Planning the ultimate tape-sim box

[cakeworks]

Hi guys! Long time - no see.

None of you will remember me but I used to hang out here years ago when i was first tinkering with a soldering iron. My diy handiwork hasn't come far since then (but my music has, which is what matters right?)

Anyway. For the last few years recording I've been playing a lot with tape. The real stuff when I can get my hands on the gear, but then when I can't I try to get as close as possible digitally. My favourite aspect of tape (apart from a dead snare recorded super hot to a big reel-to-reel) is the shitty pitch modulation which happens as the speed varies. I've got it down to a point where I'm happy on the computer, but getting it live on my guitar is proving to be more difficult than I had first thought.

I am hoping to design and build something to fill this gaping hole in my sonic soul. I've built a few pedals and amps in the past but thus far the design process has been foreign to me. I'm fully prepared to take a long time to do this. I guess I'm here looking for some advice starting out on a project like this... please 

The pedal will probably be large-ish incorporating several independently switchable circuits. First will be a bandpass filter and compressor of some kind - haven't put a lot of research into this aspect of it yet. I will before I seek any advice on it.

But the main stage in this to me will be the vibrato/pitch stage. Here is a quick sample I knocked up using a piano in Logic. It's simply a compressor and logic's 'tape delay' set to 0ms. Here's a picture of what it looks like. I've outlined in red the pitch stages.

You can hear it here http://soundcloud.com/skeletonjack/tape-sim-piano-demo I also use the effect (with a bunch of other crap) on the guitars in the song 'wonderful world' - also on that soundcloud.

I've tried the zvex instant lofi and a couple of vibratos which are fun but none quite nail that tape irregularity (the pitch pirate is super cool but still has a regularity to it). The most common thing I find is that the 'wow' is quite achievable with a very slow pitch vibrato but the 'flutter' is a lot trickier as it is generally a (seemingly) random short spike or dip in the pitch.

So I'm guessing I'd need at least two unique pitch vibrato circuits. A basic (but super slow) one for the main pitch warbling... and then a second circuit with an extreme asymetrical square wave (I'm thinking i'd use two or more LFO circuits in parallel to achieve this - as per the 'pseudorandom lfo' article on geofx - although i'm not sure exactly how to work it out).

Ideally i'd like both vibrato circuits to be able to go as slow as 0.1hz

What do you think guys? Are you still reading? Am I crazy? Am I barking up the wrong tree? Is this doable? I'm particularly interested in the lfo side of things... whether it's achievable with analogue circuitry or if I'll need to start looking into dsp and programming custom waveforms (is that even possible? i've tried researching it but I'm way too out of my depth whenever I find myself in forums for it).

Anyways... thanks for any advice you might be able to give.

Jack

[~arph]

Piet has a lot of info on W&F on his site http://www.echotapper.nl
Do a site search on wow and flutter

Here is a direct link to a microcontroller based schematic:

http://www.echotapper.nl/wp/?p=568

[runmikeyrun]

I am the last person you would want to use as a reference when it comes to delays, but i'm thinking maybe your flutter can be controlled by a momentary stomp switch?  Step on it and hold for a second to get a dip and then let off?  Or maybe control by a LDR that you can wave your foot over while playing, or that flutters with flashing stage lights?

[Processaurus]

I am the last person you would want to use as a reference when it comes to delays, but i'm thinking maybe your flutter can be controlled by a momentary stomp switch?  Step on it and hold for a second to get a dip and then let off? 

Cool idea.  Having a couple momentary switches, like one for that, one for a runaway feedback setting would be cool.  A lot of effects are just left one because performing musicians are busy doing other things, but having momentary effects available like that is musically pretty interesting.

[peterv999]

In most cases 2 flangers or even 3 flangers would be used to achieve wow and flutter. It's related to the 'tape' mechanism so if you like to simulate a specific machine you need to know the mechanics and the IPS speed of the tape or disc. In DSP's like the FV-1 it's rather easy to build using LFO's controlling delay buffers. The ratio between the W&F and the dry/wet sound components is not very clear. I believe this deviation of WOW should be less then 20 cent at 1Khz and flutter less the 20% of the WOW amplitude as some starting values.

Piet

[cakeworks]

Thanks guys, sorry it's taken me so long to reply. Just moved to New Zealand and was on tour for a couple of weeks.

And thanks Piet, that'll take me a few rereads to get. Your site is introducing me to a world that I wasn't even aware of until now.

Just to clarify as there seems to be some confusion... this wouldn't be a delay. I'm not talking about warbling repeats (I have a smmh for that). What I mean is a complex wave on a vibrato circuit - essentially... This confusion may have come from the picture of my logic delay. Please note that a) the delay time is 0ms and b) the wet is at 100% with dry at 0%

Flangers are essentially choruses with a slightly increased delay times, right? Are there flangers with wet/dry functions?What you're saying makes sense to me from this angle piet. I've found I can get something like what I want when I put my chorus and vibrato pedals on their slowest settings in series... but it's still chorused and lacks the organic vibe that an 'unpredictable' (read: too widely spaced for humans to pick up a pattern) flutter would bring.

Thanks for your patience. I'm just getting back into this game.

Jack

[peterv999]

Jack,
The 'main problem' with wow and flutter emulation is that it all evolves around time '0'. Just visualize a tape movement being altered by capstan-rubber roll putting pressure on the capstan,some idle rolls and you would recognize this 'around zero behavior'. In electronics this is normally solved with a rather small delay being your flanges where '0' is in the middle of that delay buffer or with, for clock driven stuff like BBD's, PT2399, with clock variance if the device accepts this.The FV-1 DSP unit from Spin is fully capable creating this and  much more;-). My eTap2hw design, available as a DIY kit, based on component cost only!! Caters for that already.

Cheers Piet

[cakeworks]

Piet, is it ok if i dm/email you some questions?

Cheers