Monthly Patching Workout 2017

And the Patching Workout continues. So far, I started the year with a no input reverb / resonator feedback loop, so let’s continue that tradition. I wanted to give this a try with the Space Echo and Cwejman QMMF4, so here is the RE201 QMMF Loop.

The patch is fairly simple. For the feedback loop, we have the QMMF4 main MMF output going to a Warm Audio WA76 for some compression. From there, it goes to the RE201 and then back into the modular to the QMMF main audio input (using an Intellijel Audio Interface for going in and out of the rack). The internal RE201 feedback is not used, a slight bit of the RE201 spring verb is added. The main QMMF notch out is used for recording. It’s routed through a Doepfer A199 first for some further spring reverb outside the feedback loop, into a GAP Pre73 DI input for some gain (as it was a bit quiet after adjusting everything nicely), and finally to a Mutator to smooth out the highs a bit and for some mono to stereo conversion and LFO panning (also a bit of envelope following on the filters - you can hear it all working on the noisefloor, haha). It was harder than I expected to get the gainstaging right without killing the loop or just utter destruction, but in the end I got something that I enjoyed. 

In terms of CV, we have a Cwejman RG6 at the center of everything. A bit like the Reverse Cwejman Krell patch, the RG6 provides the main clock and triggers a CTG-VC, which controls the main QMMF VCA (and thus the dynamics within the feedback loop). The RG6 random saw modulates the RG6 clock for the bursts; to get the dynamic envelope, the inverted random saw is attenuated on a VCA4MX and controlling the CTG attack and decay rates. Thus, the envelope times get faster when the clock gets faster. The normal S/H output of the RG6 modulates the main QMMF frequency. 

Finally, a PH8 controls all individual QMMF frequencies with the sine outputs and two of the QMMF bands with two (90° spaced) saws. It’s synced from the RG6 gate with gate sync, creating kind of a “LFO delay” effect (in combination with the gate length on the RG6). The PH8 is FMed from the RG6 random saw so that it also gets faster according to the clock and envelope. The quadrature sines produce the “delayed” sweeps and the saws the plops.

On the recording, I mainly tweak the QMMF frequencies and resonances (all bands are set to BP filter and in resonator mode to see how it compares with the RES4). Little bit of cutting out some boring things afterwards and some reverb, delay, EQ and compression as usual. 

Hope you enjoy it and cheers!

Jan 17: RE201 QMMF Loop

This month we have something slightly different than usual. For a change, I wanted to do something with the Roland V-Synth. I really enjoy sampling raw QMMF4 tones into it to create pads, the strange digital mangling the V-Synth can do extracts some beautifully haunting artefacts from the analog sources. Here is the V-Synthesized QMMF.

I cannot really describe all the samples in detail, as I’ve done lots of them over time… But regarding the source, they are all based on using the QMMF as an oscillator, usually cross-FMing the different bands. It’s five layered V-Synth patches in total and all of them play with the time and formant settings on the V-Synth PCM oscs (speeding the samples up and down, LFOs on the formant setting, etc.). I played all the loops manually, but the quarter notes of the bass are done by looping a sample - only the transpositions are played. After recording and arranging, I sent four of the five pads (excluding the bass) through a Mutronics Mutator for some stereo filter action on the whole thing. 

Finally, I had to add a ping track to give a bit of (admittedly wonky) rhythm to the pads. The trigger is provided by an 808 and multed to a Cwejman CTG-VC (to create a nice short trigger for pinging) and the sync input of a PH8. The CTG pings a RES4 (the straight 4/4 pattern), while the 90° and 180° saw outputs of the PH8 each ping one QMMF4 band. Those signals are mixed with a VCA4MX and further sent to a SPH2 and then stereo out to the DAW. Both the QMMF and the RES4 are FMd a tiny bit with a PH4, and the 90° and 180° triangle outputs of the PH8 gently modulate two RES4 bands. 

That’s it basically, got a bit mushy all together, but I guess I overdid it with the layering anyways :) That can often happen when layering the same synth over and over, IME. The V-Synth layers very nicely with my analog polys, but that’s not for this patch ;) To me, it truly is the modern D50, and it also helps that there is an actual D50 included in it :D

As always, a ton of delay and reverb (Valhalla VV) in Ableton, as well as some EQ and compression to bring it up a bit. Hope you enjoy it!

Aug 17: V-Synthesized QMMF

This month we have a patch using two unrelated square LFOs for trills on two VCOs. One is static and the other modulated. Here are the Dual Trills.

First, we have an Addac 701 as the first LFO doing a minor third on a Cwejman VCO6 (using the square 1 output). The 701 square also clocks an Audio Damage Sequencer 1 (I use the inverted output on the 701 so that the sequencer advances on the lower note of the minor third). Basically, the sequencer is just used to create random modulation values from the internal LFOs. The gate output also triggers an ADSR-VC2.

Then, we have a second 701 and also a second VCO6. The 701 again is in LFO mode and creates octave trills on the VCO6 (where the saw “PWM” wave is used). Both VCOs are then routed to a MMF6 and then to an Addac 802 for the final VCA and out to the DAW. The 802 VCA is gated by the ADSR-VC2, while the second envelope on the ADSR modulates the MMF6. Furthermore, there’s FM on the MMF6 from the octave trill VCO, a slow LFO from a Doepfer A143-3 and “key tracking” from the square LFO doing the minor third trill. The A143-3 also provides PWM to the two VCOs and vibrato on the octave trill VCO. Finally, as mentioned, two S/H LFOs are created internally on the clocked Sequencer 1 (using the + / 0 / - setting, which basically provides three different values that are chosen randomly). The first randomizes the ADSR decay, the second the frequency of the LFO doing the octave trill. Finally, the A143-3 LFO doing PWM on the minor third trill VCO is used to PWM the LFO doing the octave trills. 

The patch is also very much based on playing with the level of the filter audio inputs. The buildup takes a bit of time. I recorded everything in one take, no edits needed. Couldn’t resist adding a return track from a RE201 after the fact (you can hear its noisefloor rather well in the quiet beginning). As always, some delay, reverb (Valhalla Vintage Verb in plate mode), EQ and compression in Ableton to bring it together.

Hope you enjoy it and cheers!

Sept 17: Dual Trills

I wanted to do a variation on the “original” “shimmer” reverb as used by Brian Eno and others, with the gear I have. The MXR129 has made it into the Patching Workout a few times already, let’s give it another go for this one. That makes the Andromeda Shimmer Piano.

As sound source, I’m doing a variation on the clever JX10 piano patch (https://www.soundonsound.com/techniques/synthesizing-acoustic-pianos-roland-jx10-1102) with the Alesis A6. It’s based on using static hard sync to shape the timbre of the “piano”. It’s really more like an electric piano, but still a pretty decent patch - good start to explore sync as well. So, on the A6, the slave is 21.5 semitones above the master, mixed more or less equally in volume. The filters are used in parallel, one to create the “bite” and one the “growl”, if that makes any sense. On the envelopes, I programmed sort of a strange release that kind of “jumps” up a bit again shortly after having released the notes. Finally, some keyboard tracking on the filters and envelopes, plus a touch of velocity on the filter and VCA envelopes. Then I manually recorded two minutes of a pretty cheesy loop to the DAW, for which you will have to excuse me :)

Then, on to the effect. I didn’t want to go completely overkill, so I decided to restrict it to two sends. The piano recording is routed out to my Midas Venice 160 mixer on an “unconnected” channel that doesn’t get sent to the master (to be able to record wet only). Then, the send 1 + 2 are used for the two effect chains. Both are routed back to the mixer for some feedback and cross-feedback (does that term exist?) action. All three channels on the Midas are EQd quite radically to shape the effect return sound.

Firstly, the “shimmer” needs a bit of modulation to wash out the effect return. I use a recently acquired Boss RCE10 Microrack chorus, set to a subtle and slow-ish modulation. That goes to the pitch shifter, set to shift an octave up, and finally to a reverb, an Alesis Quadraverb+ set to a rather long hall. I know that the pitch shifter should rather be placed after the reverb, but the old MXR has quite a bit of delay in the processing, so I take that as a free “pre-delay” before the reverb. I tried both way and found it more fun this way. The feedback using the mixer does its thing either way.

Secondly, some delay is helpful for these types of effects. For the second effect chain, I use an old Yamaha E1005 BBD delay in series with a Soviet Lell RC, rather strange box that latter one :) The E1005 does pretty straightforward delay with a tiny bit of modulation for some chorusing and the Lell a strange reverb-ish chorus type thing. 

For the recording, the reverb and chorus of the first chain are there from the beginning, the second chain is muted. On the first chain, the “shimmer” is down first and I dial it in slowly using the dry / wet control on the MXR129. Then, the second delay chain comes in. In the “break” (without chords), I dial up the internal feedback on the MXR (“Regeneration”), because that is always fun. Generally, as you can easily hear, the feedback and cross-feedback on the mixer effect returns are dialled in more and more during the recording. The E1005 got me once with a bit of overdrive… Then, the whole process is repeated sending the recorded effect returns out again, doing slightly different fader actions on the mixer and with a different delay rate on the E1005. One could do this over and over again when used subtle, but it ended up rather over the top like this already :D 

No actual reverb and delay from the DAW this time, I think that might be a first for the MPW. However, I had to record a Roland RE201 track over the whole thing, because it’s just too nice. That’s the delay that is there right from the start. However, some EQ and compression in Ableton to bring it up.

Hope you enjoy it, cheers!

Oct 17: Andromeda Shimmer Piano

This month we have a self-playing feedback patch creating a sci-fi atmosphere. It’s based on injecting a dynamic noise into a feedback loop, controlled by a cycling Cwejman CTG-VC and RG6.

To route the feedback loop, a VCA4MX is used. The mixed out 1+2 goes to the DAW for recording, the sum out is used for the feedback loop. As mentioned, a noise source is used to excite the feedback: The S/H wave of a D-LFO is routed into a QMMF4 channel, and then into channel 1 of the VCA4MX. The QMMF4 is set to a simple LP filter with little resonance (in RES mode). For the feedback loop, the sum of the VCA4MX is routed to a Doepfer A199 spring verb first, then to a RES4, to a SPH2, and finally back into the VCA4MX channel 2 input. That’s it for the audio routing.

For the self-playing modulations, a RG6 and cycling CTG-VC are the base of the patch. The RG6 clock triggers the CTG and there’s a bit of self-modulation in the RG6 from the internally patched routing (from the random saw). The CTG opens the VCA4MX and QMMF4 channels with the noise exciter source. Furthermore, another QMMF4 channel is used as a sine osc to FM the RES4 master frequency, brought in and out manually with the output control of the QMMF4. 

The various RG6 outputs modulate different parameters in the feedback loop: The low S/H out modulates the frequency of an individual RES4 channel; the normal S/H out controls the frequencies of the D-LFO pitched noise, QMMF4 FM source, another RES4 channel, and the CTG attack time. The high S/H out modulates the CTG decay time, and the frequencies of the QMMF4 noise source filter and both sides of the SPH2. Finally, the RG6 random saw out modulates a further individual RES4 frequency, attenuated on a second VCA4MX. That’s almost everything, but to bring some more movement into the patch, a PH4 is used to crossfade between two RES4 channels (on the internal RES4 VCAs) with two sine 180° apart in phase. The other two PH4 outs modulate the hold time of the CTG and the frequency of the left SPH2 side. Lastly, a feedback CV path is brought in by patching the RES4 notch out to the frequency of the right SPH2 side, also attenuated on the second VCA4MX.

Pretty much all time parameters in the CV path are set very slow and some slew is used on the RG6 S/H outs. In the recording, the first ten seconds or so are only the pitched noise source, then I slowly bring in the feedback on the main VCA4MX channel and take it from there, mostly tweaking the various filter frequencies in the audio path and the FM amount on the RES4. Recorded in one take, no edits.

As so often, a RE201 return track was added as well as some EQ, compression, delay and reverb (Valhalla VV) in Ableton. That brought up some noises that I didn’t hear before, not sure but I suspect the Stackcables are to blame… Hope you enjoy it anyways and cheers!

Nov 17: Feedback Spacescape

This year’s last patch is a combination of using the Cwejman QMMF4 as a sine oscillator bank, with square LFOs providing tuned trills and feeding everything through a phaser with feedforward modulation from the audio source on the phaser frequency. Here’s the Multi Sine Phase Modulation :)

All four QMMF4 bands are put into self-resonance (using the osc mode), the main MMF out is then routed to the SPH2 phaser. The SPH2 L and R outs go to the DAW, the MIX out goes to a Doepfer A199 spring verb and is then mixed with the L and R (via a VCA2P). The first band of the QMMF4 is tuned low, with a D-LFO square providing an octave trill. Additionally, there’s some FM from the second QMMF4 band, which is tuned higher and just drones away without an individual trill. The third band is also tuned high and receives a fifth trill from the second D-LFO square. It is also slightly FMd from the first band. Finally, the fourth band is tuned very low with FM from the third band. Also, there is a third slow square LFO (from a PH8) transposing the whole osc bank via the main QMMF4 freq input, also tuned to an octave. All the trills are carefully tuned with a VCA4MX.

At the center of the patch is a RG6 driving the whole thing. Firstly, the gate out is used to sync the first D-LFO square with the SYNC G input, with the RG6 gate length carefully tuned so that the LFO is shortly cut off at the beginning of each sync hit. The RG6 clock out triggers the second D-LFO and the PH8 on the normal SYNC T inputs. Furthermore, the RG6 low S/H out modulates the frequency of the first D-LFO a tiny bit, the high S/H out modulates the second D-LFO. Finally, the mid S/H out modulates the level of the second QMMF4 band and the random saw out modulates the level of the third QMMF4 band, attenuated slightly via the last VCA4MX channel (all self-FM on the QMMF4 is via the pre VCA outs, so they are not affected by these amplitude modulations). 

Finally, for the feedforward modulation on the phaser, the main QMMF4 notch out is routed to a second VCA4MX. The RG6 clock is further multed to a CTG-VC, which gates the VCA4MX. The VCA4MX out then modulates the SPH2 L + R frequencies. The patch was a bit too boring and I wanted to add some harmonics, this trick is always nice on the SPH2.

For the recording, I mainly play with the QMMF4 and CTG levels. Firstly, all levels are down and only the random saw brings up the third band in the background a few times. Then I slowly bring up the first band, always playing with the feedforward phase modulation. Then, the third band is brought up properly and the second one is faded in slowly. After that, the first band (with the higher bass) is muted, the second and third are lowered a bit and the fourth low bass is brought in. From there I just jam with everything. In the end, I quickly bring up the PH8 frequency to audio rates and then back it off again. Finally, the manual six semitone transpositions that appear a few times are done by pulling the plug on the main QMMF4 frequency in and then inserting it again :D

All recorded in one take into Ableton without any edits. As so often, I added a RE201 return track after the fact. Furthermore, some reverb (Valhalla VVV), delay, EQ and compression in Ableton to bring it all together, as always.

Hope you enjoy it and cheers!

Dec 17: Multi Sine Phase Mod

Booking & enquiries: florin.buechel(at)gmail.com

Contra Communem Opinionem
Savage Grounds
Wavetest

Jan 17: RE201 QMMF Loop

C C O