Using additional waveforms in V-Synth

[Iceman777]

Hello everyone! :) How's life in the Roland Clan?

Seeing how popular Serum and similar software could become, I was thinking about loading different, custom or additional waveforms in V-Synth via USB card. In other words, choosing file with a waveform instead of a file with a usual sample. Has anyone tried that? Do you think that such approach would work? Thank you! :)

Denis / Iceman

[LA Keys]

Have a look there:viewtopic.php?f=17&t=49516

This should get you started.

LA

[Iceman777]

Thank you so much for the reply and a helpful link! :) I have imported all the possible things as samples, but somehow never tried the waveforms yet. If it works the way I think it should, then it is basically possible to turn V-Synth in a revival-of-the-past machine, but importing waveforms from the classic machines or trying to create custom waveforms, which may seem (and sound!) even more interesting! I long time dreamt of becoming a user of Omnisphere 2 that allows importing samples, but could not afford it. Then I thought that it should be possible to do almost the same thing in V-Synth, once I edit the samples/waveforms in a right way. :) Thank you again for the reply! It is so good to be here again!

- Denis

[hehehehe]

I tried this with some weird results recently. Unpredictable ugly noises no matter how i trimmed the single cycle waveforms. Seems to me like something in its sampling engine makes it better with longer samples. A traditional rompler or even sampler would do this better I believe.

[LA Keys]

I tried this with some weird results recently. Unpredictable ugly noises no matter how i trimmed the single cycle waveforms.

If you refer to 'clicking' this is because the start and end point should be at the exact same level. This is sometimes very hard or impossible to achieve. You usually want to start with a short (few cycles) wave and ideally trim at the zero crossing point. Common beginner mistake is to use a longer waves with the ADSR envelope. The ADSR SHOULD be controlled by the v-synth at the patch level, not by the sample itself. Sampling is an art that's is not easy to master but once you do you can produce amazing thing with the v-synth.

Have a look at this:https://www.gearslutz.com/board/electro ... an-do.html

LA

[Iceman777]

Thank you so much for the replies and for amazing link! I have missed it somehow! :) You are absolutely right, LA, sampling is, indeed, an art that should be mastered, even if by trial and error. A wonderful thing with V-Synth is that you can take only one sample and process it for further work. Without a doubt, editing can be difficult, but I have also understood that shorter samples often work better. While it may sound and seem like a longer sound is better, it is the shorter sample that turns out in a better way! :)

Any further sampling recommendations or ideas? What has worked the best (or worst!) for you? Have you been able to imitate classic sounds/instruments/sequences? :)

- Denis

[hehehehe]

I tried this with some weird results recently. Unpredictable ugly noises no matter how i trimmed the single cycle waveforms.

If you refer to 'clicking' this is because the start and end point should be at the exact same level. This is sometimes very hard or impossible to achieve. You usually want to start with a short (few cycles) wave and ideally trim at the zero crossing point. Common beginner mistake is to use a longer waves with the ADSR envelope. The ADSR SHOULD be controlled by the v-synth at the patch level, not by the sample itself. Sampling is an art that's is not easy to master but once you do you can produce amazing thing with the v-synth.

Have a look at this:https://www.gearslutz.com/board/electro ... an-do.html

LA

I don't refer to clicking, more like (dis)harmonic noise and artifacts. What I used was perfectly trimmed single cycle samples. Also tried trimming them on the V-Synth. The results were much more random than with a standard rompler, so my conclusion still stands.

[Saxifraga]

I am a little late to this discussion, but what cyclic waveforms (A(t),t:0..T) need to work is defined by math. Not only the amplitude at the reset point A(T) and starting point A(0) of a wave have to be the same, but also the slope aka first derivative A(t)/dt of the signal. A step in a wave gives you a multitude of high frequencies that amount to noise. The Fouriertransformation of a step function for example produces massive amounts of noise.
You can't produce good sounding waveforms without deep knowledge about Fourier-Transformations between the time-signal and the frequency domain. Trial and error dosn't work very often.