or may not be a good thing (such as a 'sour' note that just doesn't speak all that well).

Similarly 3rr minor thirds vs. chromatic.

It's all a balancing act- velocities vs. RR vs. pitch/scale vs. articulations. Each term multiplies with the others, so finding which you need and which are expendable based on your use case are the most important thing to do when planning.

Most sample libraries use C3=60="middle C", however, the official international standard is C4=60="middle C". The latter is because most samplers and such were made in Japan, which used the C3 standard. You will even find some DAWs such as FL Studio use C5, to avoid using negative C's (e.g. C4=60 has a C-1, and C3 has a C-2!). Personally, I use C3=60 most of the time as it is the most common method in use today, but as peastman (unfortunately) found out, I have at times used C4=60.
(06-11-2018, 08:19 PM)Samulis Wrote: [ -> ]Personally, I use C3=60 most of the time as it is the most common method in use today, but as peastman (unfortunately) found out, I have at times used C4=60.

Just to add some confusion, when I went looking for information on ranges of various musical instruments, I usually saw C4=60=middle C. For example:

https://www.vsl.co.at/en/Instrumentology/Woodwinds

http://www.orchestralibrary.com/reftables/rang.html


C4=middle C is apparently considered "scientific notation"

https://en.wikipedia.org/wiki/Scientific...requencies

As a result of the above links, and because it's what Sonatina Symphonic Orchestra used, I built my library around C4=middle C. I remember double checking and noticing that VSCO2 CE, at least for the violins also uses C4=middle C. Does VSCO2 CE not always use C4=middle C?
(06-12-2018, 02:59 AM)Paul Battersby Wrote: [ -> ]
(06-11-2018, 08:19 PM)Samulis Wrote: [ -> ]Personally, I use C3=60 most of the time as it is the most common method in use today, but as peastman (unfortunately) found out, I have at times used C4=60.

Just to add some confusion, when I went looking for information on ranges of various musical instruments, I usually saw C4=60=middle C. For example:

https://www.vsl.co.at/en/Instrumentology/Woodwinds

http://www.orchestralibrary.com/reftables/rang.html


C4=middle C is apparently considered "scientific notation"

https://en.wikipedia.org/wiki/Scientific...requencies

As a result of the above links, and because it's what Sonatina Symphonic Orchestra used, I built my library around C4=middle C. I remember double checking and noticing that VSCO2 CE, at least for the violins also uses C4=middle C. Does VSCO2 CE not always use C4=middle C?

Uh, no, most of VSCO 2 is C3=60. I haven't really used C4=60 since the early stuff (xylophone, marimba, etc.).

C4=middle C=60 is the "correct" method, but Kontakt, Maize, and many other samplers, DAWs, even keyboards and hardware synths/samplers use C3=60, so it sort of is the default choice for samples.
(06-12-2018, 05:16 AM)Samulis Wrote: [ -> ]Uh, no, most of VSCO 2 is C3=60. I haven't really used C4=60 since the early stuff (xylophone, marimba, etc.).

I think I understand the source of my confusion with regards to VSCO2-CE.  I do see this in SViolinVib.sfz


Code:
<region>
sample=LLVln_ArcoVib_C4_p.wav
lokey=59
hikey=61
pitch_keycenter=60
lovel=0
hivel=62


... and knowing that SSO uses C4=60 everywhere, I figured VSCO2-CE was the same. I should have looked more closely.

I now see that other VSCO2-CE strings label the middle C waveform file as C3 and play that waveform when MIDI note 60 is received.

So, my mistake. Taking a closer look, VSCO2-CE does use C3 labelled waveforms for middle C, here's ViolinEnsSusVib.sfz (B2 is being used in place of C3):

Code:
<region>
sample=VlnEns_susVib_B2_v2.wav
lokey=58
hikey=60
pitch_keycenter=59
lovel=63
hivel=127
volume=7

... but sometimes (at least for the solo violin), uses C4 labelled waveforms as middle C.

All that really matters in the end is that when I play middle C, I hear middle C regardless of how the waveforms are labelled.
(06-12-2018, 12:13 PM)Paul Battersby Wrote: [ -> ]All that really matters in the end is that when I play middle C, I hear middle C regardless of how the waveforms are labelled.

That's the idea!

I've been trying to standardize VCSL as C3=60 as I go... at least that way if someone wanted to convert it into C4=60, they could make a fork and write a neat little script.

If you notice anything slips through the cracks, let me know.
A start at that GM rompler using 11 of the VCSL instruments, though some are 'lite' versions to keep the size manageable.

Win 64 - http://www.mediafire.com/file/w8bdmi2dyz...4.zip/file

Win 32 - http://www.mediafire.com/file/u4u0n27njq...2.zip/file

Mac AU - http://www.mediafire.com/file/5gueu784x5...t.zip/file

Mac VST - http://www.mediafire.com/file/7w6cyaqmc3...t.zip/file
It's been (quite) a while since the last VCSL update, but I have some good news coming-
Melodica: https://instaud.io/3DAT (stretched)
Hammered Dulcimer: https://instaud.io/25zU (stretched)

In addition, I have a bunch of new woodwinds which will be making their way into the project at some point in a few months-
https://instaud.io/3fxS
https://instaud.io/3bt2

We're also planning on a donationware/low-cost 'professional' Kontakt version of VCSL to be developed concurrently to VCSL's free SFZ version, with lots of special features and presets. That should help offset development costs and hopefully help us record some new stuff just for VCSL (most of the stuff so far is either little things I can record in 10-20 minutes myself or things I already had lying around from tests or cancelled projects).
Wow, that dulcimer sounds wonderful! Very dynamic and expressive, which I think is slightly unusual for sampled dulcimers. At least the ones I have have very few velocity layers. Looking forward to this one!
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16