After a discussion about very fast blips and clicks with someone, I thought t a good idea to share this for inspiration.

Very sharp blips can be easily constructed with a 6dB HP filter followed by diode module. On an analog one would get this kind of fast pulse by eeding a gate or squarewave LFO through a small capacitor. Connecting the ate signal to both inputs of an analog ringmodulator mostly did the trick s well, as these were mostly AC coupled with capacitors.

Advantage over a NM AD envelope is that the complete signalpath of the pulse fter the HP filter is a red signal. This gives very fine tone control over wide range.

In the patch the crossfade mixer between two 6dB HP filters controls the ulse length. After modulating the level of the pulse it is fed into a grey nput of a sine slave, the DSP-cheapest way to get a traditional blip. The lave squarewave oscillator yields excellent results as well, a bit like the ld arcade game blips. Other oscillators, filters, etc. of course also possible.

Try the six assigned knobs to hear the controllable range.

The 12 dB SVF is a nice addition. Shorter pulses give a bass sound, if the pulse gets longer the filter is swept longer in the high region and creates the plastic click.

I'm still trying to figure out what this thing is doing... using filtered event sequencer pulses to generate blips, which eventually control the amplitude of the main osc? And the crossfaders control the relative levels of syncopated pulse patterns? In any case, this patch is too much fun. I played around with it for a looong time last night. I think I owe my neighbors an apology.

A little graph plus patch to explain, the top is the pulsetrain from e.g. a squarewave LFO or event sequencer. The middle is the signal when HP filtering with the 6dB filter. The bottom is the signal after the diode module. As you can see a very sharp AD curve, the attack being absolutely immediate and the decaylength depending on the cutoff frequency of the HP filter, from short at 12 Hz to very, very short at very high frequencies. Maybe you remember me talking about transients in waveforms on the NM Day The Hague, this is a typical example of 'extracting' the transient from a waveform. Basically it contains the dynamic energy of the squarewave, as this is the energy that e.g. swings a conus from a loudspeaker from back to forth when fed with the square LFO. Or the click from the speaker, so to speak. So now we have a short pulse with very much energy in it, enough to e.g. sweep an oscillater or excite a high resonance filter or resonator. There is even a synthesis method that, when given a filter/resonator with a certain response, extracts a sample of pulses that, when fed into this filter/resonator, resynthesises the original signal. The fun then starts when changing the rate of the pulsesample before being put into that filter/resonator again. I believe its available on the Kyma system. But the principle can be put to good use on the NM. One of the reasons why I would very much like the 6 dB filters to have a control input, to control the decal lengths. Now I simply generate a short and an even shorter pulse and crossfade between the two. I think it would be really nice and sufficient for the 6 dB filters to only have a grey input that could be controlled by e.g. the master oscillator module, or it could be made to track a master oscillator. This also as the grey input has the appears to have the same samplerate as the red signals, blue signals would be to slow to precisely control the 6 dB filters in this kind of applications.

In the same way as the HP extracts the pulse the 6 dB LP can be used to slow only the attack a bit, as is used in the HH sound. Its only the 6dB HP filter that works nice in this case, other filters simply do too much to the signal making it swing a bit around the zero line, giving a less usable pulse.

In the patch you can hear the difference between using the 6 dB HP filter to generate short pulses or the AD Envelope module. When the switch is set to input4 you hear how the pulse can excite a very simple resonator. Basically all the sound energy thats at the output of the resonator is received from the very short pulse, from nowhere else.

And a noodle...

Here's a patch which uses Rob H's recently posted idea for a very fast transient generator (gate or pulse through 6dB/oct HPF through diode). As Rob described, using audio-type modules instead of an EG creates very fast transients that have a smooth log decay shape, because these modules use a much faster clock rate than control modules like EGs.

In addition to the snappy sound (second column in the patch) I added a couple of other parallel sounds (third & fourth columns) with separate volume controls (knob 1 & 2). The transient trick can help make a sound seem much more like a real analog synth, or a real acoustic plucked sound. I gotta go make some plucked noises.

I was playing around with Rob's noodle, trying to figure out (among other things) how he got it to sound like two pulse trains (duh... 2 voices, always check the simple things first). Anyway, I started listening to some of the control signals, and playing around, and came up with this. It's still basically Rob's noodle. I just added some low frequency, panning, and mixed back in some of the 'scraped string' control signal. Pitch following is from Git Synthesizer by Kees van der Maarel.