tascam tape delays

recording tape delays
recording tape delays

A year or so ago, a friend of mine sourced me a beautiful Tascam 34 four-track reel-to-reel tape deck. The idea was it was a loaner, as my slow crawl from software to hardware in my musical adventures had made me analog-tape-curious. A while later, he himself made the move from recording to tape, to using a UA Apollo and the various UA plugs that come with, and no longer saw the need for these decks he had, so he offered the Tascam to me to have and to hold.
When I’m practicing guitar, or noodling with the modular synth, the deck sits at eye-level on my desk, on top of my Motu interface, and just kind of stares at me. It looks like a bug-eyed robot. I haven’t used it nearly enough, and until recently I wasn’t really sure what I’d use it for anyway. Yes yes, I can record on it. And it sounds really good. A natural compression with a little hiss and a lot of fun control with the varispeed… But I just hadn’t really got it going.

Last week I stumbled upon a video:

And it occurred to me that I have everything I’d need to do this. Well, actually, I didn’t have a mixer, as I’d just given mine to a different friend, and I’d not yet replaced it with the Mackie that I’d ordered. But I do have Ableton and a good audio interface (Motu 828mk3) with enough ins and outs to be able to easily route this. So that I did.

For a couple of years, tape delay emulations have been my favorite type of delay effects. I love the sound of the degraded signal after the first couple of echoes. My Strymon El Capistan is my desert island stomp-box, being the one I’d keep if I had to get rid of all of the others. I use the GSI WatKat free plug-in more than I probably should. And a year ago I created a chain of Ableton effects to simulate an over-saturated tape delay in Live.

But routing this delay up through my deck is when I finally really understood exactly what’s going on and why. I have to admit that, frankly, it’s sort of a pain in the ass to set up, and the sound isn’t really any more amazing than I get through the El Capistan. It certainly doesn’t have as much control (no moving heads, no looping) but with EQ, it definitely has the tone control and when it starts howling with feedback, it’s sublime.

I recorded a half hour of noodling with an electric piano in live (via MIDI keyboard) and another half hour of guitar playing, direct into a channel in Live and using a few plugs like the amp/cab in Ableton and U-he’s Runciter, which is a fantastic filter/fuzz plug-in. I edited the 76 minutes down as much as I could, and the result is this track.

The dry signal is panned left and the delay from the tape deck is panned right. I ran the signal from the piano and guitar into a send track in Ableton, the send track was sent directly to the tape deck’s input, and the tape deck’s output, the delays, went to a second input channel in Ableton. As the video shows, this gets a simple slap-back echo. It’s not until you use the aux send on this second delay track that the fun begins with feedback and multiple echoes.

The next day I took some bits of this recording, reversed them and slowed them down, playing with Ableton’s warp modes. This became the track “outside over there.” The sound of a reversed delay is one of my favorites and just makes the piano and guitar melt and weave in a way I really like.

Look forward to more of this kind of thing.

testing 142

I’m constantly figuring little things out about this synthesizer and I feel, in many ways, that in the five months I’ve had this modular thing I’ve learned more about synthesizers and sound than in the years and years previous working with software.

I recently added a Doepfer A142 to my system. This is a voltage controlled decay/gate. What it does is takes a trigger input, like from a LFO or in my case the MFB Seq02, and creates an envelope that consists of a very quick attack and a voltage-controlled and adjustable decay. What I was looking for was a way to create steps of different lengths from the sequencer to add some movement and, you know, funk. I could do this with my Model 10, but I like using the Model 10 so much that I figured I’d like to have something that is dedicated to sitting next to and serving the sequencer.

Soon after plugging it in, I started noticing that something wasn’t right with the A142. It didn’t have a “quick” decay, only a really muddy longish one. This was especially the case with manual control. When I plugged the CV in, I could get sharp clicky envelopes, but in manual they were anything but short even with the knob at 0. I discovered that on the PCB of the module there was a small trim pot, but I didn’t know what it did as the Doepfer site didn’t seem to refer to it.

So, as I do at times like this, I turned to the internets and started a thread at the Muffwiggler forum. I immediately got some help and much of it was quite thorough and technical. Just having others tell me that their A142’s work as expected at least narrowed it down to a problem with my particular module. One or two readers actually got deep into it and with this information I at least got the thing usable. The trimmer on the back adjusts the offset, which means that it adjusts the point at which the envelope starts getting longer. By turning it about a quarter-turn, it’s now set to where from 0 to 1, the envelope is about 1ms long. This is a good thing. A new problem begins at 1, because by 3 the envelope is more than 12 seconds. This in theory should take place at nearly a complete turn, like around 9 or 10. So the entire range of the envelope is now between 1 and 3, which means that very very small nudges to the knob result in very large changes to the length of steps from the sequencer. From incredibly staccato to very very legato.

To illustrate this, I made a video. It includes captions for clarity, but basically I set up a simple sequence in the Seq02 and make adjustments with the A142. You can see the minor little changes to the knob and hear the major adjustments to the step lengths. The advice from the forum was to solder a new resistor onto the PCB. I’ve been meaning to learn to solder…


Got a couple of new modules this week. The MFB Dual ADSR will replace the Doepfer A140. It’s got two ADSR envelopes, each with a “hold” state as well as the expected attack, decay, sustain and release. The envelopes can be triggered individually or together, and they each have two outputs. This is all great.
The Bubblesound uLFO is at its heart, as its name suggests, a low frequency oscillator. But it’s more than just that. It can go really really slow — something like 20 minutes per cycle — and it can be tracked at 1v per octave at audio rate, which basically makes it a secondary VCO. I’ve been wanting to learn more about FM synthesis and this is my first module that has a linear FM input. Both of my VCOs, Doepfer A110’s, only have exponential FM (at some soon point I plan to replace the 110s with a Cwejman VCO6 and a Harvestman Hertz Donut but they’re hard to come by right now).

uLFO and MFB dual ADSR

One funny thing about the uLFO. On Monday I noticed that the response of the SIne Shaped output seemed odd. This is a unique feature of the uLFO. It’s got a regular sine output but it’s also got a “shaped” sine, which basically creates what the Bubblesound website calls positive/negative biased non-linear triangles. The wave forms look like waves as drawn by a kid on one side, and bumps or the McDonalds logo on the other. However, I noticed that the shapes it created were the opposite from the icons indicated on the panel of the uLFO. I made a little movie to send to send to Bubblesound to make sure that I was hearing it right.

I posted this to the Muffwiggler forum, which got mixed replies. To make sure I wasn’t going nutty I ran the waves through my CueMix oscilloscope which proved that yes, I was hearing it correctly. Later, David at Bubblesound wrote and confirmed my observation. Regardless of this little quirk, this is a terrific module and I look forward to posting more here from the uLFO.

i got a sequencer

I have sequencers running out my ears on my computer. I regularly use Numerology, Reaktor, Ableton and Volta to sequence soft synths and, lately, my modular via midi. However, a good part of the reason I’ve become sort of kind of obsessed with this modular business is the tactile control of buttons, switches and knobs.
To that end, I added a sequencer in the form of the MFB Seq02 last week. This thing is incredibly versatile for such a small thing. It can run twelve separate CV sequences of up to 32 steps each, or six CV and six trigger/gates. It’s got its own clock and it can be controlled by an outside clock as well.

I’ve not had much time this week to work with this thing, but today I set up a simple four note repeating sequence and just started patching. These are variations using pretty much the same four note patterns. Each track on the Seq02 can have one of either an A or B 8-step sequence, or both in a row. A three-way switch chooses which mode. So I just made two different four-note sequences, started the sequencer, and switched back and forth.

The sounds are coming from the A110 VCOs in various waveforms, and filtered at different times with the Model 12, the A120 VCF, and the A101-2 Low Pass Gate. The A101-2 is the thing that creates the plucky acoustic sounds. When you hear it you’ll know. I used MOTU’s Volta to send various trigger and step-sequences to both the Seq02 as a clock, and other modules for various modulations. I sometimes wanted various numbers of steps, where the Seq02 only allows one to choose steps on a global level. Meaning if I want one sequence of triggers to be four steps and another to be seven, I have to rely on the software.
The delay was added in Ableton because it sounds nice.

a little sequencing?