emusonacid! legacy

Hi guys,

is there a possibility to have a squarewave LFO with its pulsewidth modulated? Trying to wrap my head around it, but can't seem to find a good solution...


cheers!

Yes, but it has some limitations and needs some efforts.
- The LFO-shape will always start with the positive phase of the square-wave.
- The pulse-width will be in the range of 12% to 100% as the best LFO-shape for this is the 12%-pulse.
- You need a sequencer/DAW for sending MIDI CC-messages (preferring one of the 3 Switch Controls: CC64, CC65, CC66) for re-triggering the LFO.
- The PWM is dependent from the LFO-speed (hz) and the song-tempo (BPM).

You have to set up a table (Excel) for the calculation of the LFO re-trigger timing.
You also have to think about the MIDI-timing: 31250 Hz transfer rate ... that's about 0.256 ms per byte
As a CC-message consists of 3 bytes, a complete CC-message needs about 0.8 ms to reach the sampler.
... and the re-trigger events are not the only MIDI-messages that you send to the sampler.
... processing will also take some time, too.

The time the LFO needs for running through a single cylce is: 1/frequency [-> ONE divided by the LFO-speed in hz]
A) LFO-speed: 0.08 hz (minimum LFO-speed)
-> Required time for a single-cycle: 12.5 seconds
B) LFO-speed: 18.01 hz (maximum LFO-speed)
-> Required time for a single-cycle: 0.055 seconds (55ms)

This means that in case A) the LFO restarts after 12.5 seconds while in case B) the LFO restarts after 55ms.
By sending a the re-trigger event the LFOs will restart earlier causing a variation of the pulse-width.

But as you have to use a sequencer/DAW the question is: For what do you need this?
When you need the LFO for re-triggering something of a voice, you can also re-trigger directly without using the LFO.
The MIDI CC-message will re-trigger the same destination of all played voices where the cord is setup accordingly.

Cheers!

Info here
http://www.emusonacid.co.uk/forum/viewtopic.php?f=1&t=3633&p=15412&hilit=Toolbox#p15412
Download PWM Thing

JAHFUNK wrote:Info here
http://www.emusonacid.co.uk/forum/viewtopic.php?f=1&t=3633&p=15412&hilit=Toolbox#p15412
Download PWM Thing

that doesn't cover lfo pwm, just pwm of a sample/square wave osc? I'm referring to PWM of the LFo which runs a squarewave...

@mosrob: Thanks for the huge effort, didn't understand a thing but I will read it over and over.

actually it's not important to be precise, the timing may be off. What I wanted to achieve with lfo pwm was a stutter effect. the rate of the lfo will control the intervals at which the sample will be played (a thing I wanted to modulate) and the PWM controls the duration of the sample. Maybe I will try someting with sampleloop/samplestart modulation and with an one-note arpeggiator, this might be easier aswell as syncable... phew!

ja-ki wrote:What I wanted to achieve with lfo pwm was a stutter effect. the rate of the lfo will control the intervals at which the sample will be played (a thing I wanted to modulate) and the PWM controls the duration of the sample.

Could you give me an example of what this sounds like in a tune?

From your discription it sounds like what you want is to ...
Re-trigger a sample with a square LFO and have the LFO rate on a C/C, with a second C/C assigned to S/start - This would create a variable rate stutter with an adjustable start point.
You also need to set Amp Decay to a controller to get good control of the duration of the sample length.
Is this correct?

VST-plugin Stutter Edit from iZotope:
http://www.izotope.com/products/audio/stutteredit/
As you can hear in the examples, can the stutter effect be much more than a re-triggered sample with a variable start-point.

Back in the days people used their DAW, recorded a phrase (->sample), cut it into several chunks of a specific size, and re-arranged these chunks and copies of these chunks on the timeline.

You can also take a phrase, create several samples with different start-point out of the phrase, assign these samples to keys in a preset and trigger the keys manually or from a sequencer/DAW.
In addition to this you can create up to 127 layers per key with different settings for filter, pitch change (up, down), fx-amount for the internal FX or output assignment to different external multi-FX, sample re-trigger,etc. and cross-switch between these layers via velocity (-> Vces-VelWin).

JAHFUNK wrote:

ja-ki wrote:What I wanted to achieve with lfo pwm was a stutter effect. the rate of the lfo will control the intervals at which the sample will be played (a thing I wanted to modulate) and the PWM controls the duration of the sample.

Could you give me an example of what this sounds like in a tune?

From your discription it sounds like what you want is to ...
Re-trigger a sample with a square LFO and have the LFO rate on a C/C, with a second C/C assigned to S/start - This would create a variable rate stutter with an adjustable start point.
You also need to set Amp Decay to a controller to get good control of the duration of the sample length.
Is this correct?

I was just sitting in front of my computer, reading your post scratching my chin and came to the point that your method should lead to the same effect!
I'm sorry I have no examples because I just wanted to try it, sounddesign fun.
Wondering what cord I have to set up to send the LFO to retrigger the sample, I'm sure there's a destination like that *firing up the machine*

Re-triggering a sample????
It's mentioned in the manual... on the last page of chapter 8 "Preset Edit", next to the cover page of chapter 9.
Source Destination Amount
"LFO~ SRetrig -1%"

I have another one, regarding LFO PWM again:

let's assume I take LFO 1 with a squarewave which will retrigger the sample as soon as it's cycle starts again (btw, thanks mosrob, didn't knew there was actually something useful in that stupid manual! )
It also controls the length of the sample while on the +1 phase, I'd cord that up to the AMP somehow
this would at first yield in a retriggered sample, always as long as the +1 phase of LFO1, so higher rate means faster retrigger aswell as shorter sampleplayback.
BUT, If I turn down the rate of LFO1 while in the +1 phase, I could get PWM when I'd use LFO2 with a saw controlling LFO1's rate. Is this correct?

The only problem I see with this is that LFO2 has to have the same starting position as LFO1... and I'm not sure if this is possible.
cheers

edit: Or I could use an AUX envelope?

edit2: basically the idea is to retrigger a sample as often as I want per given interval and to control it's playbacklenght. The variable starting point should be no problem as this could easily patched to a midi cc.
I had no luck with the envelopes so far, as I don't understand them. Can I trigger a sample via midi and it will play any given lenght via envelope and then immediately shut off? I'm wondering if the sampleretrigger will retrigger it's envelope aswell...

ja-ki wrote:BUT, If I turn down the rate of LFO1 while in the +1 phase, I could get PWM when I'd use LFO2 with a saw controlling LFO1's rate. Is this correct?

Are you sure you know what PWM really is????

On the Emu PWM can be simulated with single cycle waveforms that have been modified.
Example
Square wave ___----
Modified square wave ___---------

By adding extra positive phase to the sample but then creating a loop that is the same length as the first unmodified square wav it is possible to move the cycle start point of the loop using S/loop as a destination in cords.
The source can be either a
1. Static value
This allows the shape of the square wave to be set from 50/50 pos/neg through to 95/5 pos/neg.
2. Dynamic value
These could be LFO or envelopes and they move the s/loop start position during sample playback.
Moving the sample loop point during playback changes the pos/neg ratio and pulses the waveform shape THIS IS PWM

Due to the limitations of the Emu it is only possible to do this with micro samples.
It is not possible on longer samples that you may want too stutter or glitch.

yes I know what PWM is, but were talking different things here :)
I guess your talking about modulating an audible waveform (a sample), I'm talking about modulating the waveform of the lfo!
I'm gonna try this in a few days with an arpeggiator, upload a sample (if the whole idea works!) and let you have a listen.

cheers!

@Jahfunk:
Regarding the "steps in micro samples" you can use in the Cord settings several times the "Source -> Destination" combination "Gain4X -> Gain4X" @100% to create larger multiplier steps, like:
4 x 4 = 16
16 x 4 = 64
64 x 4 = 256
256 x 4 =1024
1024 x 4096
4096 x 4 = 16384
16384 x 4 = 65536
65536 x 4 = 262144
...
Yes, you have to victimize a few Cords to get a larger step-size.
At first you have to calculate the step-size you need, then you can setup the cords.
You have to keep in mind that the initial value (first modulation source) that has to be assigned to the first Gain4X and then you start with the multiplication.
For further adjustments you can use "DC Offset", "Summing Amp" and "Absolute Value".

Okay, no longer wait here is what it would sound like.
I realised this with an arpeggiator.
there are three values being automated:
arpeggiator rate (how often the sample will be retriggered)
arpeggiator gate (how long the sample will be played until next retrigger)
samplestart

Is this is correct?
LFO 1 will re trigger (stutter) the sample
LFO 2 is set to alter LFO 1 rate (stutter speed)

not exactly. LFO2 Rate (with a ramp down wave) would control LFO1 Rate while in the up-phase of the square wave. But as you can hear in the example an arpeggiator will do just fine