CAKEWALK RAPTURE (software instrument)

[Anderton]

Sure! The LFO, step sequencer, and EG are all in "parallel" -- think of them as being mixed together. This was really brought home to me in the wavesequencing example given earlier, where the EG provided a "master envelope" for the wave sequencing function being done by the Step Generator. It's actually very cool that it works this way.



Yes, exactly. It also means that you can use the six Elements to create a "multisampled" patch. Let me give an example.

I have a folder with five Minimoog sawtooth waves, each an octave apart. Suppose I want to make a really "fat" sound with lots of layering. There are two ways I can do this:

1. Create an SFZ file with all five waves, and load it into an Element. Then, duplicate the Element multiple times (up to six), with slight detunings and such.

2. Load each wave file into its own element, and use the key number option to limit each file to a specific range. Then I can use the Multi feature in each element to multiply the number of voices.

Why use one over the other? Well, in option (1), you have a maximum of six layers. In option (2), you can multi a waveform to create up to nine voices (layers). However, (1) is less work to program, especially because there's no Group Edit option. Does this make sense?

>

That sets keytracking for cutoff, resonance, pan, etc. (it also works with pitch, but the oscillator Keytrack parameter allows for more precise adjustments). I've used it to advantage on a tuned white noise patch to tune the filter. Unfortunately it's poorly calibrated, it's more of a qualitative than quantitative parameter, but I was still able to keytrack the filter precisely to create the tuned white noise effect.

>

They aren't even REMOTELY dumb. A dumb question would be "Is the DVD-ROM with the program edible?" or "Do you have to replace the six strings when they wear out?"

Okay, to my left on my music computer, Rapture is set up and ready to go. Let's see what it reveals tonight

[Anderton]

Hey, maybe some Rapture expert can help me out here...

I wanted to control the volume of an element from full on to full off with the mod wheel. I used the modulation matrix to assign Mod Wheel to Volume. I figured that this would override any envelope settings and control the Element volume parameter...nope.

To make a long story short, I was able to get the effect I wanted by doing the following:

* Turning the Amp Envelope Generator to ON
* Creating a full-on envelope with maximum level sustain
* Set Depth, Vel >Tim, and VelTrack to 0.0
* Set Vel > Int to 0.1. If I set it to 0.0, it didn't work right.

What am I missing? There must be a simpler way to just control an overall Element output level with the mod wheel. (In case anyone wonders, I'm trying to create a "theremin" patch where the mod wheel brings the sound in and out).

[rasputin1963]

Many thanks, Craig.

So am I right in thinking that each element can only handle one wave and one wave only? Where is the "Duplicate element" command located?

The MULTI function... is it a chorus? (ie., does it partake of the delay/phase stuff that a chorus does?

Does RAPTURE have a window for viewing the actual contour of your 3 EQ settings?

Can RAPTURE users share their custom-made patches (with custom waves) with other RAPTURE users? I'm guessing RAPTURE doesn't save the wave along with the patch, so you'd have to send your recipient both the RAPTURE patch and all relevant waves, right? Into which Windows folder would the recipient place the WAV's so RAPTURE could automatically find them?

The overall picture I'm getting of RAPTURE is that it's rather like a SoundFont... on steroids, no?

I'm also thinking that, with all RAPTURE's temporal movement, the ambient composer could almost lay down the theme of a whole song-- just with one patch alone!!

[Anderton]

 

 

One wave, or an SFZ set of multisampled waves.

 

 

 

You right-click on an Element number, select copy, then right-click on another Element and select paste.

 

 

It's not a signal processor, it actually creates more voices and detunes/spreads them. This does result in the same kind of sound as chorusing, just like detuning two oscillators slightly sounds like flanging, but it is an actual "chorus" of waves rather than a wave being chorused.

 

>

 

There's no frequency response graph, if that's what you mean.

 

 

 

Yes.

 

 

 

In the Rapture folder.

 

 

 

Well...if so, those steroids are mega-steroids, as you'll see in some of the future audio examples.

 

>

 

You are definitely right about that!

[Anderton]

This is a program where the LFOs sure do a lot more than just add vibrato. As with the other modulators, Pitch, Cutoff 1, Resonance 1, Cutoff 2, Resonance 2, Pan, and Amplitude each have their own LFO.

 

Let's look at the various parameters. Click on the attachment to see the part of the screen that relates to the LFO.

 

* Status can be on or off.

* Frequency is in play only if you turn off sync to host; it ranges from 0.01 to 40Hz.

* Sync goes from 1/8t to 128d. Each cycle equals the selected beat value. For example, if you select 1/8, there's a cycle every eighth of a beat.

* Delay determines how long it takes before the LFO kicks in, up to 10 seconds.

* Fade fades in the LFO over up to 10 seconds. You can combine Delay and Fade, so there's a certain amount of delay before the LFO fades in.

* Depth just sets the overall depth of the LFO.

 

You'll also note the waveform display. 28 waveforms come with Rapture, (26 periodic and two random) but what's really cool is you can create your own waveforms out of any WAV file; there can be a total of 100 waveforms. I did a half-wave rectified sine LFO waveform just for kicks.

 

But this is important for another reason: There's no smoothing option for the LFO, so anything like a square or sawtooth wave has a very sharp transition that might cause a click when controlling a parameter. So, you can create a waveform that's smoothed and use that. I'd rather have a smoothing control, but hey, maybe next version...

[Anderton]

Just to show what you can do with the LFO, I've prepared a little audio example. Click on the attachment to hear the LFOs at work.

Here's what's happening. An Element has a multisampled Moog sawtooth SFZ file loaded. All the modulation is done with LFOs, specifically:

* One LFO is set to a negative-going sawtooth and feeds the Filter 1 cutoff, with sync set to 1/4 (1/4 of a beat, or a sixteenth note). This produces a series of quick filter frequency decays.
* Another LFO is set the same way, but feeds the Amp. So, you have what sounds like a series of 16th notes with sharp amplitude and filter envelope decays.
* Another LFO is set for a slow sweep (every 16 beats) and varies Filter 2's cutoff.

But a sound is worth a thousand words, so check out the example.

[lawapa]

Great example of LFO's at work CA.

You can have a pitch sweep/step workin as the lfo's do there thing as well. Almost to many options in Rapture.

[Anderton]

>

The next example is going to be a fun step sequencer thing that uses only the step sequencer to accomplish the desired results.

I wouldn't say there are too many options! You don't have to use all of them But the more you know about the program, the better. For example, I tried doing the last example with the step sequencer and realized it would be a whole lot easier just to use the LFOs.

[lawapa]

I personally use step when I'm reaching for a discrete change to happen with smoothing to soften the abrupt change to something with a softer edge. Of course any change can be drawn in.

[Anderton]

>

Yes, I think the smoothing function is incredibly useful! I wish the LFOs had it too. The easiest way to create a "smooth" waveform is to take something like a sawtooth, run it through a lowpass filter, then save it as a custom LFO waveform.

[René [Cakewalk]]

Hello Craig,

Wow, that's in-depth. Thanks for the great review and comments, superb as always.

What am I missing? There must be a simpler way to just control an overall Element output level with the mod wheel. (In case anyone wonders, I'm trying to create a "theremin" patch where the mod wheel brings the sound in and out).

In the modulation matrix, we have one source which is fixed. This one is set to 1 (or 127 in MIDI terms). We have set CC127 to do that.

There are many ways you can 'connect' the modwheel to element volume. If, for instance, you'd like to add 6dB to the element volume using modwheel, you'd do:

CC1 -> Volume 1 -> 6

Those are Source, Destination and Depth respectively. Now, if we'd want to decrease the volume by 6dB, then we'd do:

CC1 -> Volume 1 -> -6

In case we'd like to increase the volume from a fixed, quiet point, we could just turn the element volume down using the element knob (not recommended, as you lose the ability to set a mix level after that), or use the CC127 to set a 'fixed' level of volume down, and then specify the modwheel to make it increase with a positive value. As follows (two rows in the matrix):

CC127 -> Volume 1 -> -36
CC1 -> Volume 1 ->36

That way, the element volume knob remains in 'allow-to-mix' status, and you get a 36dB boost from a quiet status using modwheel.

The CC127 trick is used in several patches, for many different uses, all related to adding a fixed amount of 'something', or to reverse a controller effect.

I would like to drop two comments here: first, the modulation matrix rows can 'stack'. This means that destinations can be affected by multiple sources, and same source can affect multiple destinations. This is also valid for MIDI Learn: you can learn multiple controls in the same control, or vice-versa. Second, the smooth control allows you to link one source-destination pair, and assign a smoothing factor. It is notable that smoothing isn't only used to avoid zippering noise during parameter transition, it can have an enormous creative potential for performance.


-René

[rasputin1963]

Hello Craig,


I can't get over how crystalline-clear is the sound quality of these MP3 snippets...

Is this the "anti-aliasing" that has been referred to? It certainly sounds different from other soft-synths I've heard. The patches sound fat-fat-fat, but with no unpleasant artifacts.

Can you speak more about how this ultra-clarity is achieved?

[lawapa]

I'm not craig but I have both Rapture and Dimension Pro and yes both of these beauties have that type of sound engine. You have to import noise to use in you patches.

[Anderton]

Is this the "anti-aliasing" that has been referred to? It certainly sounds different from other soft-synths I've heard. The patches sound fat-fat-fat, but with no unpleasant artifacts.

Can you speak more about how this ultra-clarity is achieved?>>

I think Rene could do a better job of that

Part of it is also that I know how to create MP3s that sound decent, which is a whole other issue. BUT the important point is that an MP3 can't make something sound BETTER than it is. So, although you're not hearing Rapture in its 100% native quality level, the fact that it makes it through the "MP3 gauntlet" sounding this good tells you something right there.

Wait'll you hear the next audio example

[Anderton]

Rene, since we have your attention...someone asked about the differences between the z3ta+ and Rapture. Given that you designed them both, it would be great if you could give your perspective on how they differ.

[Anderton]

This is arguably the crown jewel of Rapture's modulators, although the Envelope Generator is pretty hot too. Anyway, every parameter you can modulate with LFO can be modulated with step sequencer, and again, this is true for each element so you have 42 total step sequencers.

 

Click on the attachment to see three step sequencers. Through the miracle of paint programs with cut and paste, I've captured three step sequencer settings for reasons that will become clearer next post, and pasted the Element label on to each one.

 

* Status can be on or off.

* Steps range from 2 to 128. The higher settings let you "draw" fairly detailed curves.

* Frequency is in play only if you turn off sync to host; like the LFO, it ranges from 0.01 to 40Hz.

* Sync goes from 1/8t to 128d. As with the LFO, each cycle equals the selected beat value. For example, if you select 1/8, there's a cycle every eighth of a beat.

* Smooth rounds out the step transitions, giving more of a continuous waveform.

* Depth determines how much the step sequencer will modulate the assigned parameter.

 

This is all pretty straightforward; step sequencing lends itself to sample-and-hold effects, complex varying curves, and the like. But don't forget that each Element also allows for other processing...

[Anderton]

Click on the attachment to hear the step sequencers having a good time. Here's a description of how the patch was constructed.

 

I loaded a Minimoog SFZ multisampled sawtooth wave into Element 1, a similarly constructed Minimoog square wave into Element 2, and a Minimoog thin pulse wave into Element 3. (By the way, if you're searching your Rapture folder looking for the Minimoog waveforms, these are custom waveforms I created and they don't ship with Rapture.) Each Element uses one step sequencer applied to Amp.

 

Element 1 is the main element; it's panned to center, and the Step Sequencer provides a nice rhythm. Additional LFO negative-going sawtooth modulation is assigned to Cutoff 1 and Amp to give the quick decay featured in the LFO example given earlier.

 

Element 3 is panned right, and has a complementary rhythm. It similarly applies LFO to Cutoff 1 and Amp. Element 1 and 3 played together give a sort of "AdrenaLinn" vibe.

 

But the interesting part here is Element 2, which is panned right and has a relatively sparse sequence. It goes through the Large Hall reverb insert effect so when it triggers a sound, there's a reverb tail. This is a pretty cool effect that's representative of the type of creative step sequencing effects you can add in Rapture.

[René [Cakewalk]]

Rene, since we have your attention...someone asked about the differences between the z3ta+ and Rapture. Given that you designed them both, it would be great if you could give your perspective on how they differ.

Well, there're really many. To start, with Rapture we wanted an instrument you could use in any studio, regardless of what platform you feel more comfortable with. Therefore, it's VSTiDXi/DXi64/RTAS on the PC, and VSTi/AU/RTAS/UB on the Mac. z3ta+ is VST/DXi on PC only, and a Mac port might be difficult.

From an architecture point of view, all the sound designers who worked with z3ta+ agreed in a major feature request: they wanted different oscillators to be routed thru different filter types, and processed with different effects.

That wasn't really possible in z3ta+ without rewriting it from scratch, it's a complete architectural change, with busses and all. Rapture delivers that: any of the 6 elements (which could be compared with z3ta+'s 6 oscillators) allow to process each oscillator thru a whole dsp chain.

z3ta+ is still famous for the fat trance pads and leads obtained using the MULTI mode, and some detuning. Rapture expands that by allowing a selectable multi-voice multi mode, which is spreaded in the stereo field. This, and the ability to process multisamples in each element, requires a full stereo path in each element. Also from a z3ta+ user request comes the ability to change the dsp chain order, the graphical envelopes and the graphical step sequencers.

In a parallel path, we had the sfz/sfz+ family of players, using a novel technique to playback samples. We expanded the parser capabilities, enhanced the performance, and once the new engine was ready we merged it with an also expanded version of the z3ta+ wavetable-playback engine. That is the Expression Engine, which is in use in both Dimension Pro and Rapture. We oriented the straight sample-playback capabilities to Dimension Pro, and packed it with tons of contents, and the wavetable-playback capabilities to Rapture.

It is important to mention that each element in Rapture allows loading an sfz file. Any sfz file can contain a complete multisample definition, including an arbitrary number of keyboard regions, velocity regions, regions selected by the position of other things like random, sequence or MIDI CC, etc. Each region can not only play a different sample, but also can play a full wavetable oscillator.

This means that you can stack multiple oscillators using different pitchs, waveforms and others with a little sfz magic. For most uses, the interface will provide the required power, but the adventurous could really get much more mileage by doing little sfz experiments.

The other impressive achievement with Rapture is the resources efficiency. It can process the same dsp load as z3ta+ with about a half of the resources.

z3ta+ still has some unique features (waveshaper), and that will remain that way for now. It's still a tremendously successful product, with a bitrigazillion of patches and banks in the web. There's no reason to avoid people having both z3ta+ and Rapture


-René

[Anderton]

This was originally going to be about the Envelope Generators, but I found a workaround for an EG limitation that I thought you'd find interesting.

Here's the problem: There's no way I can find (maybe Rene has a solution?) to assign mod wheel to envelope decay. You can assign it to release (which isn't equivalent for what I want to do), and you can assign velocity to envelope decay; but what I wanted to do was trigger the envelope generator, and be able to vary the decay time expressively.

After not being able to come up with a solution, I thought about why I wanted to vary decay in the first place: It was so I could have a series of four notes, with decays of short-short-LONG-short to accent the backbeat.

Then it occurred to me to use the Step Generator. Click on the attachment to see the control signal that provides "dynamic decay." I set the Step Sequencer to 16 steps, and created a sharp decay that lasted for 7 steps. Then I increased the Step Sequencer length to 32 steps, and created an even sharper decay with a slightly lower amplitude in the next group of 16 steps. Then I set the Step Sequencer to 48 steps, and for the third group of 16 steps, drew a decay that covered all 16 steps (the long decay). Finally, I extended the step sequencer to 64 steps, and drew in one final short decay over the beginning of the last group of 16 steps. This envelope got applied to the Cutoff 1 and Amp envelopes to create a percussive, rhythmic effect.

Keep reading…

[Anderton]

Now it was time to fine-tune the effect. Setting Sync to 2 meant that the step sequence was 2 beats long, so each "decay" was an eighth note (four in two beats).

So far so good, but I wanted to add more expressiveness and some motion. So, I copied the Element to another Element, and added a little detuning. Then I used the Mod Wheel to control the Filter 2 cutoff for both elements, so there could be a little "filter-opening-and-closing action" in the patch.

Finally, I went to the Global FX page and used stereo delay to add motion. The left channel was set to 1/2d delay, and the right channel to 1d. Although this wasn't quite the same as being able to vary decay via mod wheel, which was what I had originally set out to do, I was still able to get something close to a dynamic decay effect. Click on the attachment to hear the patch do its rhythmic thing.

[Anderton]

It was asked about why the audio engine quality was able to make through the process of being converted into an MP3. Cakewalk provided me with results from a Dimension Pro Expression Engine Aliasing test, which also applies to the engine in Rapture.

In this example, Cakewalk's engineers loaded a 15kHz sine wave into Dimension Pro and played one octave up (C5-B5), then one octave down (B4-C4). The session was rendered at a 44.1kHz sampling rate at 32-bit floating point depth.

The sample was converted into mono, and imported into Adobe Audition’s spectral frequency view. When viewing the spectral analysis, the different colors represent degrees of amplitude. Yellow is the loudest, followed by orange, red, magenta, violet and blue.

The first picture is of Dimension Pro's engine without the sinc interpolation option enabled. Click on the attachment to see Dimension Pro without sinc interpolation. All of the extra lines you see are small bits of aliased frequencies.

The next post shows the results with sinc interpolation enabled.

[Anderton]

Click on the attachment to see the results with sinc interpolation enabled. It's pretty impressive: You can hardly see any aliased frequencies at all. This is in line with my own experience when I first tried Rapture -- I sent an email to Cakewalk saying that there was definitely aliasing in their "alias free engine." They asked if I'd enabled sinc interpolation. Ooops (read the manual!). When I did, any aliasing was no longer audible.

[lawapa]

That ability to enable and disable sync interpolation/distortion is mighty cool. If you want it you got it and if you don't it's gone You can also load a wavetable in with sfz mapping. With oscillator=on it loads up clean. or you could load it as a wavefile direct and get thet nasty distortion. What ever turns your crank

[Anderton]

The following is a true story from several years ago at the Frankfurt trade show...

Waldorf designed a new synth and they were very proud because they'd been able to get rid of aliasing. Only problem was a lot of their customers LIKED the aliasing sound of the previous model, so they put in a switch where people could choose one or the other!!!

Regarding Rapture, I'd say just leave Sinc enabled. There are lots of other ways to get nasty sounds, far nastier than just disabling sinc.

[default_damage]

Craig

thanks for all the helpful information .
I have rature on my computer now.
I got it at a local music chain store priced , the street was $199.99.

loading it into my PC was a breeze, the registration and recieving the recg. code took almost but not quite a minute.


Using a P4 2.8 hyperthreading 800fsb,
Rapture in Sonar loaded and ran four instances @ 21% cpu.

the Sound ? I liked what I heard . I found it to be very clear and detailed.
I then added the PSP vintage warmer to each instance and after that , it was true love
(hi, my name is default and I am a Vintage Warmer addict)
thanks again
tah-tah fo' nah-nah.