+- Scoring Central (http://scoringcentral.mattiaswestlund.net)
+-- Forum: Technology (http://scoringcentral.mattiaswestlund.net/forumdisplay.php?fid=5)
+--- Forum: Samples & Sample libraries (http://scoringcentral.mattiaswestlund.net/forumdisplay.php?fid=8)
+--- Thread: SSO 2.0 (/showthread.php?tid=208)
RE: SSO 2.0 - peastman - 03-31-2018
I've checked in the initial version of the "violin solo 2" instrument: https://github.com/peastman/sso. Comments and suggestions will be very welcome! Here's what I've done with it.
- Lots of general QC: removed bad samples, fixed pitches, removed silence at the beginnings of samples, made the volumes more consistent.
- I created marcato articulations by combining the arco and spiccato samples. You can use the mod wheel to control the strength of the attack.
- Instead of having two different pizzicato articulations, I combined them in round robin. There's barely any difference between the vibrato and non-vibrato versions, so this seemed like a better use of the samples.
- I added a keyswitched version.
RE: SSO 2.0 - Samulis - 03-31-2018
I gave a look-through of the Violin 2 samples. It seems that the noise is just bow noise and not actual background noise. This is the result of recording at a medium/close distance (rather than farther out as one might be used to with orchestral samples in a concert hall), and the same sort of thing is audible in VSCO 2. It can be removed, but it is quite characteristic to a violin mic'd at this distance.
I did notice a little bit of broad spectrum noise in the pizzicato samples and after a few of the releases (sounded a bit like someone fed it through a denoiser at some point and it left some "undigested crumbs"), a little pitchy stuff in the upper register overall (just pass it through a tuner and it'll be alright), and one of the tremolo samples sounded like it might have had a car in the background, but other than that, these are some fantastic samples for sure- very consistent. If you want, I can see if I can go through this early next week; I have a big project this weekend.
RE: SSO 2.0 - peastman - 03-31-2018
Thanks, that would be great! For samples with obvious clicks, thumps, squeaks, and the like, I mostly just skipped them and used one of the adjacent samples instead. I still kept all the wav files in, but they aren't used by the instruments.
RE: SSO 2.0 - Mattias Westlund - 03-31-2018
(03-31-2018, 02:49 AM)Samulis Wrote: It can be removed, but it is quite characteristic to a violin mic'd at this distance.
Yes, I would suggest leaving it in. Personally I would much rather use an instrument with natural noise included than one that has major parts of its frequency content missing from noise removal.
RE: SSO 2.0 - kneedeep - 03-31-2018
Another vote for "no noise reduction". For the reason Mattias stated. If the user deems this noise unacceptable, they may use a lowpass or even a de-esser or de-noiser of some kind in their DAW. I did hear the noise, but it became pretty unnoticeable after setting the instrument up with some (properly high-and lowpassed) reverb and EQ.
peastman: I listened to your violin 2 and it sounds good to me. I'm not 100% sure about combining the pizzicati, but I can't see it being a problem either.
It is a lot of sfz files though. What do you think about using keyswitched versions exclusively? I know I don't want to have 5 solo violin tracks in my DAW for different articulations, for performance reasons alone (sforzando, reacontrolmidi and an eq on each of them does add up). So I'm likely to go all keyswitched. I wonder if there's an advantage to keeping the non-keyswitched versions, except backwards compatibility?
RE: SSO 2.0 - peastman - 03-31-2018
It sounds like the consensus is for no noise removal. So it shall be!
The choice between keyswitches and separate files depends on the software you use and your workflow. I use the separate files in MuseScore, but the keyswitch files in Logic Pro. Because of the way the programs work, different approaches end up being better in them. I think Mattias has also said he uses a separate track for each articulation?
I've been going through the other files, and I really like the alternate solo flutes. Mattias, do you remember where the samples came from? I notice the staccato articulations only have two octaves, while all the others have three. Are all of these actual recordings, or were the trills generated synthetically?
RE: SSO 2.0 - peastman - 04-01-2018
I noticed something odd in the flute instruments. I suspect it was just a mistake, but I want to make sure before I "correct" something that was actually done for a good reason! They're sampled in minor thirds, so each sample is used for three notes. But they aren't the three I would have expected. I would expect each sample to be used for the three notes surrounding it, so the C# sample would be used for C, C#, and D, and then the E sample for D#, E, and F. But instead, the C# sample is used for B, C, and C#, and then the E sample for D, D#, and E. So it often is shifting a sample down by a whole step.
Was that intentional or just a mistake?
RE: SSO 2.0 - kneedeep - 04-01-2018
That doesn't sound right to me.