User's Guide videos discussion

Discuss John Bowen Synths - Solaris
Post Reply
John Bowen
Site Admin
Posts: 1815
Joined: Sat Apr 07, 2007 3:00 am

User's Guide videos discussion

Post by John Bowen » Thu Oct 07, 2010 8:45 am

I've just finished a series of six short, informal tutorial videos that are exclusive to pre-order customers only. I'd like to provide this space for any discussion that these users would like to have.

Movement
Posts: 11
Joined: Mon Aug 09, 2010 2:58 am

Post by Movement » Fri Oct 08, 2010 3:45 pm

To generate a little discussion I thought I'd give some info that John sent me when I asked him about the multisampling. John, if this isn't supposed to be public for whatever reason go ahead and delete this post.

Anyway, I had asked if the multisampling was going to fleshed out more so a single oscillator could have samples spread across the keyboard and different samples would be selected based on velocity as well as random (or noise) sample selection with the idea being that you could (depending on key, velocity, and some random value) choose from multiple samples like:

C3LowVelocityVariation1.wav
C3LowVelocityVariation2.wav
C3HighVelocityVariation1.wav
C3HighVelocityVariation2.wav
C4LowVelocityVariation1.wav
C4LowVelocityVariation2.wav
C4HighVelocityVariation1.wav
C4HighVelocityVariation2.wav

Here was his response on the future of multisampling in the Solaris:
Yeah, we have discussed most of these functions. Of course you can do this already with the Waldorf tables, and should be the same thing for the User Wavetables.

For the multisample object - there may be an extension to the User Wavetable format that could handle this (variable size elements in a table list), but it may be more likely that we would do this with the "wave sequencer" object planned for a future update. In this case, Shape would control the array element selected, pretty much like it works for the Waldorf stuff now.

The Random (noise) approach I have had in the to-do list as a separate value in the Mod Source list that is generated for every note on event, but we may instead implement it by using an LFO in the S/H waveshape. But not in the first OS release now!

Cheers,
John

John Bowen
Site Admin
Posts: 1815
Joined: Sat Apr 07, 2007 3:00 am

Post by John Bowen » Fri Oct 08, 2010 8:50 pm

This is in response to a part of the video where I show how we currently can use multisamples. It requires using one oscillator per sample, so technically up to 4 samples from a multisample set can be "played" across the keyboard, with the mapping done in the text file that describes each sample.

I might note that, as you can have multiple of these text files, with each one representing a sample group to be loaded, you could in fact be using the same samples, but with different mappings.
-john b.

Post Reply