OMega 3 Workstation YouTube tutorial

Thanks for these new/old versions!
The CPU is not an issue for me on the current SF I have been playing with. So I'm not sure if I need to look backward. But I will check out the blocks you just uploaded.
BTW I have no use for Racks either. That is clearly a 'marketing' strategy to get free player users to transition to full Reaktor ownership.

I always open up in 'New Ensemble' mode. I have loaded in some NI Racks, and I immediately convert them to ensembles.
But you may be onto something about 'Racks-orientation' in 6.4 that may have ramifications for my event-loop issues.

It took me a while to figure out that in my 'plucky' ensemble I originally wired up SF on the wrong 'node'.
The ensemble has a branch that leads into the NI Driver block, as I was thinking about adding a little lightweight distortion for the original version of the ensemble as a way to excite the tone, which tends at times toward the resonant side, using the NI Paul filter in a polyphonic version I hacked together.
Well, SF is quite a 'hot' instrument -- it puts out a lot of signal if it receives one in kind, so using the Driver branch into SF sent more distortion into my mixer, for the SF mixer inputs (it has its own mixer slot, along with the un-Driven output and another for the Driver output. The mixer's own outputs go into Rounds Reverb for everything).
After a couple of hours I realized (duh!) what I wanted was that SF should come from the clean input, and only as much Driver-tinged signal as mixed into a snapshot should carry distortion forward. Good thing I did, finally, 'cause I had quite a few snapshots already made.
Bottom line is I'll wait till tomorrow to send some audio...
But having a great time with your 'Delay' instrument.
(I'll have questions for you, as the control settings sometimes surprise me in how the bands wind up looking...)

 

As for the 'hotness' - there really isn't any gain per se in SF. The band filters are 0 db linear phase butterworth.
If you pass a 0 db sine wave into it you will get a 0 db sine out no matter how many band stages are on.
I was very careful to make sure that it was linear, but I haven't looked into that in a while.

What can happen, though, is that with complex signals going through band stages with varying delays per band, things get tricky.

If the stages have moderate time delays - say 250 ms or so - and if the parameters are changing every few hundred milliseconds or so, then when the signal leaves stage one and goes into stage 2 it will see very different band delays. After 5 stages it could get incomprehensible.
And there could be pitch shifts if the delays move about (which is why I think you want a granular delay line).

After each stage, the 11 audio bands are re-combined before moving on to the next stage where they get split again. So the pitch shifting might expose the next stage to very different band delays by the time the signal gets there.

And then, if the feedback is turned up, it could easily self-oscillate anywhere.

The end result could be some high gains in certain frequencies, but the total spectral energy should be conserved if feedback is 0.

For the Block devices I just posted, they will be guaranteed to never output a level outside [0..1] because of the saturator at the output.

But I didn't want that coloration in OMega 3 so I removed it. Technically, it isn't a block anymore.

The 11 band delay control setting being a surprise? I'm not sure what that means.

There are 128 snapshots. The first bunch are fairly tame, I think. It's been a while since I looked at it.

But later ones are quite fanciful. I went to a web site called OEIS:

The On-Line Encyclopedia of Integer Sequences
https://oeis.org/

and picked out some integer sequences for the controls. It was a random choice of sequences.

I rigged up some sort of test ensemble when doing that. I just found it so I post it here.

 

Well, it's early now, and I haven't turned anything on yet, but I wanted to respond to the confusion i created about what happens when I/we look at stuff that is related to what we're listening to.
I meant only that I could never predict what the 11-band display would look like based on all the different control settings.

Of course I realize that there are so many things going on in SF that my ability to comprehend them and make 'predictions' about what the resulting visual aids might look like is preposterous. At least at this point of my education!

 

Oh, those controls. I thought you were referring to the 11 band delay block, not SF.

I have no idea what the SF will sound like either, except in very simple cases.

I just try something with very small changes to a small number of parameters and listen.

It helps to start with a simple impulse, like a short snare hit or a pluck so you can hear the delay and in other cases use a long sustain to hear the pitch and timbre changes. Feedback is the last thing I add.

It is experimental by nature and design.

There was a time when I tried to put in a granular delay but I gave up on that for some reason that I no longer remember.

Short delay times sound one way and very long delays sound another and things in between sound again another way.

For the short delays, aligning the I/O time may be desirable to sync the attacks with the input mix. For really long delays and high stage count I take advantage of the cascade to pulverize the sound.

Of course, the smoother setting is crucial to remove artifacts and undesirable changes.

Some of the parameters change nature going from + to -, you can see that clearly in the desmos graphs. I think it is n1 and n3. Sometimes at 0 the thing glitches because of the 32 bit power function operating in the 64-bit world, but there may be other reasons. If this affects you a lot, maybe you can look at the parameter limits and make some changes there in the Core macros. I think you are more comfortable there than I am.

Perhaps because we are so used to hearing sounds hit our ear at the same time in frequency bands, that when the time bands arrive changed we are not used to that. Just as the Haas effect can help us locate things in space, the arrival of the spectrum in phase helps us realize timbre.

The simplest example is dispersion of sound through metals. I immediately identify that as a blast - a space gun or metal strike or something like that.

But when the dispersion is more complex, it does not occur in nature and we have no names for that.

An optic lens can be made using curved glasses of different diffraction indices stacked in cascade. The same thing is happening acoustically in the SF banks.

If your sound is already dispersed, you may be able to dial in the inverse dispersion to re-align and refocus the sound. That would take some effort but may sound rather interesting.

 

Another experiment I thought about performing was to increase the bands by making them 1/3 octave regions. That is easy to do - increase the Voice count from 11 to 33 and change the band frequency table array values with new mid-point settings.

And by extension there could be a continuous function found where the SF is 'convolved' in a sense with the input signal. That would remove the concept of bands entirely. But the math to do this is beyond me. Some sort of all-pass filter with complex coefficients...

 

No, the voice count would be changed from 55 to 165. I forgot to multiply by stages. Maximum voice count is 1024.

 

All interesting and intelligible comments for me.
I am taking SF strictly as 'experimental', which you may gather for me is right up my street.
I tweak and listen, often struck especially by what happens to the sound at phrase endings, when the input stops but the SF output may still have some spectral comments to add.
The unexpected nature of the output is tantalizing, and frankly I don't try too hard to understand how it gets the way it is.
If it's juicy, and predictable enough that it will sound that way again under similar conditions, I want to save it.
To demonstrate crudely the crude comment about 'hot' signal, I made a relatively inept video of my mixer while playing back my ensemble.
See below Mixer Slot 3 'Hot'.mp4.
Mixer input 1 is the 'clean' sound (plenty of lively motion in the sound, which I'll follow up with audio soon); Mixer input 2 is the Driver distortion (quite low settings are enough for this); Mixer input 3 is SF.
You can see that with relatively small increases in the input on Mix 1 sometimes SF on Mix 3 gets very excited! So the level set there is also quite low on the mixer.
What's interesting to my ears is that both the Driver and SF add just enough magic that the liveliness of the sound is further enhanced.
Which as I say I will send a clip of audio soon, just for 'grins'.

 

I see

 

Yes, not all that illuminating.
So I include a short sound clip now, also perhaps not very illuminating.
Trust me though that SF is adding lots of lovely color here, though I don't provide a 'B' version with SF muted.
You can catch SF's contribution especially after explosive things happen and the sound lingers with spectral dust hovering about.
Also at certain points where the relatively hot input gives rise to gushing from your experimental instrument.
Imagine you have a guitar player with a penchant for playing with pedals and knobs and switching pickups at a whim...

 

That type of sound is right up my alley as we say here.

I like it and I can clearly hear what part is modified.

With short sounds like that the effect is reproducible and more predictable with key-sync'ed LFOs and envelope modulations.

 

Since you say you don't have CPU issues, you could make two SF blocks one for left and the other for right. Use the same preset for each and just have very slight modulation changes for one track that is not in the other.

 

Aha, glad you like it!
I think the addition of SF has added a dimension of subtlety to this (and sure to be others) ensemble that I wished I could have produced myself.
I'm really revved up to have figured out a way to use SF to -- not polish but -- sculpt the sound in gently persuasive ways that take it/us away from the earthbound and inherent limitations of 'digital'.

 

Nice idea. Probably lush phasing effects would emerge? Will see about this at a later point in my education.

 

Needless to say, I'm pleased with the decision to take Omega 3 apart to learn how it works from bottom up rather than top down.
If you were to choose another likely component, what would you think would offer relatively gentle learning curve and sonic bang for buck?

 

What to change next?

That is totally up to you and the type of sound you want to make, so it is hard to answer.

I will give some general ideas:

1) Take out the instruments you don't want in the ensemble and replace them with the outboard effects you like. You could put in a small mixer and add effects in channels. Use panelsets to manage the screen real estate.

2) If you just want to change the OMegas 3 synth and not the rest, that is not hard. Macros are very easy to add on the synth, filter, insert and effects busses. Or replace ones that you don't need. There are selector list names and numbers to update, and the added macro may need to be modified to fit inside the stack macro space. Adding a new stacked macro to the 10 or so already there is fine as long as you add the macro just once and not change it because the order matters. So you will want to add a macro container - work inside that container only. Or change an older stacked macro being careful not to deleted its container - the thing that reserves its spot in the stacked macro index. I think there may be a way to change the order of things in a stacked macro, but I don't remember how to do that. Stacked macros need two small panel items in two corners, say top left and bottom right, (I call then TL and BR), to anchor the gui. I use single pixel text boxes or transparent images for that purpose.

3) If you want to add a Block, that is a bit more difficult because the Block is an instrument, so it requires its own screen area. But still, it is easy to add if you look at how the SF bank was added to the inserts for effects. I think most of the synth Blocks are monophonic? So that may be an issue. They could be added as an external input to the synth like the way the samplers were added. Be careful to check all the connections on the instrument properties so they don't interfere. Connect the MTS sequencer to the MIDI IN. The mod matrix is inside the synth, so it won't be available to external blocks/instruments unless you route out the 'a through 'x' ports or use some sort of send/receive connector.

4) You might want a different MIDI note sequencer with more steps and different ways to activate - more deterministic or more chaotic.

*) I chose to put SF as an effect insert because that's the last thing in the synth audio chain and it was most likely always going to be the last effect one would add. It could have been added in the filter insert if one wants to process its sound a lot. I did have it in stereo once but could not really fit it into the GUI, so I reluctantly made it mono.

**) You might want to change the VOL structure with a different envelope if the two there do not suit your style. Also the Control Pitch could be changed to suit your playing style. You can increase the voices count from 6 and add more playing modes. Right now there is polyphonic, monophonic and legato modes. The multi-stage envelope was made a long time ago and there may be much better ones available. I find it impossible to follow so I never change it. It does crash sometimes.

5) You may want different or more controllers - different basic LFOs and Envelopes, etc.

6) One thing I thought about was some sort of envelope follower that could be routed from some bus for a control signal and then assigned an input send and receive. That could be good for vocoder type effects and stuff like that - a poor man's convolver.

7) Reaktor has no convolution engines. Maybe some people have put in something in the U/L, I don't follow the U/L very often.

8) Inside OMega 3 Synth is a macro called Colors which has suitable, compatible background images in 9 brightness values for your macros so they fit in.

9) I have a foot pedal and use the mod wheel on an external synth, so I put those in the mod matrix for filters and effects. They are labelled 'fc' and mw'. You could adapt that to what you have. It would be a major effort to add more modulations - that would require probably hundreds of wiring changes. I would not want to do that. Every oscillator, insert and effect will need more ports added and the selector lists will need to be updated.

I think you will find every important macro and i/o pin to be documented in all the oscillators and effects, in general. I did not document something I did not understand (like the multi-stage envelope, for example).

I tried to make all the macro wiring styles consistent in screen layout and spread out and easy (for me) to read and follow. Some might not fit on an HD screen size.

 

I might say a word about the philosophy behind OMega. It came out before Blocks, but it has a different approach. I grew up with patch cords and have no intent to return to that mess, so the way I see people make patches these days is not something I would emulate.

I wanted something I could open and manage in one spot, so dragging blocks around was not available and not needed for my purpose.

I also did not want to use a DAW for managing connections with multiple Reaktor instances. I always had a powerful processor so it was never necessary and I didn't want to add another load if I didn't need it.

I also did not know exactly what pieces I wanted, so I started with a very general and haphazard assortment. I actually never changed the original assortment at all, just added the SF rather later.

So I came up with the workstation approach and tried to keep it self-contained.

There are as many valid ways to do this sort of thing as there are people.

You may notice that there really are not much in the way of Drum or Beat-making tools, nor vocal processors. I don't do that, so those who do may want to make radical changes or it may just be too different for their needs.

 

Wow, great tip! I did finally after many many months get the hang of stacked macros. Sad to say the documentation on them by NI is sketchy. Old hands at these things should (I think) contribute to the documentation in the way that Brett Lavallee helped people out with blocks' gui.

 

I want to be able to use Omega 3 for what I have heard it do, but getting to know how everything is connected is harder for someone who didn't build the ensemble, of course.
I think I will just try some of the instruments in lightweight settings to get a feel for those that can be simply extracted.

Your philosophy doesn't sound all that different from mine. I grew up on patch cords too, before there even were self-contained synthesizers. My initial experiences were via Columbia-Princeton style units wired into a patch bay (at SUNY Stony Brook with Bulent Arel, early pioneer of the tape-splicing mode of electronic music composition).
In any case, spaghetti-wire in front of already small visual widgets is nothing but difficult for my eyes.
And while I love percussion sounds, I eschew 'drummers'. No 'beats' for me, thanks. Let's not stain the plate.

I have accustomed myself to ensemble-building using blocks. It is ultimately a messy world we make this way, and yours is nicely self-contained.
I have tons of blocks-ensembles now, many of them very useable, but some of them redundant of course. Now, do i want to spend the time figuring out which ones to keep and which to unload? And what about hundreds and hundreds of nice snapshots, that also become redundant at a certain remove. So I understand your workstation approach. It allows you to get at music-making easily, and of course that really should be the main point of all these endeavors!

 

I made an ensemble called ToyChest that I used to startup a stacked macro project.
Here is a copy:

 

Here is a complete set of color backgrounds that I made available in OMega v1.0 to those who wanted to change the color scheme: