Hey guys. I'm back, this time with a solo vocal library.

It was recorded as a proof of concept for a comprehensive legato recording method when applied to vocals. Instead of sampling sustains and transitions separately, each transition leads into a full length sustain, so there's only one crossfade per legato transition. It's a pain to record, but I think the results sound really good.

Download link and a demo video are at ivyaudio.com/ClareSolo

As always, your feedback and comments are greatly appreciated, and feel free to send me anything you create using the library.

Cheers,

Simon

 

Cheers to you Simon, for your amazing products. I quite enjoyed your piano as well

 

Sounds great, but hard to work with!

Thanks

 

Thanks, very much! This will sound great through a glitcher over my beatz. Just kidding. The library is really thoughtfully executed, and it's hard to make a "soloist" library sound so real. Your piano is really great-sounding, too.

 

Thanks for the effort. It has a nice, mellow tone. But it's not really usable. The releases overlap each other, the legato can sound pitch-shifted/stretched at some intervals/points and I get a lot of latency compared to other instruments. It's like the sound is coming out a few beats/ms later or the script can't tell what point in the sample to begin from, where to stop the previous release and fade the next attack (or something like that). I found the overlapping groups a bit strange, dunno. I tried to mess with the script and releases but it didn't really change anything. The raw samples are pretty good, though. Good approach nonetheless.

 

Hey there,

I tried to cover a little bit of why it's like this in the video, but you're right, if you play notes too quickly, the release samples will overlap, and the transitions are very slow so you have to anticipate by a good deal. It works best if you place notes manually in your DAW and tweak them to match the timing you want. It's a finicky little library, and requires a bit of work to get sounding good.

Any chance you could let me know which transitions sound off? I applied light pitch correction and manually compared each sample to its original, so they should all sound pretty natural, but it's not inconceivable that I might have botched a few.

Thanks for the feedback.

 

Sorry, i hadn't watched the video when I posted. I thought it would be an announcement of shorts and/or an audio demo, so I skipped it. I jumped straight to testing it.

By "pitch-shifted" I mean like the pitch of the portamento/legato sounds a bit artificial at parts as the transitions sound like you're doing it with the pitch wheel. It's not that big of a deal though. I don't think any tone is really off but I think I hear some artifacts every now and then. Here I made a little version of the Lullaby from Pan's Labyrinth using the library:
http://uptobox.com/fb767la6sv3k

It has a bit more complex lines than your demo. You certainly need to work on it quite a bit (as a user). It could be a bit better, I'm sure. Since I needed small releases I disabled the "1" group on the releases (I changed the "allow group" to 0 and it worked quite well). If you listen carefully there is some noise/drop-out like interference at some points which I'm not sure if it's because of the change in the script or something else (maybe recording artifacts). It's pretty good overall. *yes*

One thing I noticed is that the duration/release of the notes plays a role in the character of the transition. Not just the start point.

Maybe there could be a speed knob for the transitions?

 

Hey, no problem. The transitions were left pretty much untouched as well, so I guess it's just the character of her singing giving that effect. Also I can't access that link, says it's not available in my country :/

I prefer the sound of the releases in the "1" group just because they're actual recordings rather than Kontakt fading out the note, but that's a matter of personal preference.

As for the transitions, I think the differences you're hearing are a result of the fact that the sustains do not remain at one constant volume, so it's very difficult to get a transition speed that will result in zero dip or rise in amplitude. That's another thing I was hoping to address later in scripting. Ideally I'd like to build a matrix that contains the best transition time and sample offset values for each individual transition, but that's a lot of work and I figured I may as well release the library in its present form. So yeah, definitely lots more scripting to do!

Thanks again for your feedback. This is exactly the kind of thing I'm looking for when making these libraries.

Any chance I could get a re-upload of that demo?

 

Yeah, the transition could be made more precise with time machine or a script that chooses a point in the sample that will act as the transition or speeds up the transition and/or cuts/fades the release according to the parameters set. It's a lot more work of course so it's logical you'd want to avoid it or leave it for later. Either that, or you can just use it for slower passages... It doesn't have to do everything.

Here's a Zippyshare link for the track:
http://www25.zippyshare.com/v/52217155/file.html

 

Gramofon is awesome, he was actually on a team that helped in the development of a free library. Now you know why.

Thanks so much for another excellent library Simon though some kinks need to be worked out to make it shine it's brightest. However you're in good company and plenty of people would be happy to assist. *yes*

 

Have had some time to deal serious with you newest release!

It works great with slow stuff indeed! Maybe you should consider add more vocals like: Ahh Ehh etc .... and add the possibility play several voices at the same time as a choir!

Cheers